0% found this document useful (0 votes)
18 views

101174

Uploaded by

jamieemabesi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
18 views

101174

Uploaded by

jamieemabesi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 47

Get ebook downloads in full at ebookmeta.

com

Essentials of Compilation An Incremental Approach


in Python 1st Edition Jeremy G. Siek

https://ebookmeta.com/product/essentials-of-compilation-an-
incremental-approach-in-python-1st-edition-jeremy-g-siek/

OR CLICK BUTTON

DOWNLOAD NOW

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


Essentials of Compilation
Essentials of Compilation
An Incremental Approach in Python

Jeremy G. Siek

The MIT Press


Cambridge, Massachusetts
London, England
© 2023 Jeremy G. Siek

This work is subject to a Creative Commons CC-BY-ND-NC license.

Subject to such license, all rights are reserved.

The MIT Press would like to thank the anonymous peer reviewers who provided comments on
drafts of this book. The generous work of academic experts is essential for establishing the
authority and quality of our publications. We acknowledge with gratitude the contributions of
these otherwise uncredited readers.

This book was set in Times LT Std Roman by the author. Printed and bound in the United
States of America.

Library of Congress Cataloging-in-Publication Data

Names: Siek, Jeremy, author.


Title: Essentials of compilation : an incremental approach in Python / Jeremy G. Siek.
Description: Cambridge, Massachusetts : The MIT Press, [2023] | Includes bibliographical
references and index.
Identifiers: LCCN 2022043053 (print) | LCCN 2022043054 (ebook) | ISBN 9780262048248 |
ISBN 9780262375542 (epub) | ISBN 9780262375559 (pdf)
Subjects: LCSH: Compilers (Computer programs) | Python (Computer program language) |
Programming languages (Electronic computers) | Computer programming.
Classification: LCC QA76.76.C65 S54 2023 (print) | LCC QA76.76.C65 (ebook) | DDC
005.4/53–dc23/eng/20221117
LC record available at https://lccn.loc.gov/2022043053
LC ebook record available at https://lccn.loc.gov/2022043054

10 9 8 7 6 5 4 3 2 1
This book is dedicated to Katie, my partner in everything, my children, who grew
up during the writing of this book, and the programming language students at
Indiana University, whose thoughtful questions made this a better book.
Contents

Preface xi

1 Preliminaries 1
1.1 Abstract Syntax Trees 1
1.2 Grammars 3
1.3 Pattern Matching 5
1.4 Recursive Functions 6
1.5 Interpreters 8
1.6 Example Compiler: A Partial Evaluator 10

2 Integers and Variables 13


2.1 The LVar Language 13
2.2 The x86Int Assembly Language 16
2.3 Planning the Trip to x86 21
2.4 Remove Complex Operands 23
2.5 Select Instructions 25
2.6 Assign Homes 26
2.7 Patch Instructions 27
2.8 Generate Prelude and Conclusion 27
2.9 Challenge: Partial Evaluator for LVar 28

3 Parsing 29
3.1 Lexical Analysis and Regular Expressions 29
3.2 Grammars and Parse Trees 31
3.3 Ambiguous Grammars 33
3.4 From Parse Trees to Abstract Syntax Trees 34
3.5 Earley’s Algorithm 36
3.6 The LALR(1) Algorithm 40
3.7 Further Reading 43

4 Register Allocation 45
4.1 Registers and Calling Conventions 46
4.2 Liveness Analysis 49
4.3 Build the Interference Graph 51
viii Contents

4.4 Graph Coloring via Sudoku 52


4.5 Patch Instructions 58
4.6 Generate Prelude and Conclusion 58
4.7 Challenge: Move Biasing 59
4.8 Further Reading 62

5 Booleans and Conditionals 65


5.1 The LIf Language 66
5.2 Type Checking LIf Programs 66
5.3 The CIf Intermediate Language 72
5.4 The x86If Language 72
5.5 Shrink the LIf Language 75
5.6 Remove Complex Operands 75
5.7 Explicate Control 76
5.8 Select Instructions 82
5.9 Register Allocation 83
5.10 Patch Instructions 84
5.11 Generate Prelude and Conclusion 84
5.12 Challenge: Optimize Blocks and Remove Jumps 85
5.13 Further Reading 88

6 Loops and Dataflow Analysis 91


6.1 The LWhile Language 91
6.2 Cyclic Control Flow and Dataflow Analysis 91
6.3 Remove Complex Operands 96
6.4 Explicate Control 96
6.5 Register Allocation 96

7 Tuples and Garbage Collection 99


7.1 The LTup Language 99
7.2 Garbage Collection 102
7.3 Expose Allocation 109
7.4 Remove Complex Operands 110
7.5 Explicate Control and the CTup Language 110
7.6 Select Instructions and the x86Global Language 111
7.7 Register Allocation 116
7.8 Generate Prelude and Conclusion 116
7.9 Challenge: Arrays 118
7.10 Further Reading 123

8 Functions 125
8.1 The LFun Language 125
8.2 Functions in x86 130
8.3 Shrink LFun 133
8.4 Reveal Functions and the LFunRef Language 133
Contents ix

8.5 Limit Functions 133


8.6 Remove Complex Operands 134
8.7 Explicate Control and the CFun Language 135
8.8 Select Instructions and the x86Def
callq∗ Language 136
8.9 Register Allocation 138
8.10 Patch Instructions 139
8.11 Generate Prelude and Conclusion 139
8.12 An Example Translation 141

9 Lexically Scoped Functions 143


9.1 The Lλ Language 145
9.2 Assignment and Lexically Scoped Functions 150
9.3 Uniquify Variables 151
9.4 Assignment Conversion 151
9.5 Closure Conversion 153
9.6 Expose Allocation 156
9.7 Explicate Control and CClos 156
9.8 Select Instructions 157
9.9 Challenge: Optimize Closures 158
9.10 Further Reading 160

10 Dynamic Typing 161


10.1 The LDyn Language 161
10.2 Representation of Tagged Values 165
10.3 The LAny Language 166
10.4 Cast Insertion: Compiling LDyn to LAny 170
10.5 Reveal Casts 170
10.6 Assignment Conversion 171
10.7 Closure Conversion 171
10.8 Remove Complex Operands 172
10.9 Explicate Control and CAny 172
10.10 Select Instructions 172
10.11 Register Allocation for LAny 174

11 Gradual Typing 177


11.1 Type Checking L? 177
11.2 Interpreting LCast 183
11.3 Overload Resolution 184
11.4 Cast Insertion 185
11.5 Lower Casts 187
11.6 Differentiate Proxies 188
11.7 Reveal Casts 190
11.8 Closure Conversion 191
11.9 Select Instructions 191
11.10 Further Reading 193
x Contents

12 Generics 195
12.1 Compiling Generics 201
12.2 Resolve Instantiation 202
12.3 Erase Generic Types 202

A Appendix 207
A.1 x86 Instruction Set Quick Reference 207

References 209
Index 217
Preface

There is a magical moment when a programmer presses the run button and the
software begins to execute. Somehow a program written in a high-level language is
running on a computer that is capable only of shuffling bits. Here we reveal the wiz-
ardry that makes that moment possible. Beginning with the groundbreaking work
of Backus and colleagues in the 1950s, computer scientists developed techniques
for constructing programs called compilers that automatically translate high-level
programs into machine code.
We take you on a journey through constructing your own compiler for a small
but powerful language. Along the way we explain the essential concepts, algorithms,
and data structures that underlie compilers. We develop your understanding of how
programs are mapped onto computer hardware, which is helpful in reasoning about
properties at the junction of hardware and software, such as execution time, soft-
ware errors, and security vulnerabilities. For those interested in pursuing compiler
construction as a career, our goal is to provide a stepping-stone to advanced topics
such as just-in-time compilation, program analysis, and program optimization. For
those interested in designing and implementing programming languages, we connect
language design choices to their impact on the compiler and the generated code.
A compiler is typically organized as a sequence of stages that progressively trans-
late a program to the code that runs on hardware. We take this approach to the
extreme by partitioning our compiler into a large number of nanopasses, each of
which performs a single task. This enables the testing of each pass in isolation and
focuses our attention, making the compiler far easier to understand.
The most familiar approach to describing compilers is to dedicate each chapter
to one pass. The problem with that approach is that it obfuscates how language
features motivate design choices in a compiler. We instead take an incremental
approach in which we build a complete compiler in each chapter, starting with
a small input language that includes only arithmetic and variables. We add new
language features in subsequent chapters, extending the compiler as necessary.
Our choice of language features is designed to elicit fundamental concepts and
algorithms used in compilers.
• We begin with integer arithmetic and local variables in chapters 1 and 2, where
we introduce the fundamental tools of compiler construction: abstract syntax trees
and recursive functions.
xii Preface

• In chapter 3 we learn how to use the Lark parser framework to create a parser
for the language of integer arithmetic and local variables. We learn about the
parsing algorithms inside Lark, including Earley and LALR(1).
• In chapter 4 we apply graph coloring to assign variables to machine registers.
• Chapter 5 adds conditional expressions, which motivates an elegant recursive
algorithm for translating them into conditional goto statements.
• Chapter 6 adds loops. This elicits the need for dataflow analysis in the register
allocator.
• Chapter 7 adds heap-allocated tuples, motivating garbage collection.
• Chapter 8 adds functions as first-class values without lexical scoping, similar to
functions in the C programming language (Kernighan and Ritchie 1988). The
reader learns about the procedure call stack and calling conventions and how
they interact with register allocation and garbage collection. The chapter also
describes how to generate efficient tail calls.
• Chapter 9 adds anonymous functions with lexical scoping, that is, lambda
expressions. The reader learns about closure conversion, in which lambdas are
translated into a combination of functions and tuples.
• Chapter 10 adds dynamic typing. Prior to this point the input languages are
statically typed. The reader extends the statically typed language with an Any
type that serves as a target for compiling the dynamically typed language.
• Chapter 11 uses the Any type introduced in chapter 10 to implement a gradually
typed language in which different regions of a program may be static or dynami-
cally typed. The reader implements runtime support for proxies that allow values
to safely move between regions.
• Chapter 12 adds generics with autoboxing, leveraging the Any type and type
casts developed in chapters 10 and 11.
There are many language features that we do not include. Our choices balance the
incidental complexity of a feature versus the fundamental concepts that it exposes.
For example, we include tuples and not records because although they both elicit the
study of heap allocation and garbage collection, records come with more incidental
complexity.
Since 2009, drafts of this book have served as the textbook for sixteen-week
compiler courses for upper-level undergraduates and first-year graduate students at
the University of Colorado and Indiana University. Students come into the course
having learned the basics of programming, data structures and algorithms, and
discrete mathematics. At the beginning of the course, students form groups of two
to four people. The groups complete approximately one chapter every two weeks,
starting with chapter 2 and including chapters according to the students interests
while respecting the dependencies between chapters shown in figure 0.1. Chapter 8
(functions) depends on chapter 7 (tuples) only in the implementation of efficient
tail calls. The last two weeks of the course involve a final project in which students
design and implement a compiler extension of their choosing. The last few chapters
can be used in support of these projects. Many chapters include a challenge problem
that we assign to the graduate students.
Preface xiii

Ch. 1 Preliminaries Ch. 2 Variables Ch. 3 Parsing

Ch. 4 Registers Ch. 5 Conditionals Ch. 6 Loops

Ch. 8 Functions Ch. 7 Tuples Ch. 10 Dynamic

Ch. 9 Lambda Ch. 11 Gradual Typing Ch. 12 Generics

Figure 0.1
Diagram of chapter dependencies.

For compiler courses at universities on the quarter system (about ten weeks in
length), we recommend completing the course through chapter 7 or chapter 8 and
providing some scaffolding code to the students for each compiler pass. The course
can be adapted to emphasize functional languages by skipping chapter 6 (loops)
and including chapter 9 (lambda). The course can be adapted to dynamically typed
languages by including chapter 10.
This book has been used in compiler courses at California Polytechnic State Uni-
versity, Portland State University, Rose–Hulman Institute of Technology, University
of Freiburg, University of Massachusetts Lowell, and the University of Vermont.
This edition of the book uses Python both for the implementation of the compiler
and for the input language, so the reader should be proficient with Python. There
are many excellent resources for learning Python (Lutz 2013; Barry 2016; Sweigart
2019; Matthes 2019).The support code for this book is in the GitHub repository at
the following location:
https://github.com/IUCompilerCourse/

The compiler targets x86 assembly language (Intel 2015), so it is helpful but
not necessary for the reader to have taken a computer systems course (Bryant
and O’Hallaron 2010). We introduce the parts of x86-64 assembly language that
are needed in the compiler. We follow the System V calling conventions (Bryant
and O’Hallaron 2005; Matz et al. 2013), so the assembly code that we gener-
ate works with the runtime system (written in C) when it is compiled using the
GNU C compiler (gcc) on Linux and MacOS operating systems on Intel hardware.
On the Windows operating system, gcc uses the Microsoft x64 calling conven-
tion (Microsoft 2018, 2020). So the assembly code that we generate does not work
with the runtime system on Windows. One workaround is to use a virtual machine
with Linux as the guest operating system.
xiv Preface

Acknowledgments

The tradition of compiler construction at Indiana University goes back to research


and courses on programming languages by Daniel Friedman in the 1970s and 1980s.
One of his students, Kent Dybvig, implemented Chez Scheme (Dybvig 2006), an
efficient, production-quality compiler for Scheme. Throughout the 1990s and 2000s,
Dybvig taught the compiler course and continued the development of Chez Scheme.
The compiler course evolved to incorporate novel pedagogical ideas while also
including elements of real-world compilers. One of Friedman’s ideas was to split
the compiler into many small passes. Another idea, called “the game,” was to test
the code generated by each pass using interpreters.
Dybvig, with help from his students Dipanwita Sarkar and Andrew Keep, devel-
oped infrastructure to support this approach and evolved the course to use even
smaller nanopasses (Sarkar, Waddell, and Dybvig 2004; Keep 2012). Many of the
compiler design decisions in this book are inspired by the assignment descriptions
of Dybvig and Keep (2010). In the mid 2000s, a student of Dybvig named Abdu-
laziz Ghuloum observed that the front-to-back organization of the course made it
difficult for students to understand the rationale for the compiler design. Ghuloum
proposed the incremental approach (Ghuloum 2006) on which this book is based.
I thank the many students who served as teaching assistants for the compiler
course at IU including Carl Factora, Ryan Scott, Cameron Swords, and Chris
Wailes. I thank Andre Kuhlenschmidt for work on the garbage collector and x86
interpreter, Michael Vollmer for work on efficient tail calls, and Michael Vitousek
for help with the first offering of the incremental compiler course at IU.
I thank professors Bor-Yuh Chang, John Clements, Jay McCarthy, Joseph Near,
Ryan Newton, Nate Nystrom, Peter Thiemann, Andrew Tolmach, and Michael
Wollowski for teaching courses based on drafts of this book and for their feedback.
I thank the National Science Foundation for the grants that helped to support this
work: Grant Numbers 1518844, 1763922, and 1814460.
I thank Ronald Garcia for helping me survive Dybvig’s compiler course in the
early 2000s and especially for finding the bug that sent our garbage collector on a
wild goose chase!

Jeremy G. Siek
Bloomington, Indiana
Preliminaries
1

In this chapter we introduce the basic tools needed to implement a compiler. Pro-
grams are typically input by a programmer as text, that is, a sequence of characters.
The program-as-text representation is called concrete syntax. We use concrete syn-
tax to concisely write down and talk about programs. Inside the compiler, we use
abstract syntax trees (ASTs) to represent programs in a way that efficiently sup-
ports the operations that the compiler needs to perform. The process of translating
concrete syntax to abstract syntax is called parsing and is studied in chapter 3. For
now we use the parse function in Python’s ast module to translate from concrete
to abstract syntax.
ASTs can be represented inside the compiler in many different ways, depending
on the programming language used to write the compiler. We use Python classes
and objects to represent ASTs, especially the classes defined in the standard ast
module for the Python source language. We use grammars to define the abstract
syntax of programming languages (section 1.2) and pattern matching to inspect
individual nodes in an AST (section 1.3). We use recursive functions to construct
and deconstruct ASTs (section 1.4). This chapter provides a brief introduction to
these components.

1.1 Abstract Syntax Trees

Compilers use abstract syntax trees to represent programs because they often need
to ask questions such as, for a given part of a program, what kind of language feature
is it? What are its subparts? Consider the program on the left and the diagram
of its AST on the right (1.1). This program is an addition operation that has two
subparts, a input operation and a negation. The negation has another subpart, the
integer constant 8. By using a tree to represent the program, we can easily follow
the links to go from one part of a program to its subparts.

input_int() + -8 input_int() -

8 (1.1)
Other documents randomly have
different content
of works for storage of water clearly testify. The period 69-180 ad
seems to have been marked by a considerable extension of
cultivation in these parts, and particularly in southern Numidia,
which at that time was included in the Province Africa. In this
district, between Sitifis (Setif) and Trajan’s great city Thamugadi
(Timgad), lay the commune of Lamasba[1169], the members of
which appear to have been mainly engaged in agriculture. There has
been preserved a large portion of a great inscription dealing with the
water-rights of their several farms. There is nothing to suggest that
the holders of these plots were tenants under great landlords. They
seem to be owners, not in the full sense of Roman civil law, but on
the regular provincial[1170] footing, subject to tribute. To determine
the shares of the several plots in the common water-supply was
probably the most urgent problem of local politics in this community.
The date of the inscription has been placed in the reign of
Elagabalus; but it is obviously based on earlier conditions and not
improbably a revision of an earlier scheme. It deals with the several
plots one by one, fixing the number of hours[1171] during which the
water is to be turned on to each, and making allowance for variation
of the supply according to the season of the year. A remarkable
feature of this elaborate scheme is the division of the plots into
those below the water level into which the water finds its way by
natural flow (declives), and those above water level (acclives). To
the latter it is clear that the water must have been raised by
mechanical means, and the scale of hours fixed evidently makes
allowance for the slower delivery accomplished thereby. For the
‘descendent’ water was to be left flowing for fewer hours than the
‘ascendent.’ As a specimen of the care taken in such a community to
prevent water-grabbing by unscrupulous members this record is a
document of high interest. That many others of similar purport
existed, and have only been lost to us by the chances of time, is
perhaps no rash guess.
The water-leet is called aqua Claudiana. The regulations are
issued by the local senate and people (decreto ordinis et
colonorum), for the place had a local[1172] government. Names of 43
possessors remain on the surviving portion of the stone. In form
they are generally Roman[1173]. It is noted that only three of them
have a praenomen. Of the quality of the men it is not easy to infer
anything. Some may perhaps have been Italians. Whether they, or
some of them, were working farmers must remain doubtful. At all
events they do not seem to belong to the class of coloni of whom we
shall have to speak below, but to be strictly cultivating possessors.
What labour they employed it is hardly possible to guess.

XXXVIII. FRONTINUS.
Sextus Julius Frontinus, a good specimen of the competent
departmental officers in the imperial service, was not only a
distinguished military commander but an engineer and a writer of
some merit. His little treatise[1174] on the aqueducts of Rome has for
us points of interest. From it we can form some notion of the
importance of the great water-works, not only to the city but to the
country for some miles in certain directions. For water-stealing by
the illicit tapping of the main channels was practised outside as well
as within the walls. Landowners[1175] did it to irrigate their gardens,
and the underlings of the staff (aquarii) connived at the fraud: to
prevent this abuse was one of the troubles of the curator. But in
certain places water was delivered by branch supplies from certain
aqueducts. This of course had to be duly licensed, and license was
only granted when the flow of water in the particular aqueduct was
normally sufficient to allow the local privilege without reducing the
regular discharge in Rome. The municipality of Tibur[1176] seems to
have had an old right to a branch of the Anio vetus. The aqua Crabra
had been a spring serving Tusculum[1177], but in recent times the
Roman aquarii had led off some of its water into the Tepula, and
made illicit profit out of the supply thus increased in volume.
Frontinus himself with the emperor’s approval redressed the
grievance, and the full supply of the Crabra again served the
Tusculan landlords. The jealous attention given to the water-works is
illustrated by the decrees[1178] of the Senate in the time of the
Republic and of emperors since, by which grants of water-rights can
only be made to individuals named in the grant, and do not pass to
heirs or assigns: the water must only be drawn from the reservoir
named, and used on the estate for which the license is specifically
granted.
The office of curator aquarum was manifestly no sinecure. It was
not merely that constant precautions had to be taken against the
stealing of the water. An immense staff[1179] had to be kept to their
duties, and the cleansing and repair of the channels needed prompt
and continuous attention. And it seems that some of the landowners
through whose estates the aqueducts passed gave much
trouble[1180] to the administration. Either they erected buildings in
the strips of land reserved as legal margin on each side of a channel,
or they planted trees there, thus damaging the fabric; or they drove
local roads over it; or again they blocked the access to working
parties engaged in the duties of upkeep. Frontinus quotes decrees of
the Senate dealing with these abuses and providing penalties for
persons guilty of such selfish and reckless conduct. But to legislate
was one thing, to enforce the law was another. Yet the
unaccommodating[1181] landlords had no excuse for their behaviour.
It was not a question of ‘nationalizing’ the side strips, though that
would have been amply justified in the interests of the state. But the
fact is that the old practice of Republican days was extremely tender
of private rights. If a landlord made objection to selling a part of his
estate, they took over the whole block and paid him for it. Then they
marked off the portions required for the service, and resold the
remainder. Thus the state was left unchallenged owner of the part
retained for public use. But the absence of any legal or moral claim
has not availed to stop encroachments: the draining away of the
water still goes on, with or without leave, and even the channels and
pipes themselves are pierced. No wonder that more severe and
detailed legislation was found necessary in the time of Augustus.
The writer ends by recognizing the unfairness of suddenly enforcing
a law the long disuse of which has led many to presume upon
continued impunity for breaking it. He therefore has been reviving it
gradually, and hopes that offenders will not force him to execute it
with rigour.
What stands out clearly in this picture of the water-service is the
utter lack of public spirit imputed to the landowners near Rome by a
careful and responsible public servant of good repute. There is none
of the sermonizing of Seneca or the sneers and lamentations of
Pliny. Frontinus takes things as they are, finds them bad, and means
to do his best to improve them, while avoiding the temptations of
the new broom. That a great quantity of water was being, and had
long been, diverted from the public aqueducts to serve suburban
villas and gardens, is certain. What we do not learn is whether much
or any of this was used for the market-gardens of the humble folk
who grew[1182] garden-stuff for the Roman market. It is the old
story,—little or nothing about the poor, save when in the form of a
city rabble they achieve distinction as a public burden and nuisance.
It does however seem fairly certain that licenses to abstract water
were only granted as a matter of special favour. Therefore, so far as
licensed abstraction went, it is most probable that influential owners
of suburbana were the only beneficiaries. Theft of water with
connivance[1183] of the staff was only possible for those who could
afford to bribe. There remains the alternative of taking it by eluding
or defying the vigilance of the staff. Is it probable that the poor
market-gardener ventured to do this? Not often, I fancy: we can
only guess, and I doubt whether much of the intercepted water
came his way. There was it is true one aqueduct[1184] the water of
which was of poor quality. It was a work of Augustus, intended to
supply the great pond (naumachia) in which sham sea-fights were
held to amuse the public. When not so employed, this water was
made available for irrigation of gardens. This was on the western or
Vatican side of the Tiber. Many rich men had pleasure-gardens in
that part, and we cannot be sure that even this water was in
practice serving any economic purpose.
XXXIX. INSCRIPTIONS RELATIVE TO
ALIMENTA.

It is impossible to leave unnoticed the inscriptions[1185] of this


period relative to alimenta, and Mommsen’s interpretation[1186] of
the two chief ones, though their connexion with my present subject
is not very close. In the bronze tablets recording respectively the
declarations of estate-values in the communes of Ligures Baebiani
(101 ad) and Veleia (103 ad), made with the view of ascertaining the
securities upon which the capital endowment was to be advanced,
we have interesting details of this ingenious scheme for perpetuating
charity. But neither these, nor some minor inscribed records of
bequests, nor again the experience of Pliny the younger in a
benefaction[1187] of the same kind, give us direct evidence on
labour-questions. It is in connexion with tenure of land and
management of estates that these documents mainly concern us.
The fact that there was felt to be a call for charities to encourage the
rearing of children was assuredly not a sign of social or economic
wellbeing; but this I have remarked above.
The following points stand out clearly in the interpretation of
Mommsen. The growth of large estates as against small is shewn in
both the tablets as having gone far by the time of Trajan: but not so
far as modern writers have imagined. In the case of the Ligures
Baebiani there is record of a considerable number of properties of
moderate value, indeed they are in a majority. At Veleia, though
small estates have not disappeared, there are more large ones, and
the process of absorption has evidently been more active. This was
not strange, for the former case belongs to the Hirpinian hill country
of southern Italy, the latter to the slopes of the Apennine near
Placentia, including some of the rich plain of the Po. The latter would
naturally attract capital more than the former. I have more than once
remarked that in the upland districts agricultural conditions were far
less revolutionized than in the lowlands. This seems to be an
instance in point: but the evidence is not complete. There is nothing
to shew that the estates named in these tablets were the sole
landed properties of their several owners. Nor is it probable. To own
estates in different parts of the country was a well understood policy
of landlords. How we are to draw conclusions as to the prevalence of
great estates from a few isolated local instances, without a
statement of the entire landed properties of the persons named, I
cannot see. That writers of the Empire, when they speak of
latifundia, are seldom thinking of the crude and brutal plantation-
system of an earlier time, is very true. Those vast arable farms with
their huge slave-gangs were now out of fashion, and Mommsen
points out that our records are practically silent as to large-scale
arable farming. We are not to suppose that it was extinct, but it was
probably rare.
The most valuable part of this paper is its recognition of the vital
change in Italian agriculture, the transfer of farming from a basis of
ownership to one of tenancy. The yeoman or owner-cultivator of
olden time had been driven out or made a rare figure in the most
eligible parts of Italy. The great plantations, which had largely
superseded the small-scale farms, had in their turn proved economic
failures. Both these systems, in most respects strongly contrasted,
had one point in common: the land was cultivated by or for the
owner, and for his own account. But the failure of the large-scale
plantation-system did not so react as to bring back small ownership.
Large ownership still remained, supported as it was by the social
importance attached to landowning, and occasionally by
governmental action directed to encourage investment in Italian
land. Large owners long struggled to keep their estates in hand
under stewards farming for their masters’ account. But this plan was
doomed to failure, because the care and attention necessary to
make it pay were in most cases greater than landlords were willing
to bestow. By Columella’s time this fact was already becoming
evident. He could only advise the landlords to be other than he
found them, and meanwhile point to an alternative, namely
application of the tenancy-system. It was this latter plan that more
and more found favour. The landlord could live in town and draw his
rents, himself free to pursue his own occupations. The tenant-farmer
was only bound by the terms of his lease; and, being resident, was
able to exact the full labour of his staff and prevent waste and
robbery. The custom was for the landlord to provide[1188] the
equipment (instrumentum) of the farm, or at least most of it,
including slaves. Thus he was in a sense partner of his tenant,
finding most of the working capital. Whether he had a claim to a
money rent only, or to a share of crops also, depended on the terms
of letting. It seems that rents were often in arrear, and that attempts
to recover sums due by selling up tenants’ goods did not always
cover the debts.
The typical tenant-farmer was certainly a ‘small man.’ To let the
whole of a large estate to a ‘big man’ with plenty of capital was not
the practice in Italy. Why? I think the main reason was that a big
capitalist who wanted to get the highest return on his money could
at this time do better for himself in other ventures: if set upon a
land-enterprise, he could find far more attractive openings in some
of the Provinces. Anyhow, as Mommsen says, ‘Grosspacht’ never
became acclimatized in Italy, though we find it on Imperial domains,
for instance in Africa. In connexion with this matter I am led to
remark that small tenancy ‘Kleinpacht’ seems to have existed in two
forms, perhaps indistinguishable in law, but different in their
practical effect. When a landlord, letting parcels of a big estate to
tenants, kept in hand the chief villa and its appurtenances as a sort
of Manor Farm, and tenants fell into arrear with their rent, he had a
ready means of indemnifying himself without ‘selling up’ his old
tenants and having possibly much difficulty in finding better new
ones. He could commute arrears of rent into obligations of
service[1189] on the Manor Farm. Most tenants would probably be
only too glad to get rid of the immediate burden of debt. It would
seem a better course than to borrow for that purpose money on
which interest would have to be paid, even supposing that anyone
would be willing to lend to a poor tenant confessedly in difficulties.
And such an arrangement would furnish the landlord with a fixed
amount of labour (and labour was becoming scarcer) on very
favourable terms—he or his agent would see to that. But it was not
really necessary to reserve a ‘Manor Farm’ at all, and a man owning
land in several districts would hardly do so in every estate, if in any.
Such a landlord could not readily solve the arrears-problem by
commutation. He was almost compelled[1190] to ‘sell up’ a hopeless
defaulter: and, since most of the stock had probably been supplied
by himself, there would not be much for him to sell. That such cases
did occur, we know for certain; the old tenant went, being free to
move, and to find a good new one was no easy matter, particularly
as the land was sure to have been left in a bad state. Arrears of
farm-rents had a regular phrase (reliqua colonorum) assigned to
them, and there is good reason to believe that they were a common
source of trouble. It has been well said[1191] that landlords in Italy
were often as badly off as their tenants. The truth is that the whole
agricultural interest was going downhill.
If the tenant-farmer was, as we see, becoming more and more
the central figure of Italian agriculture, we must next inquire how he
stood in relation to labour. It is a priori probable that a man will be
more ready to work with his own hands on a farm of his own than
on one hired: no man is more alive to the difference of meum and
alienum than the tiller of the soil. It is therefore not wonderful that
we find tenant-farmers employing slave labour. From the custom of
having slaves as well as other stock supplied by the landlord we may
fairly infer that tenants were, at least generally, not to be had on
other terms. Mommsen remarks[1192] that actual handwork on the
land was more and more directed rather than performed by the
small tenants. Thus it came to be more and more done by unfree
persons. This recognizes, no doubt rightly, that the system of great
estates let in portions to tenants was not favourable to a revival of
free rustic labour, but told effectively against it. He also points
out[1193] that under Roman Law it was possible for a landlord and
his slave to stand in the mutual relation of lessor and lessee. Such a
slave lessee is distinct from the free tenant colonus. It appears that
there were two forms of this relation. The slave might be farming on
his own[1194] account, paying a rent and taking the farm-profits as
his peculium. In this case he is in the eye of the law quasi colonus.
Or he might be farming on his master’s account; then he is vilicus.
In both cases he is assumed to have under him slave-labourers
supplied[1195] by the landlord, and it seems that the name vilicus
was sometimes loosely applied even in the former case. In the latter
case he cannot have been very different from the steward of a large
estate worked for owner’s account. I can only conclude that he was
put in charge of a smaller farm-unit and left more to his own
devices. Probably this arrangement would be resorted to only when
an ordinary free tenant was not to be had; and satisfactory ones
were evidently not common in the time of the younger Pliny.
So far as I can see, in this period landlords were gradually ceasing
to keep a direct control over the management of their own estates,
but the changes in progress did not tend to a rehabilitation of free
labour. One detail needs a brief special consideration. The landlord’s
agent (actor) is often mentioned, and it is clear that the actor was
generally a slave. But there is reference to the possible case[1196] of
an actor living (like his master) in town, not on the farms, and
having a wife[1197] and daughter. This suggests a freedman, not a
slave, and such cases may have been fairly numerous. Another point
for notice is the question of vincti, alligati, compediti, in this period.
Mommsen[1198] treats the chaining of field-slaves as being quite
exceptional, in fact a punishment, in Italy under the Empire. Surely it
was always in some sense a punishment. From what Columella[1199]
says of the normal employment of chained labourers in vineyard-
work I can not admit that the evidence justifies Mommsen’s
assertion. That there was a growing reluctance to use such
barbarous methods, and that local usage varied in various parts of
the country, is certain.

XL. DION CHRYSOSTOM.


We have seen that there is no lack of evidence as to the
lamentable condition of Italian agriculture in a large part of the
country. But things were no better in certain Provinces, more
particularly in Greece. Plutarch deplores[1200] the decay and
depopulation of his native land, but the most vivid and significant
picture preserved to us is one conveyed in a public address[1201] by
the famous lecturer Dion of Prusa, better known as Dion[1202]
Chrysostom. It describes conditions in the once prosperous island of
Euboea. The speaker professes to have been cast ashore there in a
storm, and to have been entertained with extraordinary kindness by
some honest rustics who were living an industrious and harmless life
in the upland parts, the rocky shore of which was notorious as a
scene of shipwrecks. There were two connected households,
squatters in the lonely waste, producing by their own exertions
everything they needed, and of course patterns of every amiable
virtue. The lecturer recounts the story of these interesting people as
told him by his host. How much of it is due to his own imagination,
or put together out of various stories, we cannot judge: but it is
manifest that what concerns us is to feel satisfied that the
experiences described were possible, and not grotesquely
improbable, in their setting of place and time. I venture to accept
the story as a sketch of what might very well have happened,
whether it actually did so or not.
We live mostly by the chase, said the hunter, with very little
tillage. This croft (χωρίον) does not belong to us either by
inheritance or purchase. Our fathers, though freemen, were poor like
ourselves, just hired herdsmen, in charge of the herds of a rich man
who owned wide farm-lands and all these mountains. When he died,
his estate was confiscated: It is said that the emperor[1203] made
away with him to get his property. Well, they drove off his live-stock
for slaughter, and our few oxen with them, and never paid our
wages. So we did the best we could, taking advantage of the
resources of the neighbourhood in summer and winter. Since
childhood I have only once visited the city[1204]. A man turned up
one day demanding money. We had none, and I told him so on my
oath. He bade me come with him to the city. There I was arraigned
before the mob as a squatter on the public land, without a grant
from the people, and without any payment. It was hinted that we
were wreckers, and had put together a fine property through that
wicked trade. We were said to have valuable farms and abundance
of flocks and herds, beasts of burden, slaves. But a wiser speaker
took a different line. He urged that those who turned the public land
to good account were public benefactors and deserved
encouragement. He pointed out that two thirds of their territory was
lying waste through neglect and lack of population. He was himself a
large landowner: whoever was willing to cultivate his land was
welcome to do so free of charge,—indeed he would reward him for
his pains—the improvement would be worth it. He proposed a plan
for inducing citizens to reclaim the derelict lands, rent-free for ten
years, and after that rented at a moderate share of the crops. To
aliens less favourable terms might be offered, but with a prospect of
citizenship in case of reclamation on a large scale. By such a policy
the evils of idleness and poverty would be got rid of. These
considerations he enforced by pointing to the pitiful state of the city
itself. Outside the gates you find, not a suburb but a hideous desert.
Within the walls we grow crops and graze beasts on the sites of the
gymnasium and the market-place. Statues of gods and heroes are
smothered in the growing corn. Yet we are forsooth to expel these
hard-working folks and to leave men nothing to do but to rob or
steal.
The rustic, being called upon to state his own case, described the
poverty of the squatter families, the innocence of their lives, their
services to shipwrecked seafarers, and so forth. On the last topic he
received a dramatic confirmation from a man in the crowd, who had
himself been one of a party of castaways hospitably relieved three
years before by these very people. So all ended well. The stress laid
on the simple rusticity of the rustic, and the mutual distrust and
mean jealousy of the townsfolk, shew in numerous touches that we
have in this narrative a highly coloured scene. But the picture of the
decayed city, with its ancient walls a world too wide for its shrunk
population, is companion to that of the deserted countryside. Both
panels of this mournful diptych could have been paralleled in the
case of many a city and territory in Italy and Greece. The moral
reflexions, in which the lecturer proceeds to apply the lessons of the
narrative, are significant. He enlarges on the superiority of the poor
to the rich in many virtues, unselfishness in particular. Poverty in
itself is not naturally an evil. If men will only work with their own
hands, they may supply their own needs, and live a life worthy of
freemen. The word αὐτουργεῖν occurs more than once in this
spirited appeal, shewing clearly that Dion had detected the plague-
spot in the civilization of his day. But he honestly admits the grave
difficulties that beset artisans in the various trades practised in
towns. They lack necessary[1205] capital: everything has to be paid
for, food clothing lodging fuel and what not, for they get nothing free
but water, and own nothing but their bodies. Yet we cannot advise
them to engage in foul degrading vocations. We desire them to live
honourably, not to sink below the standards of the greedy usurer or
the owners of lodging-houses or ships or gangs of slaves. What then
are we to do with the decent poor? Shall we have to propose turning
them out of the cities and settling them on allotments in the
country? Tradition tells us rural settlement prevailed throughout
Attica of yore: and the system worked well, producing citizens of a
better and more discreet type than the town-bred mechanics who
thronged the Assemblies and law-courts of Athens.
It may be said that Dion is a mere itinerant philosopher, who
travels about seeing the world and proposing impracticable remedies
for contemporary evils in popular sermons to idle audiences. But he
knew his trade, and his trade was to make his hearers ‘feel better’
for attending his discourses. When he portrays the follies or vices of
the age, he is dealing with matters of common knowledge, and not
likely to misrepresent facts seriously. When he suggests remedies, it
matters little that there is no possibility of applying them. Present
company are always excepted, and the townsfolk who listened to the
preacher would neither resent his strictures on city life nor have the
slightest intention of setting their own hands to the spade or plough.
That there was a kind of moral reaction[1206] in this period, and that
lecturers and essayists contributed something to the revival of
healthier public sentiment, I do not dispute; though I think too much
success is sometimes[1207] ascribed to their good intentions. At any
rate they cannot be credited with improving the conditions of rustic
life. To the farmer the voice of the great world outside was
represented by the collectors of rents and taxes, the exactors of
services, not by the sympathetic homilies of popular teachers.

XLI. NEW TESTAMENT WRITERS.


The authors of the books of the New Testament, whom it is
convenient to view together as a group of witnesses bearing on the
condition of a part of the Roman East under the early Empire, supply
some interesting matter. We read of an agriculture that includes
corn-growing, the culture of vines, and pastoral industry: the olive,
and above all the fig-tree, appear as normal objects of the
countryside. Plough spade and sickle, storehouse threshing-floor and
winepress, are the familiar appliances of rustic life, as they had been
from time immemorial. Farmers need not only hard work, but
watchfulness and forethought, for the business of their lives. Live
stock have to be protected from beasts of prey, and need endless
care. And the rustic’s outlook is ever clouded by the fear of drought
and murrain. All this is an ordinary picture, common to many lands:
only the anxiety about water-supply is perhaps specially Oriental.
The ox and the ass are the chief beasts of draught and burden. In
short, country life goes on as of old, and much as it still does after
many changes of rulers.
From the way in which farmers are generally spoken of I infer that
they are normally peasant[1208] landowners. That is to say, not
tenants of an individual landlord, but holding their farms with power
of sale and right of succession, liable to tribute. The Roman state is
strictly speaking the owner, having succeeded to the royal ownership
assumed by the Seleucid kings. But that there was also letting[1209]
of estates to tenant-farmers is clear, for we read of collection of
rents. At the same time we find it suggested, apparently as a moral
rather than legal obligation, that the toiling farmer has the first
claim[1210] on the produce, and the ox is not to be muzzled. Such
passages, and others insisting on honesty and the duty of labour,
keep us firmly reminded of the moral aims pervading the works of
these writers. In other words, they are more concerned to define
what ought to be than to record what is. Many of the significant
references to rustic matters occur in parables. But we must not
forget that a parable would have little force if its details were not
realistic.
Of the figures appearing on the agricultural scene we may
distinguish the wealthy landlord[1211], whether farming for his own
account or letting his land to tenants: the steward[1212] farming for
his lord’s account: the tenant-farmer: probably the free peasant on a
small holding of his own. Labour is represented by the farmer
working with his own hands, and by persons employed simply as
labourers. These last are either freemen or slaves. Slavery is
assumed as a normal condition, but a reader can hardly help being
struck by the notable passages in which the wage-earner appears as
a means of illustrating an important point. Does the occurrence of
such passages suggest that in these Oriental surroundings wage-
service was as common a system as bond-service, perhaps even
more so? I hesitate to draw this conclusion, for the following reason.
Accepting the fact of slavery (as the writers do), there was not much
to be said beyond enjoining humanity on masters and conscientious
and respectful service on slaves. But the relation between hirer and
hired, presumably a bargain, opened up far-reaching issues of
equity, transcending questions of formal law. Hence we hear much
about it. That the workman is worthy of his meat (ἐργάτης ...
τροφῆς) is a proposition of which we have an earlier[1213] version,
referring to slaves. The cowardice of the hireling shepherd points a
notable moral. The rich who defraud the reaper of his hire[1214]
meet with scathing denunciation. For to him that worketh the reward
is not reckoned[1215] of grace but of debt.
This last proposition seems to furnish a key to the remarkable
parable[1216] of the Labourers in the Vineyard, which has been
subjected to many diverse interpretations. If we accept the view that
the wages represent the Kingdom of God, and that this reward is
granted not of debt but of grace, it is clear that great stress is laid
on the autocratic position of the householder (οἰκοδεσπότης). His
treatment of the hired labourers is an assertion of entire indifference
to what we call ‘economic’ considerations. How it is to be interpreted
as equitable, theologians must decide, or be content to leave
modern handworkers to draw their own conclusions. My interest in
the matter may be shewn in the question whether this householder
is to be regarded as a typical figure, or not. I trust I am guilty of no
irreverence in saying that to me he seems a purely hypothetical
character. That is to say that I take the gist of the parable to be this:
if an employer chose to deal with his hirelings on such arbitrary
principles, he would be acting within his rights. I do not infer that
such conduct was likely in ordinary life, or even that a concrete case
of its occurrence had ever been known. I cannot believe that in a
country where debts[1217] and usury are referred to as matters of
course, and where masters entrusted money[1218] to their slaves for
purposes of trade, where sales of land[1219] were an ordinary
business transaction, a sane individualistic capitalist would act as the
man in this parable. Those who think differently must clear up their
own difficulties. I would add that this parable, the details of which
seem to me non-realistic, only occurs in one of the Gospels. Is it
possible that it is based on some current Oriental story?

XLII. MARTIAL AND JUVENAL.


Among the witnesses, other than technical writers, from whom we
get evidence as to the conditions of agriculture under the Empire,
are two poets, Martial and Juvenal. The latter, a native of Aquinum
in the old Volscian part of Latium, never shook off the influence of
his connexion with rural Italy. The former, a native of Bilbilis in
Spain, was one of the gifted provincials who came to Rome as the
literary centre of the world. He spent more than thirty years there,
and made an unrivalled name as a writer of epigrams, but his heart
was in Spain. The attitude of these two men towards the facts of
their time is very different, and the difference affects the value of
their evidence. In the satires of Juvenal indignant rhetoric takes up a
high moral position, and declaims fiercely against abominations. Now
this attitude is beset with temptations to overstate an evil rather
than weaken effect. Moreover, in imperial Rome it was necessary to
be very careful: not only were personal references dangerous, but it
was above all things necessary to avoid provoking the Emperor. Yet
even Emperors could (and did) view attacks upon their predecessors
with indifference or approval: while vicious contemporaries were not
likely to put on the cap if their deceased counterparts were assailed.
So the satirist, confining his strictures mainly to the past, is not often
a contemporary witness of the first order. It is fortunate that his
references to rustic conditions are not much affected by this
limitation: but they mostly refer to the past. Martial on the contrary
is a mere man of his time. His business is not to censure, still less to
reform, but to find themes for light verse such as will hit the taste of
average Roman readers. He soon discovered that scandal was the
one staple topic of interest, and exploited it as a source of ‘copy’
down to the foulest dregs. Most of the characters exposed appear
under fictitious Greek names, but doubtless Roman gossips applied
the filthy imputations to each other. We need not suppose that
Martial’s ruling passion was for bawdy epigram. But he knew what
would hit the taste of an idle and libidinous world. For himself,
nothing is clearer than that he found life in the great city a sore trial,
not solely from the oppressive climate at certain seasons of the year.
He was too clever a man not to suffer weariness in such
surroundings. He had to practice the servility habitually displayed by
poor men towards the rich and influential, but he did not like it. It
seems to have been through patronage that he got together
sufficient wealth to enable him eventually to retire to his native
country. The din and dirt and chronic unrest of Rome were to him,
as to Juvenal, an abomination: and from these ever-present evils
there was, for dwellers in mean houses or crowded blocks of sordid
flats, no escape. Both writers agree that the Rome of those days was
only fit for the wealthy to live in. Secure in his grand mansion on one
of the healthiest sites, with plenty of elbow-room, guarded against
unwelcome intrusions by a host of slaves and escorted by them in
public, the millionaire could take his life easily: he could even sleep.
Martial had his way to make as a man of letters, and needed to keep
brain and nerves in working order. For this, occasional retirement
from the urban pandemonium was necessary. So he managed to
acquire a little suburban[1220] property, where he could spend days
in peace and quiet. Many of his friends did the same. To keep such a
place, however small, in good order, and to grow some country
produce, however little, it was necessary to have a resident[1221]
vilicus. He had also a vilica, and there would probably be a slave or
two under them. The poet was now better off, and doing as others
did. These suburbana, retreats for the weary, were evidently
numerous. Their agricultural significance was small. Martial often
pokes fun at the owners who withdraw to the country for a holiday,
taking with them[1222] their supplies of eatables bought in the
markets of Rome. Clearly the city markets were well supplied: and
this indicates the existence of another class of suburban properties,
market-gardens on a business footing, of which we hear little
directly. An industry of this kind springs up round every great centre
of population: how far it can extend depends on the available means
of delivering the produce in fair marketable condition. Round Rome it
had no doubt existed for centuries, and was probably one of the
most economically sound agricultural undertakings in central Italy.
That it was conducted on a small scale and was prosperous may be
the reason why it attracted little notice in literature.
Though Martial cannot be regarded as an authority on Italian
agriculture, it so happens that passages of his works are important
and instructive, particularly in connexion with matters of land-
management and farm-labour. He gives point to his epigrams by
short and vivid touches, above all by telling contrasts. Now this style
of writing loses most of its force if the details lack reality. He was
therefore little tempted to go beyond the truth in matters of ordinary
non-bestial life, such as agricultural conditions; we may accept him
as a good witness. To begin with an all-important topic, let us see
what we get from him on the management of land, either for the
landlord’s account under a slave vilicus, or by letting it to a free
colonus. In explaining the gloomy bearing of Selius, he
remarks[1223] that it is not due to recent losses: his wife and his
goods and his slaves are all safe, and he is not suffering from any
failures of a tenant or a steward. Here colonus as opposed to vilicus
must mean a free tenant, who might be behindhand with his rent or
with service due under his lease. The opposition occurs elsewhere,
as when he refers[1224] to the produce sent in to a rich man in Rome
from his country estates by his steward or tenant. So too on the
birthday of an eminent advocate all his clients and dependants send
gifts; among them[1225] the hunter sends a hare, the fisherman
some fish, and the colonus a kid. The venator and piscator are very
likely his slaves. In protesting[1226] against the plague of kissing as it
strikes a man on return to Rome, he says, ‘all the neighbours kiss
you, and the colonus too with his hairy unsavoury mouth.’ It seems
to imply that the rustic tenant would come to Town to pay his
respects to his landlord. Barring the kiss, the duty of welcoming the
squire makes one think of times not long gone by in England. In one
passage[1227] there is a touch suggestive of almost medieval
relations. How Linus has managed to get through a large inherited
fortune, is a mystery in need of an explanation. He has not been a
victim of the temptations of the great wicked city. No, he has always
lived in a country town, where economy was not only possible but
easy. Everything he needed was to be had cheap or gratis, and there
was nothing to lead him into extravagant ways. Now among the
instances of cheapness is the means of satisfying his sexual passions
when they become unruly. At such moments either the vilica or the
duri nupta coloni served his turn. The steward’s consort would be his
slave, and there is no more to be said: but the tenant-farmer’s wife,
presumably a free woman, is on a different footing. There is no
suggestion of hoodwinking the husband, for the situation is treated
as a matter of course. It would rather seem that the landlord is
represented as relying on the complaisance of a dependent boor. If I
interpret the passage rightly, we have in it a vivid sidelight on the
position of some at least of the coloni of the first century ad. That
vilici and coloni alike were usually clumsy rustics of small manual
skill, is suggested by two passages[1228] in which they are credited
with bungling workmanship in wood or stone. Perhaps we may
detect reference to a colonus in an epigram on a man who spends
his money lavishly on his own debaucheries but is meanly niggardly
to necessitous friends. It says ‘you sell ancestral lands to pay for a
passing gratification of your lust, while your friend, left in the lurch,
is tilling land[1229] that is not his own.’ That is, you might have made
him a present of a little farm, as many another has done; but you
have left him to sink into a mere colonus. Enough has now been said
to shew that these tenant-farmers were a humble and dependent
class of men, and that the picture drawn from passages of Martial
corresponds to that drawn above in Weber’s interpretation of
Columella.
It is not necessary to set out with the same fulness all the
evidence of Martial on agricultural matters regarded from various
points of view. The frequent reference to the land is a striking fact:
like his fellow-countryman Columella, he was clearly interested in the
land-system of Italy. He shews wide knowledge of the special
products of different districts; a knowledge probably picked up at
first in the markets of Rome, and afterwards increased by
experience. No writer draws the line more distinctly between
productive and unproductive estates. That we hear very much more
of the latter is no wonder: so long as the supremacy of Rome was
unshaken, and money poured into Italy, a great part of the country
was held by wealthy owners to whom profit was a less urgent
motive than pleasure or pride. To what lengths ostentation could go
is seen[1230] in the perverse fancy of a millionaire to have a real rus
in urbe with grounds about his town house so spacious that they
included a real vineyard: here in sheltered seclusion he could have a
vintage in Rome. This is in truth the same vulgar ambition as that
(much commoner) of the man who prides himself on treating guests
at his country mansion to every luxury procurable in Rome. It is
merely inverted.
At this point it is natural to ask whence came the vast sums
lavished on these and other forms of luxury. Italy was not a great
manufacturing country. The regular dues from the Provinces flowed
into the treasuries, not openly into private pockets. Yet a good deal
of these monies no doubt did in the end become the reward of
individuals, as salaries or amounts payable to contractors, etc. These
however would not by themselves suffice to account for the
immense squandering that evidently took place. A source of
incomes, probably much more productive than we might at first
sight imagine, existed in the huge estates owned by wealthy
Romans in the lands beyond the seas. Martial refers[1231] to such
properties at Patrae in Achaia, in Egypt, etc. The returns from these
estates, however badly managed, were in the total probably very
large. And they were no new thing. In Varro and in Cicero’s letters
we find them treated as a matter of course: the case of Atticus and
his lands in Epirus is well known. Pliny[1232] tells us of the case of
Pompey, and also of the six land-monopolizers whom Nero found in
possession of 50% of the Province of Africa. The practice of usury in
the subject countries was no longer so widespread or so
remunerative as it had been in the last period of the Republic, but it
had not ceased, and the same is true of the farming of revenues.
Commerce was active: but we are rather concerned with the means
of paying for imported goods than with the fact of importation. The
anxiety as to the supply of corn from abroad shews itself in the
gossip[1233] of quidnuncs as to the fleet of freight-ships coming from
Alexandria. Puteoli and Ostia were doubtless very busy; all we need
note is that someone must have made money[1234] in the business
of transport and delivery. These considerations may serve to explain
the presence of so much ‘money in the country’ as we say, and the
resulting extravagance. But all this social and economic fabric rested
on the security guaranteed by the imperial forces on land and sea.
One of Martial’s epigrams[1235] is of special interest as describing
a manifestly exceptional estate. It was at or near Baiae, the famous
seaside pleasure-resort, which had been the scene of costly fancies
and luxurious living for more than a hundred years. The point of the
poem lies in the striking contrast of this place compared with the
unproductive suburbanum[1236] of another owner, which is kept
going by supplies from the Roman market. For the place is a genuine
unsophisticated country farm, producing corn and wine and good
store of firewood, and breeding cattle swine sheep and various kinds
of poultry and pigeons. When rustic neighbours come to pay their
respects, they bring presents, such as honey in the comb, cheese,
dormice, a kid, a capon. The daughters[1237] of honest tenants bring
baskets of eggs. The villa is a centre of hospitality; even the slaves
are well fed. The presence of a slave-household brought from Town
is particularly dwelt on: what with fishing and trapping and with
‘light work’ in the garden, these spoilt menials, even my lord’s pet
eunuch, are happy enough. There are also young home-bred slaves
(vernae) probably the offspring of the farm-slaves. The
topsyturvydom of this epigram is so striking that one may suspect
Martial of laughing in his sleeve at the eccentric friend whose farm
he is praising. In any case this cannot be taken seriously as a
realistic picture of a country seat practically agricultural. The owner
evidently drew his income from other sources. And the sort of man
who treated himself to an eunuch can hardly have been much of a
farmer, even near Baiae. The mention of probi coloni illustrates what
has been said above as to tenants, and that a farm could be
described in such words as rure vero barbaroque is a candid
admission that in too many instances a place of the kind could only
by courtesy be styled a farm, since the intrusion of ‘civilization’ (that
is, of refined and luxurious urban elements) destroyed its practical
rustic character. That the estate in question produced enough to
feed the owner and his guests, his domestics brought from Rome,
and the resident rustic staff as well, is credible. But there is nothing
to shew that it produced any surplus for the markets: it may have
done something in this direction, but that it really paid its way,
yielding a moderate return on the capital sunk in land slaves and
other farm-stock, is utterly incredible.
Whether in town or country, the life sketched by Martial is that of
a society resting on a basis of slavery. At the same time the supply
of new slaves[1238] was not so plentiful as it had been in days before
the Roman Peace under Augustus. Serviceable rustic slaves were
valuable nowadays. Addressing Faustinus, the wealthy owner of the
above Baian villa and several others, the poet says ‘you can send
this book[1239] to Marcellinus, who is now at the end of his
campaign in the North and has leisure to read: but let your
messenger be a dainty Greek page. Marcellinus will requite you by
sending you a slave, captive from the Danube country, who has the
making of a shepherd in him, to tend the flocks on your estate by
Tibur.’ Each friend is to send the other what the other lacks and he is
in a position to supply. This is a single instance; but the suggested
do ut des is significant. As wars became rarer, and prisoners fewer,
the disposal of captives would be a perquisite of more and more
value. That the normal treatment of slaves was becoming more and
more humane, is certain. But whether humanitarian sentiment in
Stoic forms, as preached by Seneca and others, had much to do with
this result, is more doubtful. The wisdom of not provoking discontent
among the slaves, particularly in the country, was well understood.
The decline of the free rustic population had made the absence of a
regular police force a danger not to be ignored. Improved conditions
were probably in most cases due to self-interest and caution much
more than to humane sentiment. In Martial’s day we may gather
from numerous indications that in general the lot of slaves was not a
hard one if we except the legal right of self-disposal. Urban
domestics were often sadly spoilt, and were apt to give themselves
great airs outside the house or to callers at the door. But I believe
that in respect of comfort and happiness the position of a steward
with a slave-staff in charge of a country place owned by a rich man
was in most cases far pleasanter. Subject to the preparation for the
master’s occasional visits and entertainment of his guests, these
men were left very much to their own devices. The site of the villa
had been chosen for its advantages. So long as enough work was
done to satisfy the owner, they, his caretakers, enjoyed gratis for the
whole year[1240] the privileges and pleasures which he paid for
dearly and seldom used.
It seems certain that it was on such estates that most of the
slave-breeding took place. It was becoming a more regular practice,
as we see from Columella. And it had advantages from several
points of view. The slave allowed to mate with a female partner and
produce children was more effectively tied to the place than the
unmated labourer on a plantation was by his chain. So long as the
little vernae were not brutally treated (and it seems to have been a
tradition to treat them well), the parents were much less likely to
join in any rebellious schemes. And, after all, the young of slaves
were worth money, if sold; while, if kept by the old master, they
would work in what was the only home they had known: they would
be easier to train and manage than some raw barbarian from
Germany or Britain or the Sudan. But it must not be forgotten that
the recognition of slave-breeding foreboded the eventual decline of
slavery—personal slavery—as an institution, at least for purposes of
rustic life. I know of no direct evidence[1241] as to the class or
classes from which the unfree coloni of the later Empire were drawn.
But it seems to me extremely probable that many of the coloni of
the period with which we are just now concerned were home-bred
slaves manumitted and kept on the estate as tenants. This
conjecture finds a reason for manumission, as the freedman would
be capable of a legal relation, which the slave was not. The
freedman’s son would be ingenuus, and would represent, in his
economic bondage under cover of legal freedom, a natural stage in
the transition from the personal slave to the predial serf.
That there were vernae on the small suburban properties, the
rest-retreats of Martial and many others, is not to be doubted. But
they can hardly have been very numerous. These little places were
often but poorly kept up. The owners were seldom wealthy men,
able to maintain many slaves. Economy and quiet were desired by
men who could not afford ostentation. The normal use of the epithet
sordidus[1242] (not peculiar to Martial) in speaking of such places,
and indeed of small farmsteads in general, is characteristic of them
and of the undress life led there. The house was sometimes in bad
condition. To patch up a leaky roof[1243] a present of a load of tiles
was welcome. A man buys a place the house (casa) on which is
horribly dark and old: the poet remarks that it is close to the
pleasure-garden (hortos) of a rich man. This explains the purchase:
the buyer will put up with bad lodging for the prospect of good
dinners at his neighbour’s table. The difficulty of finding a purchaser
for an estate of bad sanitary record, and the damage done to
riparian farms by the Tiber floods, are instances[1244] of the ordinary
troubles of the little landowners near Rome. A peculiar nuisance,
common in Italy, was the presence in some corner of a field of the
tomb[1245] of some former owner or his family. A slice of the land,
so many feet in length and breadth, was often reserved[1246] as not
to pass with the inheritance. What the heir never owned, that he
could not sell. So, when the property changed hands, the new owner
had no right to remove what to him might be nothing but a
hindrance to convenient tillage. Altars[1247] taken over from a
predecessor may also have been troublesome at times, but their
removal was probably less difficult.
The picture of agricultural conditions to be drawn from Juvenal
agrees with that drawn from Martial. But, as said above, the point of
view is different in the satirist, whose business it is to denounce
evils, and who is liable to fall into rhetorical exaggeration. And to a
native of central Italy the tradition of a healthier state of things in
earlier ages was naturally a more important part of his background
than it could be to a man from Spain. Hence we find vivid
scenes[1248] drawn from legend, shewing good old Romans, men of
distinction, working on the land themselves and rearing well-fed
families (slaves included) on the produce of meagre little plots of
two iugera. An ex-consul[1249] breaks off his labours on a hillside,
shoulders his mattock, and joins a rustic feast at the house of a
relative. The hill-folk of the Abruzzi are patterns of thrifty
contentment, ready to earn their bread[1250] with the plough. But
the civic duties are not forgotten. The citizen has a double function.
He serves the state in arms and receives a patch of land[1251] as his
reward for wounds suffered. He has to attend the Assembly before
his wounds[1252] are fully healed. In short, he is a peasant soldier
who does a public duty in both peace and war. The vital need of the
present day[1253] is that parents should rear sons of this type. Here
we have the moral which these scenes, and the frequent references
to ancient heroes, are meant to impress on contemporaries. A
striking instance[1254] from historical times is that of Marius, who is
represented as having risen from the position of a wage-earning
farm-labourer to be the saviour of Rome from the barbarians of the
North. But the men of the olden time led simple lives, free from the
extravagance and luxury of these days and therefore from the
temptations and ailments that now abound. The only wholesome
surroundings[1255] now are to be found in out-of-the way country
corners or the homes of such frugal citizens as Juvenal himself. But
these are mere islets in a sea of wantonness bred in security: luxury
is deadlier[1256] than the sword, and the conquered world is being
avenged in the ruin of its conqueror. Perhaps no symptom on which
he enlarges is more significant and sinister from his own point of
view than that betrayed in a passing reference by the verbal
contrast[1257] between paganus and miles. The peasant is no longer
soldier: and in this fact the weightiest movements of some 250 years
of Roman history are virtually implied.
So much for an appeal to the Roman past. But Juvenal, like Vergil
before him, was not content with this. He looks back to the primitive
age[1258] of man’s appearance on earth and idealizes the state of
things in this picture also. Mankind, rude healthy and chaste, had
not yet reached the notion of private property: therefore theft was
unknown. The moral is not pressed in the passage where this
description occurs; but it is worth noting because the greed of men
in imperial Rome, and particularly in the form of land-grabbing and
villa-building, is a favourite topic in the satires. All this side of
contemporary life, viewed as the fruit of artificial appetites and
unnecessary passions, is evidence of a degeneracy that has been
going on ever since the beginnings of society. And the worst of it is
that those who thrive on present conditions are the corrupt the
servile and the mean, from whom no improvement can be hoped for.
Juvenal’s picture of present facts as he sees them is quite enough to
justify his pessimism. As a means of arresting degeneration he is
only able to suggest a change[1259] of mind, in fact to urge people
to be other than they are. But he cannot shew where the initiative is
to be found. Certainly not in the mongrel free populace of Rome, a
rabble of parasites and beggars. Nor in the ranks of the wealthy
freedmen into whose hands the chief opportunities of enrichment
have passed, thanks to the imperial jealousy of genuine Romans and
preference of supple aliens. These freedmen are the typical
capitalists: they buy up everything, land included; and Romans who
despise these upstarts have nevertheless to fawn on them. Nor
again are leaders to be found in the surviving remnant of old
families. It is a sad pity, but pride of birth, while indisposing them to
useful industry, does not prevent them from debauchery or from
degrading themselves in public. Financial ruin and charges of high
treason are destroying them: even were this not so, who would look
to such persons for a wholesome example? Neither religion with its
formalities and excitements, nor philosophy with its professors
belying their moral preaching, could furnish the means of effecting
the change of heart needed for vital reform.
No, it was not from the imperial capital, the reeking hotbed of
wickedness, that any good could come. And when Juvenal turns to
the country it is remarkable how little comfort he seems to find in
the rural conditions of Italy. Like other writers, he refers to the
immense estates[1260] that extended over a great part of the
country, both arable and grazing lands (saltus), the latter in
particular being of monstrous size. We cannot get from him any hint
that the land-monopoly, the canker of the later Republic, had been
effectually checked. Nor indeed had it. One of the ways in which rich
patrons[1261] rewarded clients for services, honourable or (as he
suggests) often dishonourable, was to give the dependant a small
landed estate. The practice was not new. Maecenas had given
Horace his Sabine farm. But the man who gave away acres must
have had plenty of acres to give. True, some of the great landlords
had earned[1262] their estates by success in an honourable
profession: but the satirist is naturally more impressed by the cases
of those, generally freedmen, whose possessions are the fruit of
corrupt compliance or ignoble trades. These upstarts, like the
Trimalchio of Petronius, live to display their wealth, and the
acquisition of lands[1263] and erection of costly villas are a means to
this end. The fashion set by them is followed by others, and over-
buying and over-building are the cause of bankruptcies. Two
passages[1264] indicate the continued existence of an atrocious evil
notorious in the earlier period of the latifundia, the practice of
compelling small holders to part with their land by various outrages.
The live stock belonging to a rich neighbour are driven on to the
poor man’s farm until the damage thus caused to his crops forces
him to sell—of course at the aggressor’s price. A simpler form,
ejectment without pretence of purchase, is mentioned as an instance
of the difficulties in the way of getting legal redress, at least for
civilians. There would be little point in mentioning such wrongs as
conceivable possibilities: surely they must have occurred now and
then in real life. The truth, I take it, was that the great landlord
owning a host of slaves had always at disposal a force well able to
carry out his territorial ambitions; and possession of power was a
temptation to use it. The employment of slaves in rural border-raids
was no new thing, and the slave, having himself nothing to lose,
probably found zest in a change of occupation.
In Juvenal agriculture appears as carried on by slave labour, and
the employment of supplementary wage-earners is ignored; not
unnaturally, for it was not necessary to refer to it. The satirist
himself[1265] has rustic slaves, and is proud that they are rustic,
when they on a special occasion come in to wait at his table in
Rome. Slaves are of course included[1266] in the stock of an estate,
great or small, given or sold. All this is commonplace: what is more
to the satirist’s purpose is the mention[1267] of a member of an
illustrious old family who has come down in the world so low as to
tend another man’s flocks for hire. And this is brought in as a
contrast to the purse-proud insolence of a wealthy freedman. But
more remarkable is the absence of any reference to tenant coloni.
Even the word colonus does not occur in any shade of meaning. This
too may fairly be accounted for by the fact that little could have
been got out of references to the system for the purposes of his
argument. It was, as he knew, small peasant landowners, not
tenants, that had been the backbone of old Rome; and it was this
class, viewed with the sympathetic eye of one sighing for perished
glories, that he would have liked to restore. It is a satirist’s bent to
wish for the unattainable and protest against the inevitable. For
himself, he can sing the praises of rustic simplicity and cheapness
and denounce the luxury and extravagance of Roman society,
though he dare not assail living individuals. And in exposing the
rottenness of the civilization around him he attacks the very vices
that had grown to such portentous heights through the development
of slavery. Idleness bore its fruit, not only in the debauchery and
gambling that fostered unholy greed and crimes committed to
procure the money that was ever vanishing, but in the degradation
of honest labour. Pampered menials were arrogant, poor citizens
servile. And vast tracts of Italian land bore witness to the mournful
fact that the land system, so far from affording a sound basis for
social and economic betterment, was itself one of the worst
elements of the situation.
At this stage it is well to recall the relation between agriculture
and military service, the farmer-soldier ideal. The long-since existing
tendency for the soldier to become a professional, while the free
farmer class was decaying, had never obliterated the impression of
this ideal on Roman minds. The belief that gymnastic exercises on
Greek models were no effective substitute for regular manual labour
in the open air as guarantees of military ‘fitness’ is still strong in
Juvenal. It shews itself in his pictures of life in Rome, where such
exercises were practised for the purpose of ‘keeping fit’ and ‘getting
an appetite,’ much as they are now. Followed by baths and massage
and luxurious appliances of every kind, this treatment enabled the
jaded city-dweller to minimize the enervating effects of idleness
relieved by excitements and debauchery. He significantly lays stress
on the fact that these habits were as common among women as
among men. The usual allowance must be made for a satirist’s
exaggeration; but the general truth of the picture is not to be
doubted. The city life was no preparation for the camp with its rough
appliances and ever-present need for the readiness to endure
cheerfully the hardships of the field. The toughness of the farm-
labourer was proverbial: the Latin word durus is his conventional
epithet. In other words, he was a model of healthy hardness and
vigour. Now to Juvenal, as to others, the best object of desire[1268]
was mens sana in corpore sano, and he well knew that to secure the
second gave the best hope of securing the first. We might then
expect him to recommend field work as the surest way to get and
keep vigorous health. Yet I cannot find any indication of this precept
save the advice to a friend to get out of Rome and settle on a
garden-plot in the country. He says ‘there live devoted[1269] to your
clod-pick; be the vilicus of a well-tended garden.’ I presume he
means ‘be your own steward, and lend a hand in tillage as a steward
would do.’ But an average vilicus would be more concerned to get
work out of his underlings than to exert himself, and Juvenal is not
very explicit in his advice, the main point being to get his friend out
of Rome. I have reserved for comparison with this passage one from
Martial[1270]. In a couplet on a pair of halteres (something rather
like dumb-bells) he says ‘Why waste the strength of arms by use of
silly dumb-bells? If a man wants exercise, he had better go and dig
in a vineyard.’ This is much plainer, but one may doubt whether it is
seriously meant to be an ordinary rule of life. Probably it is no more
than a sneer at gymnastic exercises. For Martial well knew that
muscle developed by the practice of athletics[1271] is very different
from the bodily firmness and capacity for continuous effort under
varying conditions that is produced by a life of hard manual labour.
And the impression left on a reader’s mind by epigrammatist and
satirist alike is that in Rome and in the most favoured and accessible
parts of Italy the blessing of ‘corporal soundness’ was tending to
become a monopoly of slaves. For when Juvenal declares[1272] that
nowadays the rough fossor, though shackled with a heavy chain,
turns up his nose at the garden-stuff that fed a Manius Curius in the
olden days, hankering after the savoury fleshpots of the cook-shop,
we need not take him too seriously.

XLIII. PLINY THE YOUNGER.


The younger Pliny, one of the generation who remembered
Vespasian, lived through the dark later years of Domitian, and
rejoiced in the better times of Nerva and Trajan, is one of our most
important witnesses. Not being a technical writer on agriculture, it
was not his business to dwell on what ought to be done rather than
what was being done. Being himself a great landowner as well as a
man of wide interests and high reputation, he knew the problems of
contemporary land-management from experience, and speaks with
intelligence and authority. He was not a man of robust constitution,
and like many others he found much refreshment in rural
sojournings. He is remarkable for keen appreciation of beautiful
scenery. Adopted by his uncle, the author of the Natural History,
well-educated and in touch with the literary circles and the best
social life of Rome, his letters illustrate the intellectual and moral
influences that prevailed in cultivated households of honest
gentlemen. In particular he is to us perhaps the very best example
of the humanizing tendency of the current philosophies of the day in
relation to the subject of slavery. He is deeply interested in
promoting manumissions[1273] whenever he gets a chance. His

You might also like