Visit https://ebookfinal.

com to download the full version and

explore more ebook

Graph Algorithms Practical Examples in Apache Spark

and Neo4j 2020-06-05 - Third Release Edition Mark
Needham

_____ Click the link below to download _____

https://ebookfinal.com/download/graph-algorithms-
practical-examples-in-apache-spark-and-
neo4j-2020-06-05-third-release-edition-mark-needham/

Explore and download more ebook at ebookfinal.com

Here are some recommended products that might interest you.
You can download now and explore!

Graph Algorithms and Applications 7 Giuseppe Liotta

https://ebookfinal.com/download/graph-algorithms-and-
applications-7-giuseppe-liotta/

ebookfinal.com

Practical AI on the Google Cloud Platform 2020-01-30:

First Early Release Edition Micheal Lanham

https://ebookfinal.com/download/practical-ai-on-the-google-cloud-
platform-2020-01-30-first-early-release-edition-micheal-lanham/

ebookfinal.com

CIM Coursebook 05 06 Marketing Management in Practice CIM

Coursebook 2005-2006 Edition John Williams

https://ebookfinal.com/download/cim-coursebook-05-06-marketing-
management-in-practice-cim-coursebook-2005-2006-edition-john-williams/

ebookfinal.com

Handbook of Graph Theory Combinatorial Optimization and

Algorithms 1st Edition Subramanian Arumugam

https://ebookfinal.com/download/handbook-of-graph-theory-
combinatorial-optimization-and-algorithms-1st-edition-subramanian-
arumugam/
ebookfinal.com
A Java Library of Graph Algorithms and Optimization 1st
Edition Hang T. Lau

https://ebookfinal.com/download/a-java-library-of-graph-algorithms-
and-optimization-1st-edition-hang-t-lau/

ebookfinal.com

The New Europe Architectural Design 05 06 2006 Vol 76 N 3

1st Edition Valentina Croci

https://ebookfinal.com/download/the-new-europe-architectural-
design-05-06-2006-vol-76-n-3-1st-edition-valentina-croci/

ebookfinal.com

Algorithms in a Nutshell A Desktop Quick Reference 2

(Early Release) Edition George T. Heineman

https://ebookfinal.com/download/algorithms-in-a-nutshell-a-desktop-
quick-reference-2-early-release-edition-george-t-heineman/

ebookfinal.com

Practical Yield Line Design 1st edition, 24-06-03 Edition

Gerard Kennedy

https://ebookfinal.com/download/practical-yield-line-design-1st-
edition-24-06-03-edition-gerard-kennedy/

ebookfinal.com

Photography Foundations for Art and Design Third Edition A

Practical Guide to Creative Photography Mark Galer

https://ebookfinal.com/download/photography-foundations-for-art-and-
design-third-edition-a-practical-guide-to-creative-photography-mark-
galer/
ebookfinal.com
Graph Algorithms Practical Examples in Apache Spark
and Neo4j 2020-06-05 - Third Release Edition Mark
Needham Digital Instant Download
Author(s): Mark Needham; Amy E. Hodler
ISBN(s): 9781492047681, 1492047686
Edition: 2020-06-05 - Third Release
File Details: PDF, 21.57 MB
Year: 2020
Language: english
 Graph Algorithms
Practical Examples in
Apache Spark and Neo4j

Mark Needham and Amy E. Hodler

Beijing Boston Farnham Sebastopol Tokyo

Graph Algorithms
by Mark Needham and Amy E. Hodler
Copyright © 2019 Amy Hodler and Mark Needham. All rights reserved.
Printed in the United States of America.
Published by O’Reilly Media, Inc., 1005 Gravenstein Highway North, Sebastopol, CA 95472.
O’Reilly books may be purchased for educational, business, or sales promotional use. Online editions are
also available for most titles (http://oreilly.com). For more information, contact our corporate/institutional
sales department: 800-998-9938 or corporate@oreilly.com.

Acquisitions Editor: Jonathan Hassell Indexer: Judy McConville

Development Editor: Jeff Bleiel Interior Designer: David Futato
Production Editor: Deborah Baker Cover Designer: Karen Montgomery
Copyeditor: Tracy Brown Illustrator: Rebecca Demarest
Proofreader: Rachel Head

May 2019: First Edition

Revision History for the First Edition

2019-04-15: First Release
2019-05-16: Second Release
2020-06-05: Third Release

See http://oreilly.com/catalog/errata.csp?isbn=9781492047681 for release details.

The O’Reilly logo is a registered trademark of O’Reilly Media, Inc. Graph Algorithms, the cover image of a
European garden spider, and related trade dress are trademarks of O’Reilly Media, Inc.
While the publisher and the authors have used good faith efforts to ensure that the information and
instructions contained in this work are accurate, the publisher and the authors disclaim all responsibility
for errors or omissions, including without limitation responsibility for damages resulting from the use of
or reliance on this work. Use of the information and instructions contained in this work is at your own
risk. If any code samples or other technology this work contains or describes is subject to open source
licenses or the intellectual property rights of others, it is your responsibility to ensure that your use
thereof complies with such licenses and/or rights.
This work is part of a collaboration between O’Reilly and Neo4j. See our statement of editorial independ‐
ence.

978-1-492-05781-9
[LSI]
 Table of Contents

Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

Foreword. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
What Are Graphs? 2
What Are Graph Analytics and Algorithms? 3
Graph Processing, Databases, Queries, and Algorithms 6
OLTP and OLAP 7
Why Should We Care About Graph Algorithms? 8
Graph Analytics Use Cases 12
Conclusion 13

2. Graph Theory and Concepts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Terminology 15
Graph Types and Structures 16
Random, Small-World, Scale-Free Structures 17
Flavors of Graphs 18
Connected Versus Disconnected Graphs 19
Unweighted Graphs Versus Weighted Graphs 19
Undirected Graphs Versus Directed Graphs 21
Acyclic Graphs Versus Cyclic Graphs 22
Sparse Graphs Versus Dense Graphs 23
Monopartite, Bipartite, and k-Partite Graphs 25
Types of Graph Algorithms 27
Pathfinding 27
Centrality 27
Community Detection 27

iii
 Summary 28

3. Graph Platforms and Processing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

Graph Platform and Processing Considerations 29
Platform Considerations 29
Processing Considerations 30
Representative Platforms 31
Selecting Our Platform 31
Apache Spark 32
Neo4j Graph Platform 34
Summary 38

4. Pathfinding and Graph Search Algorithms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

Example Data: The Transport Graph 41
Importing the Data into Apache Spark 44
Importing the Data into Neo4j 44
Breadth First Search 45
Breadth First Search with Apache Spark 46
Depth First Search 48
Shortest Path 49
When Should I Use Shortest Path? 50
Shortest Path with Neo4j 51
Shortest Path (Weighted) with Neo4j 54
Shortest Path (Weighted) with Apache Spark 55
Shortest Path Variation: A* 58
Shortest Path Variation: Yen’s k-Shortest Paths 60
All Pairs Shortest Path 62
A Closer Look at All Pairs Shortest Path 62
When Should I Use All Pairs Shortest Path? 64
All Pairs Shortest Path with Apache Spark 64
All Pairs Shortest Path with Neo4j 65
Single Source Shortest Path 68
When Should I Use Single Source Shortest Path? 69
Single Source Shortest Path with Apache Spark 69
Single Source Shortest Path with Neo4j 71
Minimum Spanning Tree 73
When Should I Use Minimum Spanning Tree? 74
Minimum Spanning Tree with Neo4j 74
Random Walk 77
When Should I Use Random Walk? 78
Random Walk with Neo4j 78
Summary 80

iv | Table of Contents
5. Centrality Algorithms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Example Graph Data: The Social Graph 83
Importing the Data into Apache Spark 84
Importing the Data into Neo4j 85
Degree Centrality 85
Reach 85
When Should I Use Degree Centrality? 86
Degree Centrality with Apache Spark 87
Closeness Centrality 88
When Should I Use Closeness Centrality? 89
Closeness Centrality with Apache Spark 90
Closeness Centrality with Neo4j 92
Closeness Centrality Variation: Wasserman and Faust 94
Closeness Centrality Variation: Harmonic Centrality 95
Betweenness Centrality 97
When Should I Use Betweenness Centrality? 99
Betweenness Centrality with Neo4j 100
Betweenness Centrality Variation: Randomized-Approximate Brandes 102
PageRank 104
Influence 104
The PageRank Formula 105
Iteration, Random Surfers, and Rank Sinks 107
When Should I Use PageRank? 108
PageRank with Apache Spark 109
PageRank with Neo4j 111
PageRank Variation: Personalized PageRank 112
Summary 113

6. Community Detection Algorithms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

Example Graph Data: The Software Dependency Graph 118
Importing the Data into Apache Spark 120
Importing the Data into Neo4j 120
Triangle Count and Clustering Coefficient 121
Local Clustering Coefficient 121
Global Clustering Coefficient 122
When Should I Use Triangle Count and Clustering Coefficient? 122
Triangle Count with Apache Spark 123
Triangles with Neo4j 123
Local Clustering Coefficient with Neo4j 125
Strongly Connected Components 126
When Should I Use Strongly Connected Components? 127
Strongly Connected Components with Apache Spark 128

Table of Contents | v
 Strongly Connected Components with Neo4j 129
Connected Components 132
When Should I Use Connected Components? 132
Connected Components with Apache Spark 133
Connected Components with Neo4j 134
Label Propagation 135
Semi-Supervised Learning and Seed Labels 137
When Should I Use Label Propagation? 137
Label Propagation with Apache Spark 138
Label Propagation with Neo4j 139
Louvain Modularity 142
When Should I Use Louvain? 145
Louvain with Neo4j 146
Validating Communities 152
Summary 152

7. Graph Algorithms in Practice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153

Analyzing Yelp Data with Neo4j 154
Yelp Social Network 154
Data Import 155
Graph Model 155
A Quick Overview of the Yelp Data 156
Trip Planning App 160
Travel Business Consulting 166
Finding Similar Categories 171
Analyzing Airline Flight Data with Apache Spark 177
Exploratory Analysis 178
Popular Airports 178
Delays from ORD 180
Bad Day at SFO 182
Interconnected Airports by Airline 184
Summary 191

8. Using Graph Algorithms to Enhance Machine Learning. . . . . . . . . . . . . . . . . . . . . . . . . . 193

Machine Learning and the Importance of Context 193
Graphs, Context, and Accuracy 194
Connected Feature Engineering 195
Graphy Features 197
Graph Algorithm Features 198
Graphs and Machine Learning in Practice: Link Prediction 200
Tools and Data 200
Importing the Data into Neo4j 202

vi | Table of Contents
 The Coauthorship Graph 203
Creating Balanced Training and Testing Datasets 204
How We Predict Missing Links 209
Creating a Machine Learning Pipeline 210
Predicting Links: Basic Graph Features 211
Predicting Links: Triangles and the Clustering Coefficient 223
Predicting Links: Community Detection 227
Summary 234
Wrapping Things Up 234

A. Additional Information and Resources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235

Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239

Table of Contents | vii

 Preface

The world is driven by connections—from financial and communication systems to

social and biological processes. Revealing the meaning behind these connections
drives breakthroughs across industries in areas such as identifying fraud rings and
optimizing recommendations to evaluating the strength of a group and predicting
cascading failures.
As connectedness continues to accelerate, it’s not surprising that interest in graph
algorithms has exploded because they are based on mathematics explicitly developed
to gain insights from the relationships between data. Graph analytics can uncover the
workings of intricate systems and networks at massive scales—for any organization.
We are passionate about the utility and importance of graph analytics as well as the
joy of uncovering the inner workings of complex scenarios. Until recently, adopting
graph analytics required significant expertise and determination, because tools and
integrations were difficult and few knew how to apply graph algorithms to their
quandaries. It is our goal to help change this. We wrote this book to help organiza‐
tions better leverage graph analytics so that they can make new discoveries and
develop intelligent solutions faster.

What’s in This Book

This book is a practical guide to getting started with graph algorithms for developers
and data scientists who have experience using Apache Spark™ or Neo4j. Although our
algorithm examples utilize the Spark and Neo4j platforms, this book will also be help‐
ful for understanding more general graph concepts, regardless of your choice of
graph technologies.
The first two chapters provide an introduction to graph analytics, algorithms, and
theory. The third chapter briefly covers the platforms used in this book before we
dive into three chapters focusing on classic graph algorithms: pathfinding, centrality,
and community detection. We wrap up the book with two chapters showing how

ix
graph algorithms are used within workflows: one for general analysis and one for
machine learning.
At the beginning of each category of algorithms, there is a reference table to help you
quickly jump to the relevant algorithm. For each algorithm, you’ll find:

• An explanation of what the algorithm does

• Use cases for the algorithm and references to where you can learn more
• Example code providing concrete ways to use the algorithm in Spark, Neo4j, or
both

Conventions Used in This Book

The following typographical conventions are used in this book:
Italic
Indicates new terms, URLs, email addresses, filenames, and file extensions.
Constant width
Used for program listings, as well as within paragraphs to refer to program ele‐
ments such as variable or function names, databases, data types, environment
variables, statements, and keywords.
Constant width bold
Shows commands or other text that should be typed literally by the user.
Constant width italic
Shows text that should be replaced with user-supplied values or by values deter‐
mined by context.

This element signifies a tip or suggestion.

This element signifies a general note.

x | Preface
 This element indicates a warning or caution.

Using Code Examples

Supplemental material (code examples, exercises, etc.) is available for download at
https://bit.ly/2FPgGVV.
This book is here to help you get your job done. In general, if example code is offered
with this book, you may use it in your programs and documentation. You do not
need to contact us for permission unless you’re reproducing a significant portion of
the code. For example, writing a program that uses several chunks of code from this
book does not require permission. Selling or distributing a CD-ROM of examples
from O’Reilly books does require permission. Answering a question by citing this
book and quoting example code does not require permission. Incorporating a signifi‐
cant amount of example code from this book into your product’s documentation does
require permission.
We appreciate, but do not require, attribution. An attribution usually includes the
title, author, publisher, and ISBN. For example: “Graph Algorithms by Amy E. Hodler
and Mark Needham (O’Reilly). Copyright 2019 Amy E. Hodler and Mark Needham,
978-1-492-05781-9.”
If you feel your use of code examples falls outside fair use or the permission given
above, feel free to contact us at permissions@oreilly.com.

O’Reilly Online Learning

For almost 40 years, O’Reilly has provided technology and
business training, knowledge, and insight to help companies
succeed.

Our unique network of experts and innovators share their knowledge and expertise
through books, articles, and our online learning platform. O’Reilly’s online learning
platform gives you on-demand access to live training courses, in-depth learning
paths, interactive coding environments, and a vast collection of text and video from
O’Reilly and 200+ other publishers. For more information, please visit http://
oreilly.com.

Preface | xi
How to Contact Us
Please address comments and questions concerning this book to the publisher:

O’Reilly Media, Inc.

1005 Gravenstein Highway North
Sebastopol, CA 95472
800-998-9938 (in the United States or Canada)
707-829-0515 (international or local)
707-829-0104 (fax)

We have a web page for this book, where we list errata, examples, and any additional
information. You can access this page at https://oreil.ly/graph-algorithms.
To comment or ask technical questions about this book, send email to bookques‐
tions@oreilly.com.
For news and more information about our books and courses, see our website at
http://www.oreilly.com.
Find us on Facebook: http://facebook.com/oreilly
Follow us on Twitter: http://twitter.com/oreillymedia
Watch us on YouTube: http://www.youtube.com/oreillymedia

Acknowledgments
We’ve thoroughly enjoyed putting together the material for this book and thank all
those who assisted. We’d especially like to thank Michael Hunger for his guidance, Jim
Webber for his invaluable edits, and Tomaz Bratanic for his keen research. Finally, we
greatly appreciate Yelp permitting us to use its rich dataset for powerful examples.

xii | Preface
 Foreword

What do the following things all have in common: marketing attribution analysis,
anti-money laundering (AML) analysis, customer journey modeling, safety incident
causal factor analysis, literature-based discovery, fraud network detection, internet
search node analysis, map application creation, disease cluster analysis, and analyzing
the performance of a William Shakespeare play. As you might have guessed, what
these all have in common is the use of graphs, proving that Shakespeare was right
when he declared, “All the world’s a graph!”
Okay, the Bard of Avon did not actually write graph in that sentence, he wrote stage.
However, notice that the examples listed above all involve entities and the relation‐
ships between them, including both direct and indirect (transitive) relationships.
Entities are the nodes in the graph—these can be people, events, objects, concepts, or
places. The relationships between the nodes are the edges in the graph. Therefore,
isn’t the very essence of a Shakespearean play the active portrayal of entities (the
nodes) and their relationships (the edges)? Consequently, maybe Shakespeare could
have written graph in his famous declaration.
What makes graph algorithms and graph databases so interesting and powerful isn’t
the simple relationship between two entities, with A being related to B. After all, the
standard relational model of databases instantiated these types of relationships in its
foundation decades ago, in the entity relationship diagram (ERD). What makes
graphs so remarkably important are directional relationships and transitive relation‐
ships. In directional relationships, A may cause B, but not the opposite. In transitive
relationships, A can be directly related to B and B can be directly related to C, while A
is not directly related to C, so that consequently A is transitively related to C.
With these transitivity relationships—particularly when they are numerous and
diverse, with many possible relationship/network patterns and degrees of separation
between the entities—the graph model uncovers relationships between entities that
otherwise may seem disconnected or unrelated, and are undetected by a relational

xiii
database. Hence, the graph model can be applied productively and effectively in many
network analysis use cases.
Consider this marketing attribution use case: person A sees the marketing campaign;
person A talks about it on social media; person B is connected to person A and sees
the comment; and, subsequently, person B buys the product. From the marketing
campaign manager’s perspective, the standard relational model fails to identify the
attribution, since B did not see the campaign and A did not respond to the campaign.
The campaign looks like a failure, but its actual success (and positive ROI) is discov‐
ered by the graph analytics algorithm through the transitive relationship between the
marketing campaign and the final customer purchase, through an intermediary
(entity in the middle).
Next, consider an anti-money laundering (AML) analysis case: persons A and C are
suspected of illicit trafficking. Any interaction between the two (e.g., a financial trans‐
action in a financial database) would be flagged by the authorities, and heavily scruti‐
nized. However, if A and C never transact business together, but instead conduct
financial dealings through safe, respected, and unflagged financial authority B, what
could pick up on the transaction? The graph analytics algorithm! The graph engine
would discover the transitive relationship between A and C through intermediary B.
In internet searches, major search engines use a hyperlinked network (graph-based)
algorithm to find the central authoritative node across the entire internet for any
given set of search words. The directionality of the edge is vital in this case, since the
authoritative node in the network is the one that many other nodes point at.
With literature-based discovery (LBD)—a knowledge network (graph-based) applica‐
tion enabling significant discoveries across the knowledge base of thousands (or even
millions) of research journal articles—“hidden knowledge” is discovered only
through the connection between published research results that may have many
degrees of separation (transitive relationships) between them. LBD is being applied to
cancer research studies, where the massive semantic medical knowledge base of
symptoms, diagnoses, treatments, drug interactions, genetic markers, short-term
results, and long-term consequences could be “hiding” previously unknown cures or
beneficial treatments for the most impenetrable cases. The knowledge could already
be in the network, but we need to connect the dots to find it.
Similar descriptions of the power of graphing can be given for the other use cases lis‐
ted earlier, all examples of network analysis through graph algorithms. Each case
deeply involves entities (people, objects, events, actions, concepts, and places) and
their relationships (touch points, both causal and simple associations).
When considering the power of graphing, we should keep in mind that perhaps the
most powerful node in a graph model for real-world use cases might be “context.”
Context may include time, location, related events, nearby entities, and more. Incor‐

xiv | Foreword
porating context into the graph (as nodes and as edges) can thus yield impressive pre‐
dictive analytics and prescriptive analytics capabilities.
Mark Needham and Amy Hodler’s Graph Algorithms aims to broaden our knowledge
and capabilities around these important types of graph analyses, including algo‐
rithms, concepts, and practical machine learning applications of the algorithms.
From basic concepts to fundamental algorithms to processing platforms and practical
use cases, the authors have compiled an instructive and illustrative guide to the won‐
derful world of graphs.

— Kirk Borne, PhD

Principal Data Scientist and Executive Advisor
Booz Allen Hamilton
March 2019

Foreword | xv
 CHAPTER 1
Introduction

Graphs are one of the unifying themes of computer science—an abstract representation that
describes the organization of transportation systems, human interactions, and telecommuni‐
cation networks. That so many different structures can be modeled using a single formalism
is a source of great power to the educated programmer.
—The Algorithm Design Manual, by Steven S. Skiena (Springer),
Distinguished Teaching Professor of Computer Science at Stony Brook University

Today’s most pressing data challenges center around relationships, not just tabulating
discrete data. Graph technologies and analytics provide powerful tools for connected
data that are used in research, social initiatives, and business solutions such as:

• Modeling dynamic environments from financial markets to IT services

• Forecasting the spread of epidemics as well as rippling service delays and outages
• Finding predictive features for machine learning to combat financial crimes
• Uncovering patterns for personalized experiences and recommendations

As data becomes increasingly interconnected and systems increasingly sophisticated,

it’s essential to make use of the rich and evolving relationships within our data.
This chapter provides an introduction to graph analysis and graph algorithms. We’ll
start with a brief refresher about the origin of graphs before introducing graph algo‐
rithms and explaining the difference between graph databases and graph processing.
We’ll explore the nature of modern data itself, and how the information contained in
connections is far more sophisticated than what we can uncover with basic statistical
methods. The chapter will conclude with a look at use cases where graph algorithms
can be employed.

1
What Are Graphs?
Graphs have a history dating back to 1736, when Leonhard Euler solved the “Seven
Bridges of Königsberg” problem. The problem asked whether it was possible to visit
all four areas of a city connected by seven bridges, while only crossing each bridge
once. It wasn’t.
With the insight that only the connections themselves were relevant, Euler set the
groundwork for graph theory and its mathematics. Figure 1-1 depicts Euler’s progres‐
sion with one of his original sketches, from the paper “Solutio problematis ad geome‐
triam situs pertinentis”.

Figure 1-1. The origins of graph theory. The city of Königsberg included two large islands
connected to each other and the two mainland portions of the city by seven bridges. The
puzzle was to create a walk through the city, crossing each bridge once and only once.

While graphs originated in mathematics, they are also a pragmatic and high fidelity
way of modeling and analyzing data. The objects that make up a graph are called
nodes or vertices and the links between them are known as relationships, links, or
edges. We use the terms nodes and relationships in this book: you can think of nodes
as the nouns in sentences, and relationships as verbs giving context to the nodes. To
avoid any confusion, the graphs we talk about in this book have nothing to do with
graphing equations or charts as in Figure 1-2.
Looking at the person graph in Figure 1-2, we can easily construct several sentences
which describe it. For example, person A lives with person B who owns a car, and
person A drives a car that person B owns. This modeling approach is compelling
because it maps easily to the real world and is very “whiteboard friendly.” This helps
align data modeling and analysis.
But modeling graphs is only half the story. We might also want to process them to
reveal insight that isn’t immediately obvious. This is the domain of graph algorithms.

2 | Chapter 1: Introduction
Figure 1-2. A graph is a representation of a network, often illustrated with circles to rep‐
resent entities which we call nodes, and lines to represent relationships.

What Are Graph Analytics and Algorithms?

Graph algorithms are a subset of tools for graph analytics. Graph analytics is some‐
thing we do—it’s the use of any graph-based approach to analyze connected data.
There are various methods we could use: we might query the graph data, use basic
statistics, visually explore the graphs, or incorporate graphs into our machine learn‐
ing tasks. Graph pattern–based querying is often used for local data analysis, whereas
graph computational algorithms usually refer to more global and iterative analysis.
Although there is overlap in how these types of analysis can be employed, we use the
term graph algorithms to refer to the latter, more computational analytics and data
science uses.

What Are Graph Analytics and Algorithms? | 3

 Network Science
Network science is an academic field strongly rooted in graph theory that is concerned
with mathematical models of the relationships between objects. Network scientists
rely on graph algorithms and database management systems because of the size, con‐
nectedness, and complexity of their data.
There are many fantastic resources for complexity and network science. Here are a
few references for you to explore.

• Network Science, by Albert-László Barabási, is an introductory ebook

• Complexity Explorer offers online courses
• The New England Complex Systems Institute provides various resources and
papers

Graph algorithms provide one of the most potent approaches to analyzing connected
data because their mathematical calculations are specifically built to operate on rela‐
tionships. They describe steps to be taken to process a graph to discover its general
qualities or specific quantities. Based on the mathematics of graph theory, graph algo‐
rithms use the relationships between nodes to infer the organization and dynamics of
complex systems. Network scientists use these algorithms to uncover hidden infor‐
mation, test hypotheses, and make predictions about behavior.
Graph algorithms have widespread potential, from preventing fraud and optimizing
call routing to predicting the spread of the flu. For instance, we might want to score
particular nodes that could correspond to overload conditions in a power system. Or
we might like to discover groupings in the graph which correspond to congestion in a
transport system.
In fact, in 2010 US air travel systems experienced two serious events involving multi‐
ple congested airports that were later studied using graph analytics. Network scien‐
tists P. Fleurquin, J. J. Ramasco, and V. M. Eguíluz used graph algorithms to confirm
the events as part of systematic cascading delays and use this information for correc‐
tive advice, as described in their paper, “Systemic Delay Propagation in the US Air‐
port Network”.
To visualize the network underpinning air transportation Figure 1-3 was created by
Martin Grandjean for his article, “Connected World: Untangling the Air Traffic Net‐
work”. This illustration clearly shows the highly connected structure of air transpor‐
tation clusters. Many transportation systems exhibit a concentrated distribution of
links with clear hub-and-spoke patterns that influence delays.

4 | Chapter 1: Introduction
Figure 1-3. Air transportation networks illustrate hub-and-spoke structures that evolve
over multiple scales. These structures contribute to how travel flows.

Graphs also help uncover how very small interactions and dynamics lead to global
mutations. They tie together the micro and macro scales by representing exactly
which things are interacting within global structures. These associations are used to
forecast behavior and determine missing links. Figure 1-4 is a foodweb of grassland
species interactions that used graph analysis to evaluate the hierarchical organization
and species interactions and then predict missing relationships, as detailed in the
paper by A. Clauset, C. Moore, and M. E. J. Newman, “Hierarchical Structure and the
Prediction of Missing Links in Network”.

What Are Graph Analytics and Algorithms? | 5

Figure 1-4. This foodweb of grassland species uses graphs to correlate small-scale interac‐
tions to larger structure formation.

Graph Processing, Databases, Queries, and Algorithms

Graph processing includes the methods by which graph workloads and tasks are car‐
ried out. Most graph queries consider specific parts of the graph (e.g., a starting
node), and the work is usually focused in the surrounding subgraph. We term this
type of work graph local, and it implies declaratively querying a graph’s structure, as
explained in the book Graph Databases, by Ian Robinson, Jim Webber, and Emil
Eifrem (O’Reilly). This type of graph-local processing is often utilized for real-time
transactions and pattern-based queries.
When speaking about graph algorithms, we are typically looking for global patterns
and structures. The input to the algorithm is usually the whole graph, and the output
can be an enriched graph or some aggregate value such as a score. We categorize such
processing as graph global, and it implies processing a graph’s structure using compu‐
tational algorithms (often iteratively). This approach sheds light on the overall nature
of a network through its connections. Organizations tend to use graph algorithms to
model systems and predict behavior based on how things disseminate, important
components, group identification, and the overall robustness of the system.
There may be some overlap in these definitions—sometimes we can use processing of
an algorithm to answer a local query, or vice versa—but simplistically speaking
whole-graph operations are processed by computational algorithms and subgraph
operations are queried in databases.
Traditionally, transaction processing and analysis have been siloed. This was an
unnatural split based on technology limitations. Our view is that graph analytics

6 | Chapter 1: Introduction
drives smarter transactions, which creates new data and opportunities for further
analysis. More recently there’s been a trend to integrate these silos for more real-time
decision making.

OLTP and OLAP

Online transaction processing (OLTP) operations are typically short activities like
booking a ticket, crediting an account, booking a sale, and so forth. OLTP implies
voluminous low-latency query processing and high data integrity. Although OLTP
may involve only a small number of records per transaction, systems process many
transactions concurrently.
Online analytical processing (OLAP) facilitates more complex queries and analysis
over historical data. These analyses may include multiple data sources, formats, and
types. Detecting trends, conducting “what-if ” scenarios, making predictions, and
uncovering structural patterns are typical OLAP use cases. Compared to OLTP,
OLAP systems process fewer but longer-running transactions over many records.
OLAP systems are biased toward faster reading without the expectation of transac‐
tional updates found in OLTP, and batch-oriented operation is common.
Recently, however, the line between OLTP and OLAP has begun to blur. Modern
data-intensive applications now combine real-time transactional operations with ana‐
lytics. This merging of processing has been spurred by several advances in software,
such as more scalable transaction management and incremental stream processing,
and by lower-cost, large-memory hardware.
Bringing together analytics and transactions enables continual analysis as a natural
part of regular operations. As data is gathered—from point-of-sale (POS) machines,
manufacturing systems, or internet of things (IoT) devices—analytics now supports
the ability to make real-time recommendations and decisions while processing. This
trend was observed several years ago, and terms to describe this merging include
translytics and hybrid transactional and analytical processing (HTAP). Figure 1-5 illus‐
trates how read-only replicas can be used to bring together these different types of
processing.
According to Gartner:
[HTAP] could potentially redefine the way some business processes are executed, as
real-time advanced analytics (for example, planning, forecasting and what-if analysis)
becomes an integral part of the process itself, rather than a separate activity performed
after the fact. This would enable new forms of real-time business-driven decision-
making process. Ultimately, HTAP will become a key enabling architecture for intelli‐
gent business operations.

Graph Processing, Databases, Queries, and Algorithms | 7

Figure 1-5. A hybrid platform supports the low latency query processing and high data
integrity required for transactions while integrating complex analytics over large
amounts of data.

As OLTP and OLAP become more integrated and begin to support functionality pre‐
viously offered in only one silo, it’s no longer necessary to use different data products
or systems for these workloads—we can simplify our architecture by using the same
platform for both. This means our analytical queries can take advantage of real-time
data and we can streamline the iterative process of analysis.

Why Should We Care About Graph Algorithms?

Graph algorithms are used to help make sense of connected data. We see relation‐
ships within real-world systems from protein interactions to social networks, from
communication systems to power grids, and from retail experiences to Mars mission
planning. Understanding networks and the connections within them offers incredible
potential for insight and innovation.
Graph algorithms are uniquely suited to understanding structures and revealing pat‐
terns in datasets that are highly connected. Nowhere is the connectivity and interac‐
tivity so apparent than in big data. The amount of information that has been brought
together, commingled, and dynamically updated is impressive. This is where graph
algorithms can help make sense of our volumes of data, with more sophisticated ana‐
lytics that leverage relationships and enhance artificial intelligence contextual infor‐
mation.
As our data becomes more connected, it’s increasingly important to understand its
relationships and interdependencies. Scientists that study the growth of networks

8 | Chapter 1: Introduction
have noted that connectivity increases over time, but not uniformly. Preferential
attachment is one theory on how the dynamics of growth impact structure. This idea,
illustrated in Figure 1-6, describes the tendency of a node to link to other nodes that
already have a lot of connections.

Figure 1-6. Preferential attachment is the phenomenon where the more connected a
node is, the more likely it is to receive new links. This leads to uneven concentrations and
hubs.

In his book, Sync: How Order Emerges from Chaos in the Universe, Nature, and Daily
Life (Hachette), Steven Strogatz provides examples and explains different ways that
real-life systems self-organize. Regardless of the underlying causes, many researchers
believe that how networks grow is inseparable from their resulting shapes and hierar‐
chies. Highly dense groups and lumpy data networks tend to develop, with complex‐
ity growing along with data size. We see this clustering of relationships in most real-
world networks today, from the internet to social networks like the gaming
community shown in Figure 1-7.
The network analysis shown in Figure 1-7 was created by Francesco D’Orazio of Pul‐
sar to help predict the virality of content and inform distribution strategies. D’Orazio

Why Should We Care About Graph Algorithms? | 9

found a correlation between the concentration of a community’s distribution and the
speed of diffusion of a piece of content.

Figure 1-7. This gaming community analysis shows a concentration of connections

around just 5 of 382 communities.

This is significantly different than what an average distribution model would predict,
where most nodes would have the same number of connections. For instance, if the
World Wide Web had an average distribution of connections, all pages would have
about the same number of links coming in and going out. Average distribution mod‐
els assert that most nodes are equally connected, but many types of graphs and many
real networks exhibit concentrations. The web, in common with graphs like travel
and social networks, has a power-law distribution with a few nodes being highly con‐
nected and most nodes being modestly connected.

Power Law
A power law (also called a scaling law) describes the relationship between two quanti‐
ties where one quantity varies as a power of another. For instance, the area of a cube is
related to the length of its sides by a power of 3. A well-known example is the Pareto
distribution or “80/20 rule,” originally used to describe the situation where 20% of a
population controlled 80% of the wealth. We see various power laws in the natural
world and networks.

Trying to “average out” a network generally won’t work well for investigating relation‐
ships or forecasting, because real-world networks have uneven distributions of nodes

10 | Chapter 1: Introduction
and relationships. We can readily see in Figure 1-8 how using an average of character‐
istics for data that is uneven would lead to incorrect results.

Figure 1-8. Real-world networks have uneven distributions of nodes and relationships
represented in the extreme by a power-law distribution. An average distribution assumes
most nodes have the same number of relationships and results in a random network.

Because highly connected data does not adhere to an average distribution, network
scientists use graph analytics to search for and interpret structures and relationship
distributions in real-world data.
There is no network in nature that we know of that would be described by the random
network model.
—Albert-László Barabási, Director, Center for Complex Network Research, North‐
eastern University, and author of numerous network science books

The challenge for most users is that densely and unevenly connected data is trouble‐
some to analyze with traditional analytical tools. There might be a structure there, but
it’s hard to find. It’s tempting to take an averages approach to messy data, but doing so
will conceal patterns and ensure our results are not representing any real groups. For
instance, if you average the demographic information of all your customers and offer
an experience based solely on averages, you’ll be guaranteed to miss most communi‐
ties: communities tend to cluster around related factors like age and occupation or
marital status and location.

Why Should We Care About Graph Algorithms? | 11

Furthermore, dynamic behavior, particularly around sudden events and bursts, can’t
be seen with a snapshot. To illustrate, if you imagine a social group with increasing
relationships, you’d also expect more communications. This could lead to a tipping
point of coordination and a subsequent coalition or, alternatively, subgroup forma‐
tion and polarization in, for example, elections. Sophisticated methods are required
to forecast a network’s evolution over time, but we can infer behavior if we under‐
stand the structures and interactions within our data. Graph analytics is used to pre‐
dict group resiliency because of the focus on relationships.

Graph Analytics Use Cases

At the most abstract level, graph analytics is applied to forecast behavior and pre‐
scribe action for dynamic groups. Doing this requires understanding the relation‐
ships and structure within the group. Graph algorithms accomplish this by
examining the overall nature of networks through their connections. With this
approach, you can understand the topology of connected systems and model their
processes.
There are three general buckets of questions that indicate whether graph analytics
and algorithms are warranted, as shown in Figure 1-9.

Figure 1-9. The types of questions graph analytics answer

Here are a few types of challenges where graph algorithms are employed. Are your
challenges similar?

• Investigate the route of a disease or a cascading transport failure.

• Uncover the most vulnerable, or damaging, components in a network attack.
• Identify the least costly or fastest way to route information or resources.
• Predict missing links in your data.

12 | Chapter 1: Introduction
 • Locate direct and indirect influence in a complex system.
• Discover unseen hierarchies and dependencies.
• Forecast whether groups will merge or break apart.
• Find bottlenecks or who has the power to deny/provide more resources.
• Reveal communities based on behavior for personalized recommendations.
• Reduce false positives in fraud and anomaly detection.
• Extract more predictive features for machine learning.

Conclusion
In this chapter, we’ve looked at how data today is extremely connected, and the impli‐
cations of this. Robust scientific practices exist for analysis of group dynamics and
relationships, yet those tools are not always commonplace in businesses. As we evalu‐
ate advanced analytics techniques, we should consider the nature of our data and
whether we need to understand community attributes or predict complex behavior. If
our data represents a network, we should avoid the temptation to reduce factors to an
average. Instead, we should use tools that match our data and the insights we’re seek‐
ing.
In the next chapter, we’ll cover graph concepts and terminology.

Conclusion | 13
 CHAPTER 2
Graph Theory and Concepts

In this chapter, we set the framework and cover terminology for graph algorithms.
The basics of graph theory are explained, with a focus on the concepts that are most
relevant to a practitioner.
We’ll describe how graphs are represented, and then explain the different types of
graphs and their attributes. This will be important later, as our graph’s characteristics
will inform our algorithm choices and help us interpret results. We’ll finish the chap‐
ter with an overview of the types of graph algorithms detailed in this book.

Terminology
The labeled property graph is one of the most popular ways of modeling graph data.
A label marks a node as part of a group. In Figure 2-1, we have two groups of nodes:
Person and Car. (Although in classic graph theory a label applies to a single node, it’s
now commonly used to mean a node group.) Relationships are classified based on
relationship type. Our example includes the relationship types of DRIVES, OWNS,
LIVES_WITH, and MARRIED_TO.
Properties are synonymous with attributes and can contain a variety of data types,
from numbers and strings to spatial and temporal data. In Figure 2-1 we assigned the
properties as name-value pairs, where the name of the property comes first and then
its value. For example, the Person node on the left has a property name: "Dan", and
the MARRIED_TO relationship has a property of on: Jan 1, 2013.
A subgraph is a graph within a larger graph. Subgraphs are useful as a filters such as
when we need a subset with particular characteristics for focused analysis.

15
A path is a group of nodes and their connecting relationships. An example of a simple
path, based on Figure 2-1, could contain the nodes Dan, Ann, and Car and the DRIVES
and OWNS relationships.

Figure 2-1. A labeled property graph model is a flexible and concise way of representing
connected data.

Graphs vary in type, shape, and size as well the kind of attributes that can be used for
analysis. Next, we’ll describe the kinds of graphs most suited for graph algorithms.
Keep in mind that these explanations apply to graphs as well as subgraphs.

Graph Types and Structures

In classic graph theory, the term graph is equated with a simple (or strict) graph where
nodes only have one relationship between them, as shown on the left side of
Figure 2-2. Most real-world graphs, however, have many relationships between nodes
and even self-referencing relationships. Today, this term is commonly used for all
three graph types in Figure 2-2, so we also use the term inclusively.

16 | Chapter 2: Graph Theory and Concepts

Figure 2-2. In this book, we use the term graph to include any of these classic types of
graphs.

Random, Small-World, Scale-Free Structures

Graphs take on many shapes. Figure 2-3 shows three representative network types:
Random networks
In a completely average distribution of connections, a random network is formed
with no hierarchies. This type of shapeless graph is “flat” with no discernible pat‐
terns. All nodes have the same probability of being attached to any other node.
Small-world networks
A small-world network is extremely common in social networks; it shows local‐
ized connections and some hub-and-spoke pattern. The “Six Degrees of Kevin
Bacon” game might be the best-known example of the small-world effect.
Although you associate mostly with a small group of friends, you’re never many
hops away from anyone else—even if they are a famous actor or on the other side
of the planet.
Scale-free networks
A scale-free network is produced when there are power-law distributions and a
hub-and-spoke architecture is preserved regardless of scale, such as in the World
Wide Web.

Graph Types and Structures | 17

Figure 2-3. Three network structures with distinctive graphs and behaviors

These network types produce graphs with distinctive structures, distributions, and
behaviors. As we work with graph algorithms, we’ll come to recognize similar pat‐
terns in our results.

Flavors of Graphs
To get the most out of graph algorithms, it’s important to familiarize ourselves with
the most characteristic graphs we’ll encounter. Table 2-1 summarizes common graph
attributes. In the following sections we look at the different flavors in more detail.

Table 2-1. Common attributes of graphs

Graph attribute Key factor Algorithm consideration
Connected versus Whether there is a path between any Islands of nodes can cause unexpected behavior, such as
disconnected two nodes in the graph, irrespective of getting stuck in or failing to process disconnected
distance components.
Weighted versus Whether there are (domain-specific) Many algorithms expect weights, and we’ll see significant
unweighted values on relationships or nodes differences in performance and results when they’re
ignored.
Directed versus Whether or not relationships explicitly This adds rich context to infer additional meaning. In some
undirected define a start and end node algorithms you can explicitly set the use of one, both, or no
direction.
Cyclic versus acyclic Whether paths start and end at the Cyclic graphs are common but algorithms must be careful
same node (typically by storing traversal state) or cycles may prevent
termination. Acyclic graphs (or spanning trees) are the basis
for many graph algorithms.
Sparse versus dense Relationship to node ratio Extremely dense or extremely sparsely connected graphs can
cause divergent results. Data modeling may help, assuming
the domain is not inherently dense or sparse.

18 | Chapter 2: Graph Theory and Concepts

Graph attribute Key factor Algorithm consideration
Monopartite, Whether nodes connect to only one Helpful for creating relationships to analyze and projecting
bipartite, and k- other node type (e.g., users like movies) more useful graphs.
partite or many other node types (e.g., users
like users who like movies)

Connected Versus Disconnected Graphs

A graph is connected if there is a path between all nodes. If we have islands in our
graph, it’s disconnected. If the nodes in those islands are connected, they are called
components (or sometimes clusters), as shown in Figure 2-4.

Figure 2-4. If we have islands in our graph, it’s a disconnected graph.

Some algorithms struggle with disconnected graphs and can produce misleading
results. If we have unexpected results, checking the structure of our graph is a good
first step.

Unweighted Graphs Versus Weighted Graphs

Unweighted graphs have no weight values assigned to their nodes or relationships.
For weighted graphs, these values can represent a variety of measures such as cost,
time, distance, capacity, or even a domain-specific prioritization. Figure 2-5 visualizes
the difference.

Flavors of Graphs | 19
Figure 2-5. Weighted graphs can hold values on relationships or nodes.

Basic graph algorithms can use weights for processing as a representation for the
strength or value of relationships. Many algorithms compute metrics which can then
be used as weights for follow-up processing. Some algorithms update weight values as
they proceed to find cumulative totals, lowest values, or optimums.
A classic use for weighted graphs is in pathfinding algorithms. Such algorithms
underpin the mapping applications on our phones and compute the shortest/cheap‐
est/fastest transport routes between locations. For example, Figure 2-6 uses two dif‐
ferent methods of computing the shortest route.

Figure 2-6. The shortest paths can vary for otherwise identical unweighted and weighted
graphs.

20 | Chapter 2: Graph Theory and Concepts

Without weights, our shortest route is calculated in terms of the number of relation‐
ships (commonly called hops). A and E have a two-hop shortest path, which indicates
only one node (D) between them. However, the shortest weighted path from A to E
takes us from A to C to D to E. If weights represent a physical distance in kilometers,
the total distance would be 50 km. In this case, the shortest path in terms of the num‐
ber of hops would equate to a longer physical route of 70 km.

Undirected Graphs Versus Directed Graphs

In an undirected graph, relationships are considered bidirectional (for example,
friendships). In a directed graph, relationships have a specific direction. Relationships
pointing to a node are referred to as in-links and, unsurprisingly, out-links are those
originating from a node.
Direction adds another dimension of information. Relationships of the same type but
in opposing directions carry different semantic meaning, expressing a dependency or
indicating a flow. This may then be used as an indicator of credibility or group
strength. Personal preferences and social relations are expressed very well with direc‐
tion.
For example, if we assumed in Figure 2-7 that the directed graph was a network of
students and the relationships were “likes,” then we’d calculate that A and C are more
popular.

Figure 2-7. Many algorithms allow us to compute on the basis of only inbound or out‐
bound connections, both directions, or without direction.

Flavors of Graphs | 21
Exploring the Variety of Random
Documents with Different Content
Epic Verse Quintus Ennius fl. Epic (on Roman Copies the
Ennius 200 history and Homeric
b.c. legend) called hexameter
Annales. and borrows
the Olympic
deities.
Called the
“Roman
Homer,” but
crude and
inartistic.
Publius VIRGIL lived AENEID (epic of Copies Odyssey in The basis f
VERGILI 70- Aeneas, first six books subseq
US Maro 19 legendary (wanderings), epics.
b.c. founder of and Iliad in Utilized
Roman last six Dante,
people). (battles); Ariosto
borrows Tasso,
images and in Italy
incidents Model
from all the Fre
Greek epic, e
writers. But Voltaire
more Henria
descriptive,
philosophical,
and
fastidious of
expression.

Marcus Lucan lived Pharsalia (epic of Departing further Two favour

Annaeus a.d. civil wars from Greek writers
LUCANU 39- between directness. the ea
S 65 Caesar and Cleverness in middle
Pompey). rhetoric, ages.
epigram, Statius
description, affecte
satire, etc., Dante
aimed at. Boccac
Fondness for (who
details of founds
horror. Teseide
upon h

Publius Statius fl. Thebaid (epic of Clever and facile

Papinius (“Stace” a.d. Thebes and its verse:
STATIUS in 70 heroes). elegant
Chaucer) simile, etc.,
of more
importance
than the
matter.
Lyric Verse Gaius Valerius Catullus lived Poems (odes, Imitates metre His works l
CATULLU 84- epigrams, and and style of during
S 54 occasional Greek lyrists Middle
b.c. pieces— (Sappho, Ages, a
especially love- etc.) and always
poems to Alexandrian read th
“Lesbia”). elegists Horace
(Callimachus,
etc.). Most
Greek of all
Romans in
his simplicity
and
spontaneity.

Quintus HORACE lived ODES AND EPODES Avowed imitation The model
HORATIU 65-8 (love, politics, of Sappho, lyrists
S Flaccus b.c. vers de Alcaeus, writers
société, Pindar, social
moralizings). Anacreon. verse i
Adapts Greek France
lyrical Italy.
metres. Deft
but
unimpassione
d.

Elegiac Verse Albius Tibullus Tibullus fl. Elegies (of affection Direct imitations Exerted an
Sextus Propertius 20 and of of Ovid
b.c. sentiment). Alexandrian
Aurelius
Propertius Greek
elegists.
Publius OVID fl. Various Poems, e.g., The Greek models Ovid was a
OVIDIUS 43 Heroides (in are less the ea
Naso b.c.- the form of epigrammatic elegies
a.d. letters). Tristia, . Ovid affects affecte
17. Amores, etc. pointed the Ro
couplets represe
(compare by the
Pope). and of
English
and po
made g
Chauce
are mu
favour
his Me
not in
verse,
Satiric Verse Gaius Lucilius Lucilius fl. Satires on politics, A native Latin Models for
120 literature growth much
b.c. (fragments), (Greek satire Italian
etc. takes a satire
different form (Aretin
and etc.), a
medium). French
Quintus HORACE 65-8 SATIRES AND (Regni
HORATIU b.c. EPISTLES Satyre
S Flaccus (genial). Ménipp
Boileau
etc.).
Aulus Persius Persius fl. Satires (crabbed
Flaccus a.d. style)
60

Decimus JUVENAL fl.SATIRES (polished,

Junius a.d. terse,
JUVENAL 120 trenchant).
IS
Didactic (and Titus LUCRETIUS fl. DE RERUM In form follows
“philosoph LUCRETI 60 NATURA (“the old Greek
ical”) US Carus b.c. Constitution of philosophical
Verse Nature”). poets, and in
matter
expounds the
philosophy of
Epicurus.

Publius VIRGIL lived GEORGICS (poems The idea taken [See Hesiod
VERGILI 70- on from Hesiod.
US Maro 19 husbandry).
b.c.
Quintus HORACE lived ARS POETICA (an An ill-digested The source
HORATIU 65-8 essay in rechauffé of the sha
S Flaccus b.c. literary Aristotle and criticis
criticism). later Greek Boileau
critics. and his
school
Pastoral Verse VIRGIL BUCOLICS (or From Theocritus, Imitated by
ECLOGUES). but moralized Mantua
and and
sometimes Sannaz
artificial. in Italy
and Ma
in Fran
Epigram Marcus MARTIAL fl. EPIGRAMS (various The conception of The moder
Valerius a.d. subjects). Greek is entir
MARTIAL 90. epigram is form. M
IS polish and
delicacy. Of
Latin it is
chiefly point
and sting.
History Gaius Julius CAESAR lived COMMENTARIES
CAESAR 100- (on the Gallic
44 and the Civil
b.c. Wars; simple,
straightforward
narrative).
Gaius Sallust fl. Catilina and Attempts to
SALLUST 45 Jugurtha imitate
IUS b.c. Thucydides.
Crispus A
commonplace
moralizer.
Titus LIVIUS LIVY Ob. HISTORY (of Adopts the Greek
Patavinus a.d. Rome); (rich custom of
17 style, ample, putting
pathetic). verbatim
speeches into
the mouths
of his
characters.
Gaius TACITUS fl. HISTORIES and Aims at the
Cornelius a.d. ANNALS (of condensation
TACITUS 100 Emperors); of
(epigrammatic, Thucydides.
terse,
satirical).
Oratory Marcus CICERO (and lived Speeches (59 extant, Follower of The model
Tullius “Tully.”) 106- e.g., Philippics, Demosthenes most
CICERO 43 Against Verres, , but in a French
b.c. etc.). more rotund oratory
and loaded and
style. preach
(Bossu
etc.)
Marcus Quintilian fl. The Training of the (Other
Fabius a.d. Orator. Seneca
Quintilian 100 followe
us

Letter-Writing CICERO 106- Letters (“To A specially Roman Type follow

43 Atticus,” “To department in Fran
b.c. Friends”). of literature. (Mada
Gaius Pliny (the fl. Letters (to friends, de
PLINIUS Younger.) a.d. to Trajan, Sévign
Secundu 100 etc.). etc.).
s
Philosophy CICERO Academica, De A reproducer of
Officiis, etc. Greek
systems in
popular
expositions.
SENECA A brilliantly
epigrammatic
moralizer on
old lines of
thought.

Fable Phaedrus Phaedrus fl. Fables Reproducer of See “Aesop

a.d. Aesop
15
Encyclopaedic Gaius Plinius Pliny (the fl. Natural History Storehouse
Secundu Elder.) a.d.
s 70
 III
LITERARY CURRENTS OF THE DARK AGES

Latin literature, despite its decline after the classical period, is

marked by a number of names which merit eminence in their several
domains. The era succeeding the silver age hardly deserves to be
called leaden. Literature does, indeed, both descend from the
Virgilian and Ciceronian style of language, and also adopt a less
classic attitude in its themes and sentiment, but it is not without a
life and value of its own. Some of the writers are pagan, some are
Christian, but their religious professions are not to be determined by
their dates. Apuleius, the African writer, a professional rhetorician
and man of letters, who wrote his prose Metamorphoses or Golden
Ass in the second century, is, of course, a pagan, and by no means a
model one. The work just mentioned, probably based on current
folk-tales, is entirely fiction, narrating the story of a man turned by
sorcery into an ass, and describing his adventures, scandalous,
distressful, or amusing, in the hands of robbers and other low types
of a society which, we may trust, was not really so bad as it is here
painted. Yet into this otherwise not very edifying work there comes
the exquisite story of Cupid and Psyche, which has been so
frequently translated or recast in literature—best of all by William
Morris in the Earthly Paradise—and so frequently utilized as the
subject of pictorial or plastic art.
From the beginning of the third century until the fifth, Christian
views find their exponents in Tertullian, Lactantius, Ambrose,
Prudentius, Jerome, and Augustine. To Jerome is due in particular
that Latin version of the Bible of which the present Vulgate
represents successive partial revisions, to Augustine the City of God,
to Ambrose the initiation of the Christian hymn, and to Prudentius its
development. Christian also is the Gallo-Roman poet Ausonius, of
the later fourth century, but his verse is by no means dedicated to
Christian teaching. In him appears what might seem to be a
modern, if not fully “romantic,” partiality for affectionate observation
of natural scenery, best illustrated by his well-known description of
the stream and banks of the Moselle.
Meanwhile among pagan writers must be reckoned Ammianus, a
picturesque and interesting historian, who undertook to bring the
work of Tacitus up to the year 378; Macrobius, whose Saturnalia
discourses in a desultory fashion on a variety of literary and social
topics; and Claudian, the composer of polished poems on
contemporary history, in which extremely skilful polish of verse is
united to brilliant gifts of description. The religion of Boethius, the
last man of letters who can be said to linger on the border of the
classical world, but who in style and thought stands nearer to it than
many an earlier writer, is doubtful. In all probability he was a pagan,
but he concerned himself, not with religion, but with philosophy as
reflected from Plato. His De Consolatione, or Consolations of
Philosophy, is a prose work interspersed with verses, and in virtue of
this production, which often rises to great excellence, Boethius stood
to the Dark Ages for the exemplar of the philosopher. His place in
mediaeval reading was a very high one, and may be gauged from
the fact that in England Alfred the Great translated his Consolations
into Anglo-Saxon, though with insertions and comments of his own.
To Chaucer, as to all the mediaeval world, “Boece” was part of the
staple library.

During the centuries from the decay of the literature of Rome till
the emergence of the modern literatures of western Europe there
occur the great migrations of conquering peoples and the forming of
the new nations. The Gothic conquests of Italy and Spain, the
movements of the Franks, Burgundians, and Lombards, the story of
Alaric or Theodoric, of Pharamond or Clovis, belong to history, as do
the settlement of the Anglo-Saxon and Danish tribes in Britain and
the occupation of Normandy by the men from whom it is named.
Against these Teutonic triumphs and their influence to the north
must be set the Moslem triumphs and influence to the south. Not
only did the Moors conquer and hold for centuries the greater part of
the Spanish peninsula, but Sicily also passed for five generations into
the hands of the Saracens. In England a national history commences
with Alfred at the close of the ninth century, but its development,
both from a literary and social point of view, was deeply modified at
the Norman Conquest. In France and Germany the empire of
Charlemagne, the great fact of the eighth century, had done much
towards consolidating culture and reviving learning. At the end of
the eleventh century began the Crusades, which helped to bring the
western nations into closer touch with each other and also into
contact with the Greek world and with legends of the east.
Meanwhile, the Christians of north-west Spain were gradually
winning back their country from the Moors, but, in the process,
absorbing no little of their Arabic culture.
By the twelfth century the modern Romance tongues, Italian,
French, Provençal, and Castilian, are sufficiently formed for literary
purposes, and the speakers of those languages have attained to the
position of steady and settled communities. Though the English
language is temporarily in abeyance for literary uses, the English
nation is free from further disturbance, while nevertheless it is now
happily placed in direct communication with continental tendencies
and ideas. In the meantime it must not be forgotten that Europe had
now become Christian, and that in the west the teaching of one
great Church was common to all the nations.
This long period of disintegration and reconstruction is for the
most part so little studied, and is, in fact, comparatively so
studiously ignored, that we are apt to forget how long it actually
was. The literary productions of nearly seven hundred years are
regarded as of so little moment that we forget there were any at all.
Yet for a proper comprehension of the inter-relations of literature as
affecting the development of our own, it is necessary to form some
conception of the various literary currents of these “Dark Ages.”
 As might perhaps be anticipated after a survey of the historical
movements and situations, we have to reckon with:
(1) Such Latin literature, of classical or later date, as survived after
the wreck of the empire and still formed part of, at least, the higher
reading.
(2) Such new productions in Latin as appeared before the new
tongues were formed.
(3) The matter and influence of the literature of the Church,
comprising the Hebrew Scriptures, chiefly in the shape of the
Vulgate, commentaries, moral works, and also religious legends,
lives of saints, and the like.
(4) The material and spirit brought in by the Teutons in the shape
of their own old epics and sagas, with the myths which formed their
basis.
(5) The Celtic feeling, traditions, and compositions which made
their way into the répertoire of such countries as contained a Celtic
population.
(6) The learning, literary matter, and literary art of the Saracens,
whether introduced by way of Spain or by that of Sicily, and whether
derived from Oriental or from Greek sources.
(7) Literary influences from the Greek world, including remnants
of classical and post-classical compositions, mediaeval productions of
Byzantium, and tales of the East which had been rendered into a
Greek form.
It is difficult to disentangle these various threads, which interlace
each other in complex ways, but on the whole the most satisfactory
procedure will be to make a note or two upon each. Such notes will
necessarily be brief to the point of mere hinting.
(1) It was, perhaps, to be expected that, with the decline of Latin
culture, the “fittest” part of Latin literature to survive in the
knowledge of the semi-barbarized west should be that which lacked
the highest artistic qualities. It is only with the dawn of the first
Renaissance, which led up to and was assisted by the great Tuscan
trio, that the true classics began to reappear among the common
reading of men of superior learning. Virgil, indeed, was not wholly
forgotten, nor was Cicero, and in the age of Charlemagne there was
promise of a much wider scope. But, unless with the piously inclined
—and often even with them—the Dark Ages were more interested in
scraps of miscellaneous information containing a spice of the
wonderful, derived and garbled from Pliny, in stories with a similar
spice of the marvellous and, by preference, of the licentious, such as
are to be found in the Satyricon of Petronius and the Golden Ass of
Apuleius, or in traditions of the art of love culled from Ovid and
crudely transmitted. For those of a more serious turn there were the
mild philosophizing of Boethius, the history of the Spaniard Orosius,
and the encyclopaedic educational medley of Martianus Capella; and,
for the religious, the hymns of Prudentius served as a model. Yet,
though sparingly met, the reading of literary Latin never quite failed,
and verses, for example, continued to be written as much in the
style of Claudian as writers could command. Latin comedy was not
unknown to the monasteries, since the German nun Hrotswith is
found in the tenth century composing prose imitations of Terence. It
is impossible, in the defect of our material, to tell with any precision
the extent to which Latin reading was directly kept in vogue. Capella
and Orosius, at least, were accepted as standard works, but in
respect of the legends, stories, mythologies, and pseudo-marvels of
natural history, such matter as shows itself at the birth of the new
literatures had in a large measure come back in roundabout ways
and through other channels.
(2) It is more easy to name the chief Latin productions of the Dark
Ages themselves. If we regard Boethius as the last figure in Roman
literature proper, the series consists mainly—for the sixth century—of
the voluminous writings of Cassiodorus, historical and educational;
the informal History of the Franks by Gregory of Tours; and the work
of the Goth Jordanes concerning his own people. To the seventh
century belong the Christian and didactic Moralia of Gregory the
Great, and the encyclopaedic Origines of Isidore of Seville. To these
we must add two writers of Great Britain; the one, Gildas, who
wrote in Wales, in the middle of the sixth century, his dolorous
account of the conquest of Britain by the Saxons; the other, the
Englishman Bede, whose Ecclesiastical History and biographical
works belong to the eighth. The age of Charlemagne, with its
vigorous encouragement of education, consolidated all the
promiscuous learning of the time, in which style plays a part
altogether subordinate to the multifarious contents. From a literary
point of view the creations above-named are of little moment to our
subject, except in so far as information and misinformation from this
uncritical mass of material found its way into all the work of our pre-
Renaissance writers. Their chief merit is that they kept the channels
of classical influence from being completely blocked. We must,
however, note one important innovation in literary form. This was
the introduction of rhyme into Latin hymns. The exact source of the
novelty is unknown, but it began as early as the fourth century, and,
together with Arabic influence, it helps to account for the use of
rhyme which became current in the neo-Latin countries before their
modern languages produced a real literature.
(3) The Hebraic influence which came through Christianity is as
obvious as it was far-reaching. Every step in the Christianizing of
Europe meant the conveying, not only of new sentiment and new
ways of regarding things, but also of new materials in the way of
Biblical history, however distorted in perspective. From the new
doctrines of self-mortification there grew legends of the saints; from
the traditions of their sanctity, legends of miracles; from the
persecutions, legends of the martyrs. Both the Old and the New
Testament already existed in a Latin form even before the more
competent and authoritative version of Jerome (about a.d. 400). It
should also be observed that the Bible which was thus rendered
accessible was then read, far more than in later times, as a book
containing matter interesting in itself, and therefore to be utilized
and recast in story, apart from its uses in theology. Meanwhile round
the original Scriptures both the earlier and later “Fathers” built up
large masses of comment. When we remember that in the Dark
Ages it was the churchmen who kept alive literary cultivation and
production, and that the Bible narrative, the legends, martyrologies,
and Christian doctrines were conveyed to every mind by sermons
and other agencies, it is manifest how extensive must have been the
effect upon thought and matter before the newly forming literatures
emerged. On actual literary art and style it is true that there could
be but little palpable influence, until, or unless, the Bible came to
exist and to be widely read, as it eventually did in England, in the
vernacular. But of this something more will be said in a later section
of this book.
(4) With the Teutonic invaders of France there came in the spirit of
feudal relationship. For centuries this spirit survived. Combined with
the Celtic exaltation which is so pronounced in the Arthurian
legends, and also with the sentiments of Christianity, it became
embellished into the well-known mediaeval conception of knighthood
with its vows of utter loyalty and self-devotion. The way was thus
prepared for the knightly, or chivalrous, romances which are to be
described in the chapter on the literature of France.
But, besides this feudal spirit shown in the Franks, there had
already existed among the Germanic tribes before their settlement in
France or Britain an orally transmitted literature. Its form was epic,
and its themes the superhuman exploits of heroes among scenes of
slaughter and carousal, in contest with huge monsters, and under
the dispensation of rude pagan deities, Woden, Thor, and the rest of
the Teutonic pantheon. Between the fourth and sixth centuries this
heroic poetry of Germany grows into appreciable form, and both the
Franks of the Continent and the Anglo-Saxons of England bring with
them their several portions. In Germany itself it is much later that
the Nibelungen Lied is edited into a connected shape; but to
England there came in the sixth century the epic legend of Beowulf,
of which the source is to be found in events of southern Sweden and
the Western Baltic dimly recorded. This poem was edited in Christian
times, and with some Christian additions, during the literary
flourishing of northern England in the early part of the eighth
century. Another poem carried from the mainland by the Englecyn
was the Song of Widsith (the Far-Traveller), a wandering gleeman
who has much to say of the deeds and generosity of the Gothic and
other German chiefs among whom he roamed “as his fate willed,”
and to whom he “unlocked his word-hoard.”
In point of matter this Germanic contribution to Dark Age
literature is perhaps of little account. But its vigour of action and
strenuous temper did no little towards determining the virility of the
French chansons de geste, which formed so large a portion of
English reading in the pre-Chaucerian period. In point of form it is
necessary to note that the Anglo-Saxon method of versification,
based on accent, alliteration and assonance, is naturally inherited
from the German tradition. With very slight modification the method
of Anglo-Saxon poetry is also that of Langland in his Piers Plowman
of the Chaucerian age. Though this was subsequently abandoned by
English poets in favour of the French system of rhyme and
numbered syllables, the use—all the more artistic for being disguised
—of alliteration and accent has survived as one of the chief formal
beauties of all our poetry.
Whereas the Teutonic poetry, when it came in contact with
Christianity in England or France, soon lost its characteristic themes,
its mythology, and much of its savagery, the older matter and spirit
still flourished among the pagan Norsemen, and were re-imported
into northern England with the invasions of the Danes.
(5) More distinguishable and pronounced effects upon the
literature of western Europe were produced by a backward invasion
of the Celtic themes and temperament. There was much Celtic blood
in northern Italy and in Spain, still more in France and the British
Islands. When once the Celtic temperament emerged in literature it
was sure of a ready and wide response. Perhaps no such emergence
would have happened if it had not been that in the Dark Ages the
Christian scholars—the only authors of that epoch—were especially
cultured and ardent in Wales and Ireland. The racial and patriotic
feelings of the British Celts were pathetic and intense, and, whether
among those in western Britain, or among the emigrants to Brittany,
the exploits of their race were celebrated in song marked by a high
spirit of pride, as well as by a peculiar mysticism and a remarkable
sentiment of chivalry and romance. The actual contributions of the
Celts to our own literary making are the subject of brief remark
elsewhere.
(6) The influence of the East during the period before the first
Renaissance was of no small importance. The language through
which it came, but not often the language in which it originated, was
the Arabic of the Saracens, whether as invaders of Spain or of Sicily.
It is precisely while the literary state of Europe was at its lowest that
the Saracenic culture was at its height. Into Spain, where the Moors
had established themselves in splendour and opulence, there
followed all the learning of the Semitic East, in philosophy, natural
science, and medicine, together with the literary forms of the Arabs
and the music of their accompaniments. Though the western
Saracens were politically altogether separate from the Caliphate of
Bagdad, the literary language was common to the Moslem world,
and men of learning and artistic gift—whether Arabs or Jews—were
equally welcome at either end. In the reign of Abd-ur-Rahman, in
the early part of the tenth century, there particularly flourished in
Moorish Spain the light verse of love and its gay surroundings.
Meanwhile Cordova developed what was practically a University, to
which congregated all manner of Oriental talent, and in which
studies in science and philosophy were prosecuted with zeal. Nor
was the diffusion of all this culture restricted to the Arabs or their
Spanish subjects. Many Christians from other parts of western
Europe sought a knowledge of mathematics or medicine at Cordova,
nor were these severer accomplishments all that the visitor would
acquire.
To literature proper the true Arabs would have contributed little. In
their original home their poetry had mostly taken the shape of the
qasîda, a loosely connected ode, in which an introduction concerning
the forsaken camping ground was regularly followed by reflections
on the singer’s love affairs, and these by thoughts concerning his
desert wanderings, his steed, and finally his chief. Of most
importance to us, perhaps, is the fact that this Arabic verse was in
rhyme, and that short odes, or ghazels, of fourteen lines, appear to
anticipate the sonnet, a form which arose in Sicily in a court
frequented by cultured Moslems.
After the establishment of Islam, the new religion at first exercised
a cramping effect, but the same fondness for rhyme (which, indeed,
was associated with notions of sacred or magic power) introduced it
even into the prose of the Koran. When the Saracen conquest had
extended widely and included Persia, the superior culture of the
Persians gave them, from about a.d. 750, a predominating influence
at the court of the Abbasids at Bagdad. Arabic literature, therefore,
widened its forms and themes, and its poetry now embraced lyrics of
love and wine, satires and elegies, largely of Persian origin. Of this
poetry in general it may be said that it is marked by a peculiar
predilection for sententious wisdom in the shape of proverbs and
aphorisms, and for fables and allegories which convey similar
maxims. These, we shall find, appear in full force in Spain, where
they are converted into part of the earliest literature in Spanish. For
the collection of such fables the Arabs and Persians could reach a
hand in either direction. From the west they could take the Greek
fables of Aesop and convert them into the Arabic fables of Loqman;
from the east they could gather the Indian fables of Pilpay (or
Bidpai), translated first from the Indian Pancatantra into the Middle
Persian (better called Pehlevi), and thence by Muqaffa into the
Arabic Kalila and Dimna. In this collection the actions of the beasts
serve subtly to convey to a prince rules of wise conduct, more moral
than the later principles of Macchiavelli. The Orientals showed an
equal passion for purely romantic stories, provided that they
contained wonderful and magical occurrences, much prowess, and
luscious suggestions of magnificence and pleasure. The Thousand
and One Nights, better known as the Arabian Nights, form an
immense body of such compositions, which have been perpetually
translated and re-translated, and which are still among the standard
books of the world.
 But the Saracens were by no means sunk in sententiousness or
frivolity. They were impassioned for philosophy and science,
especially the sciences of mathematics, astronomy, medicine, and
chemistry. For their acquirements in these directions they were
indebted to the Greeks, and chiefly through the Syrians of
Mesopotamia. Here Hellenism, introduced by Alexander, had grown
into peculiar strength, nor was the Greek blood itself inconsiderable.
From the Syrians the Arabs derived their knowledge both of Aristotle
and of Plato, although, from their practical turn of mind, it was the
Aristotelian philosophy which they mainly affected, and which
passed into the famous Arabic translation of Averrhoes. Carried to
Cordova, much of this learning, and particularly that derived from
Aristotle, was disseminated through western Europe. The Arabic
influence on thought, reflected from Greece, was therefore great.
From a more purely literary point of view, we must reckon with the
introduction of Oriental apologues and tales, although many of
these, as will be seen immediately, come in also from eastern
Europe through a Greek medium.
In point of form it is impossible not to conclude that the minstrelsy
and poetry which prevailed in Moorish Spain contributed liberally to
the fashioning of the troubadour poetry of Provence. The itinerant
Arab minstrel was not welcomed solely by Moors; he played his part
among the true Spaniards, and Spaniards themselves turned
minstrels after the same fashion. The eastern, or Catalonian, part of
Spain was in language virtually identical with the neighbouring south
of France, and no border separated the Catalonian minstrelsy from
the Provençal districts. In 1112 the Count of Barcelona became the
ruler of Provence, and in his train followed all the poetry and song
which had grown familiar in Catalonia. It is dangerous to attempt to
decide the more and less of direct borrowing; but the manner of the
troubadour, his rhymes, his themes of the tenso, the planh, and the
morning and evening songs, so closely recall the machinery and
devices of the Saracens, that the affiliation can hardly be denied.
(7) Direct effect of Greece upon Europe to the West was in
abeyance from the fall of Rome till the Renaissance. Occasionally
indeed, but very seldom, we hear of scholars who read some Greek,
and Theodore of Tarsus actually visited and taught in Anglo-Saxon
England in the later seventh century. But such influence of Greek
work as appears during the dark and mediaeval times comes only in
circuitous ways and from inferior writers of inferior matter. It for the
most part appears in stories derived from the post-classical Greek
romances, or from Oriental tales first translated into Greek and then
recast into Latin.
Greek romance itself—beginning as early as the second century,
but mostly produced at uncertain dates from the fourth century
onwards—at once betrays an Oriental atmosphere. Its genesis is not
so much in the Greek mind as in the eastern mind, with which the
empire of Alexander had brought the Greeks into contact. The
writers commonly hail from Asia Minor or Alexandria, and the scenes
and adventures are apt to be Babylonian, Syrian, or Egyptian. Their
chief features are much the same. A number of unlikely and
inconsequent adventures, comprising separations and stratagems of
lovers, travels, voyages, dangers, pirates, magic, murders,
descriptions, and dreams, are tediously repeated with unessential
variations. One of the first examples, it is true, deals with
wanderings in the north and west, among Celts and Cimmerians.
This is the Marvels beyond Thule of Antonius Diogenes. But the
Babylonica of Iamblichus and the Aethiopica (or Theagenes and
Chariclea) of Heliodorus have their mise-en-scène in the east, with
events and wonders in the Oriental style. The latter work enjoyed a
special vogue, and portions of its contents were not scorned even in
comparatively late times by Italians like Tasso and Guarini, and
Frenchmen like Hardy and Racine. This, together with the Leucippe
and Cleitophon of Achilles Tatius, and the pastoral romance of
Daphnis and Chloe by Longus, played no small part in the
conception of the French sentimental romances of the seventeenth
century, beginning with D’Urfé and carried on by Scudéry and La
Calprenède. The work of Longus is on the whole the most important,
since it contains the new element of pastoral setting and description
and some novelty of simple sentiment. In the Dark Ages themselves
we cannot tell how far these productions were known in any direct
form in the west, but at least we know that nothing travels more
quickly than stories. Another romance, with the usual elements of
love and adventure, and with the addition of “recognition of the
long-lost,” was the famous story of Apollonius of Tyre, of which we
possess only the Latin version, through which the tale was passed
westward. This work was favourite reading in the age with which we
are here concerned. It was translated even into Anglo-Saxon, and
later came in again as an English version of a French rendering.
Of a different character was the Barlaam and Josaphat of John of
Damascus, an ecclesiastical writer of the eighth century. The story is
derived from Buddhist sources in India. Though magic plays its part,
the whole is naturally of a moral and theological turn. The mediaeval
world found it vastly interesting, for after its conversion into Latin by
Vincent of Beauvais in the thirteenth century, it passed into nearly
every European language which could pretend to a literature.
Meanwhile, through Greek versions, there came in tales of purely
non-Greek construction. Chief among these was the work known to
more modern times as the Seven Wise Masters, originally an Indian
production, styled the Parables of Sandabar. This was turned into
Persian or Arabic, then into Greek under the name of Syntipas,
thence, in the thirteenth century, into Latin as Dolopathos, and
thence again versified into French.
All this material appears and reappears in the fabliaux of France,
in the earliest novelle of the Italians, and naturally in Boccaccio.
Other productions popular in the Dark Ages, of special note as the
storehouse upon which the French trouvères in particular drew for
their classical cycles of romances (to be dealt with in the chapter on
French literature), were those attributed to “Dares Phrygius,” “Dictys
Cretensis,” and “Callisthenes.” If we place the two former under the
head of Greek work, it is because of their ascription to Greek writers
and their possible derivation, at least in part, from lost Greek
sources. They deal with the story of Troy, ostensibly from
complementary points of view—a Trojan and a Greek. That there
actually was some sort of history by Dares of Phrygia appears from a
passage in Aelian; but the book On the Destruction of Troy, in which
mediaeval readers put their simple trust, is a Latin production of a
date probably not earlier than the sixth century a.d., although it
pretends to be a translation of Dares by the classical writer Nepos.
Similarly an actual Greek Dictys of Crete apparently did write an
account of the Trojan war and the Greek heroes, but the book in
actual use was but a fourth-century production in Latin, asserting
itself to be a translation. Portions of these two compilations were
versified, transfused, and invested with an atmosphere of mediaeval
chivalry, by Trouvères, including the Norman-English Benoît de
Sainte-More, whom again Guido Colonna, in the thirteenth century,
exploited for his Latin History of the Trojan War, a work which
became the standard reference for “matter of Troy” as it appears in
Chaucer, Lydgate (Troy Book), and Gower. It is not from Homer, but
from these pseudo-classical accounts, that we derive such episodes
as those of Troilus and Cressida.
For the cycle of Alexander the same generation of Trouvères and
their English followers were indebted to a late Byzantine writer, who
pretended to be the Greek Callisthenes, contemporary of the great
Macedonian. In point of fact his History of Alexander is an
imaginative mixture of passages culled from history with eastern
stories and marvels. It is, of course, in a Latin version that this
farrago became known to the authors of the romans.
We must not forget the vogue during these ages, devoted as they
were to tale and apologue, of the fables of “Aesop.” Of these there
were in mediaeval times various versions and collections, some
derived directly from the Latin Phaedrus, who had versified from the
Greek; others from the later Greek remodelling by Babrius; others
again from an Arabic collection, which combined a compilation of the
Greek with a compilation from the Indian Fables of Bidpai (or
Pilpay). One early version, of uncertain provenance, was that of King
Alfred; and it was apparently a general massing of all this material
which, after passing through German and French hands, became the
famous Esope of Caxton.
 To all that literary matter which pretended to classical antiquity the
Middle Ages, entirely lacking historical perspective, gave the
comprehensive name of “Roman.” How freely that term was used,
and how miscellaneous had been the sources of legend, is
manifested in the strange medley of the thirteenth or fourteenth
century, known as the Gesta Romanorum, in which fragments of
classical history, legends of saints, and Oriental stories, are
combined without the least notion of their relations or
contradictions. To the Gesta every writer, whether in England,
beginning with Chaucer, or in Italy, beginning with Boccaccio, had
free recourse for the matter of his poems or his plots.
 IV
FRENCH LITERATURE AND ENGLISH

It will be remembered that the influence of the literature of Greece upon that of England
has been exerted in various ways, direct and indirect, and at various epochs; and that it
continues still to operate upon us rather more than less, affecting both the matter and the
form of what is written in our midst. The literature of Latin, again, has always exercised an
influence on every generation, Latin forms and thoughts being imbedded in our English
writings beyond all enucleation or analysis. The literature of mediaeval and Renaissance Italy,
we shall find, had indeed much to do with shaping and polishing the literature of England
during the three hundred years from the time of Chaucer to the time of Milton, but since the
last-named period it has played little part in determining what our authors shall say, or how
they shall say it.
Prior in date and influence to that of Italy comes the literature of France, with the debts in
substance or in manner which we are bound to acknowledge to our neighbours across the
Channel. Our purpose does not require that we should pretend to traverse the whole history
of French literature. If we dwell upon a certain number of salient topics or famous names, it
is because they in particular represent the chief types in the development of French literary
history, and either directly or indirectly affect the evolution of our own.
France has, during civilized times, been politically and socially, as well as geographically, so
near to us; Englishmen and Englishwomen have been generally so well acquainted with the
French language and French books, that it is beyond possibility to determine exactly what
effect French models have had and are having upon us, just as on the other hand it is
beyond possibility to analyse exactly the effect which English models have had and are
having upon France. But, without aiming at this impossible exactitude, we may at least make
ourselves aware of such periods and manners of French influence as yield themselves readily
to the student’s survey.
It will be found that, though the influence of French literature has been felt in every
generation, there are two great periods in particular during which the creations and the
critical principles of Frenchmen have dominated those of our own authors. The one is the
period between the Conquest and the rise of Chaucer; the second is the period which began
in the seventeenth century with writers of the age of Waller and Dryden, and continued till
towards the end of the eighteenth century, that is to say, till the time of Cowper and Burns.
Approximate dates are, perhaps, necessary here, and the following may roughly serve. From
about the year 1100 till about the year 1370, and from about the year 1660 till 1780,
England took its cue in many departments of literary work from the matter, the form, and the
critical principles of contemporary France. Doubtless at all times there have been borrowings
to and fro, but these are the periods when the borrowings have been most one-sided and
most palpable. The interval from the maturity of Chaucer till the earlier part of the
seventeenth century was more especially the era of Italian influence, introducing and
supporting that mightier influence from pagan Greece and pagan Rome which began in what
is justly styled the Renaissance. Again, since the latter part of the eighteenth century, the
time heralded by Cowper and crowned by Burns, the English have emancipated themselves
from direct literary imitation of the French, although, as is briefly stated at the end of this
chapter, there have been no few currents of French influence upon various classes of our
writers, and, from them, upon the reading public.
Let it then be repeated that two periods especially concern us—the period of the Norman
and Plantagenet kings preceding and reaching up to Chaucer, and that period which
embraces the literature of the reigns of Charles II, James II, William III, Anne, and the first
two Georges; or, to put the latter period more plainly and more suitably in a literary
connection, the age of Dryden, of the “Social School,” of the comedy of Wycherley and
Congreve, of the essays of Addison and the Spectator, of the verse of Pope, of the prose of
David Hume and Samuel Johnson.
The former period corresponds to the era of influence from the Provençal Troubadours and
the Northern French Trouvères, from the epic chansons de gestes, the several kinds and
cycles of “romances,” the allegories, fabliaux, and other creations of which we must take
some special account. The second period answers in particular to the names of Corneille,
Racine, Molière, Boileau, Voltaire, and of a number of famous French novelists, letter-writers,
and critics. How and in what manner these authors came to tyrannize so completely and so
long over English literature will require some terse statement. For the rest, in the period from
the writers of the romans and allegories down to Corneille, and again in the period from
Rousseau to the present, we shall speak of French authors only as links in the chain of
French development in itself, with a passing reference to any value they may have
individually for the literature of England.

The greater part of the land of Gaul—the modern France—was at an early date occupied
mainly by Celts, akin to, though not precisely identical with, the present Bretons of the north-
west corner of the country. There were also Germans in the north-east, and Euskarians in
the south-west. Under the Roman empire the land was gradually overrun with Roman
settlers, Roman merchants, and Roman soldiers, and Latin naturally became the official
language, the language of high society, of literature, and of education.
The mixed people in process of time thus came to speak a provincial Latin, and to call
themselves “Romans.” In reality they were very far from being true Romans, and their
speech almost was as far from being true Latin. It was both corrupted and also broken up
into local dialects. It was, in fact, a blend of Latin with influences from the various native
peculiarities. Early in the fifth century a body of Franks, a German people and speaking a
German language, invaded the heart of Gaul and permanently held its northern half. It is
from them, the Franks, that the whole country obtained the name of “France.” These
conquerors brought many a German word—mostly of war and feudalism—into the language
of the conquered, and likewise hastened the corruption of their “Latin” syntax. The old Latin
of culture became more and more widely severed from everyday speech, and hence
“Romance,” the corrupted language of these modern “Romans” of Gaul, was regularly used
as a term in direct opposition to the old and literary “Latin.” It came, in fact, to mean the
vulgar tongue. It was about the year 800 that, in the northern half of Gaul, the popular or
Romance speech was formally recognized. In the tenth century, the Northmen descended on
much of this region, and became its masters. Meanwhile the southern half of Gaul, which
had been subdued by other German peoples, the Visigoths and Burgundians, was forming its
own particular corruption of the Latin, and, among the dialects which arose in that division,
the dialect of Provence, in the south-east, took earliest shape and clearest predominance.
 Before entering upon any account of “French” literature, we must remove from our minds
the conception of France and French as they are, and try to see them as they were in the
eleventh and twelfth centuries. North of the Loire are various provinces and a distinctly
marked Romance language, the langue d’oïl, or “French.” A Celtic attachment, which has
immigrated from Great Britain, exists in Brittany, much Norman blood in the north, and a
Frankish influence has modified the Gallo-Roman staple. South of the Loire are other states,
and, for the most part, another Romance language, the langue d’oc, or “Provençal,” with
leanings rather to an Italian kindred on one side, and a Catalonian-Spanish on the other.
Strictly speaking, the langue d’oc extended over the country south of a line drawn from
about Charente to the Alps, while Provençal is properly the language of the south-eastern
portion of that area. Corresponding to the two divisions of Gaul there arose two different
forms and two different spirits of literature, one “French,” one “Provençal.” Later it was a
joining of these two forms and spirits (though with a very distinct predominance of the
northern) which produced modern French literature, or “French literature” in the ordinary
sense; and it was both of them, though chiefly the northern, which largely controlled
England during two centuries before Chaucer, and so contributed to the making of that poet
and his age.
There is a fact too often forgotten by students of English writing and even of English
history. It is that until Chaucer’s time England was only a portion of the King of England’s
dominions; the rest was on the continent, in France. Under Henry II, King of England in the
later twelfth century, more than half of modern France, namely, Normandy and other
provinces north of the Loire, Poitou, Aquitaine and Gascony south of it, were part of an
English or Franco-English empire. At the peace of Brétigny in 1360, Poitou, Guienne, and
Gascony were still left a portion of the realms of Edward III. This fact of the oneness of
England and much of France is of very great importance to early English literature. The court
and official tongue, and, in a large degree, the literary language, of England were in any
case French. The intercourse between England and the langue d’oïl, and (though less
continued) between England and the langue d’oc, was, moreover, intimate and frequent. The
writers and minstrels were, in a considerable measure, common to England and to both
northern and southern France. No few of the writers belonging to old literature in French,
e.g., Walter Map, and Benoît de Ste. More, had their homes in England; among them was
Marie de France. The channels of communication were constantly open, and the current
flowing and ebbing through.

A concise account must first be given of the two Gallic literatures, “French” and Provençal.
Provençal flourished early, and enjoyed but a brief life. We may, therefore, trace this branch
first, then the northern French, and afterwards compare and combine the two.
Though northern France had its song of Roland and other chansons at as early a date as
the love lyrics of Provence, yet, if literature implies conscious art and system, Provençal
composition is—with the exception of Anglo-Saxon—the first real literature of modern
Europe; it stimulated Spain on the one hand, and Italy on the other; but it was in advance of
either. It is earlier than Dante, and, although it is appreciably indebted to Ovid, and in some
degree to Virgil, it is anterior to the influence of the classical forms or spirit of the first
Renaissance. It began, helped by Moorish or “Arabian” impulses and lessons, in the eleventh
century, enjoyed a brilliant existence for two hundred years, and died with the dying langue
d’oc. Though it was never enriched and made immortal by the work of any one transcendent
genius, it can boast a large number of composers possessed at least of talent and taste.
Provençal verses became models for all neighbouring countries. Frederick Barbarossa in
Germany, Richard Cœur-de-Lion in Anglo-French England, Alfonso II in Aragon, Frederick II
in Sicily, these royal personages went out of their way to compose in the fashionable style
and rhythm of Provence. They became, in fact, troubadours. The terms needs some
explanation. A troubadour is not properly a wandering minstrel carrying a guitar. That
itinerant minstrel is an inferior order of person, the jongleur (in Provençal joglar). He stands
to the troubadour as the Anglo-Saxon “gleeman” to the “maker.” The troubadour was the
“finder,” the poet, generally a noble, a knight, sometimes even a prince. It is no doubt true
that the jongleur, who originally sang the troubadour’s ditties, was fain, like other inferiors, to
assume the higher rank, try his own hand at composing variations, and call himself a
troubadour, and so the title became degraded. It is true also that the real troubadours
frequently chanted their own songs of love and glory, and so helped to cause confusion
between themselves and the mere jongleur minstrels. But the troubadour proper was one
who travelled sumptuously mounted and attired, to be the honoured guest of châtelains and
princes.
Nearly all this Provençal literature of three centuries of troubadours is lyric, not epic. It is
generally singing, not narrating, and its theme is chiefly personal feelings. Rhymes, which
had, it is true, been sparsely employed in monkish compositions in Latin, were then novel
things in European literature, although long and universally used by the Saracens. The
Provençal poets cultivated rhymes which grew more and more varied and complicated; with
careful elaboration of soft and harmonious sounds they sang of two things, and almost only
of two, to wit, love and glory, gallantry and chivalry in both senses and connections. The
verses were love verses or martial verses, celebrating loyalty in love and valour in arms. As a
class they are without pretension to any profoundness of imagination or to any sublimity.
Their excellence is their music, not any translatable substance of thought. It must be
confessed that the songs and subjects lacked variety; the same tricks of expression and
“conceits,” the same nouns and adjectives, the same situations, the same “fantastic
sentimentality,” would reappear monotonously, and would inevitably suggest the artificial and
unreal. One could hardly be expected to read extensively in the cansos, or love-songs, of
those who called themselves the “gentle troubadours,” without a feeling of satiety. The
serenade (serena), the morning greeting (alba), the dispute of lovers (tenso or joc parti), the
lament (planh), which were recognized species of troubadour effort, inevitably suffered from
exhaustion of material. Nevertheless one cannot but be impressed with the chivalrous idea of
love which many of the Provençal poets professed, and according to which they nearly
always treat that passion, vaunting a devoted tenderness and a delicate and sentimental
worship. The influence of this idea, as still further refined and ennobled in Tuscany, is
palpable in the attitude of Dante towards Beatrice, and of Petrarch towards Laura; there are
many traces of its influence in Chaucer and his contemporaries. Through the Petrarchist
sonneteers it again reaches England in the Elizabethan age.
It must be enough merely to mention the names of Bernard de Ventadour, Bertran de
Born, Pierre Vidal, and Arnaud Daniel, among famous troubadours. But a word may be said
of that remarkable institution, the “Court of Love,” to which a poem of Chaucer (or more
probably of some one with a large share of Chaucer’s mind) owes its conception and its title.
During the later generations of the “gentle troubadours,” the way to speak and think of love
and gallantry was reduced to a system. It was made a science—called el gai saber, “the gay
science”—which every poet was supposed to understand and to have at his finger-ends. One
favourite form of poetry was the tenso, a dispute or altercation between troubadours upon
delicate questions and scruples of behaviour and feeling in affairs of love. It became the
fashion for noble ladies in those idle, rather frivolous, but doubtless not unhappy days to
hold mock courts, in which poets sang one against the other, like opposing advocates;
whereafter the court gave its decision, or arrêt d’amour, and awarded prizes to that
troubadour whose arguments and verse were most in keeping with the code prescribed by
the gay science. “Is it a greater grief to lose a lover by death or by unfaithfulness?” may
serve as an example of the subjects particularly favoured in these poetical courts of the
ladies of Gascony, of the Countess of Champagne, or of Queen Eleanor.
Such, briefly, was the genuine Provençal literature—lyrics of love and bravery, with here
and there a pastoral, and here and there a poem of censure or satirical criticism. But true
epics and romances of adventure, sustained allegories, witty tales of common life, they had
practically none. For these we must look to northern France, to the land not of the
troubadours, but of the Trouvères. Trouvère is the French form corresponding to the
Provençal troubadour, and equally means the “finder,” who is indeed the “poet.” But in
northern France there existed different social conditions and a different clime; there was also
the sterner stuff which belongs to Franks and Normans, while in Brittany there was the Celt,
with all his melancholy fire and imaginative and mystical emotion. The lyric literature of the
north blossomed, indeed, somewhat later than that of Provence, and is largely an imitation
of it. The romances of the trouvères are also distinctly infused with the ideas and style of the
lyric south. Nevertheless the great mass of the poetry of northern France is of its own
creation in both matter and spirit. It is the poetry of the epic, the allegory, and the tale; the
poetry of the romance of heroic adventures, of satirical teaching, and of stories to amuse; in
other words, it produced the chanson de geste, the roman, and the fabliau. It is in every way
stronger than the creations of the south, in seriousness, in vigour, in variety, in invention.
According to Ten Brink, prouesse, the masculine side of chivalry, is more northern, while
courtoisie, the feminine side, is more Provençal. But the difference goes much deeper than
any pair of terms can express it.
The old French poems of heroic adventure, blent with more or less of that other-world
imaginative quality known as “romance,” fall into three main cycles of subjects. They deal
with Charlemagne and his Paladins (in which case they are more truly epic in character, and
are called chansons de geste, a term properly thus restricted to incidents in supposed French
history), or with King Arthur and his knights (together with the once independent legends of
Tristram), or with classical heroes, whether of the Trojan legend or, like Alexander, of actual
history. These three cycles have been named the “Carlovingian” (or “French”), the
“Arthurian” (or “British”), and the “classical” (or “Grecian”). Or we may make four by
subdividing the last into “Trojan” and “Alexandrian.” At the time of their composition the cycle
which dealt with the classical subjects of antiquity was said to deal with “matter of Rome.” All
antiquity was “Rome,” and all ancient heroes “Romans.” We find, then, songs of Roland and
Oliver, romances of Tristram or Launcelot, romances of Alexander the Great, and many more.
Some romans, called d’aventures, are independent of any cycle and make no pretence at all
to be history. The twelfth and thirteenth centuries are rich in the poetical narratives which
tell of heroic feats, or miracles of devotion and loyalty, mixed with much of the supernatural
withal. This was the day of the Crusades, of conquests of the Saracens, single combats,
adventures in distant lands, where dwarfs and enchanters, dragons and giants, were
supposed to dwell; and nothing pleased the venturous barons more than to be told such
tales to the music of the itinerant jongleurs. A further variety of these songs of exploit was
the lai, which is too short and too lyrical to be an epic or a roman, and is rather the song of
an epic episode. The allegorical poems—which they also called romans, as being similarly
akin to epic and written in the vulgar tongue—long and tedious as they are to us, were not
disdained by Chaucer, and gave the cue to several conspicuous works of the Chaucerians
down to the sixteenth century. That most famous of all, the Roman de la Rose, was
translated by our master of English undefiled. This poem, begun by Guillaume de Lorris as an
“Art of Love,” after the manner of Ovid (as filtered through Provence), was continued a
generation later by Jean de Meung as a satirical miscellany of learning and legend. It is all
about a lover who sought to pluck a rose, about his difficulties in reaching it, about the
abstract qualities which help or hinder him, about personified virtues and vices, such as
Dame Idleness, who lets him into the garden, Avarice, Meanness, Hatred, who stand in his
way, Fair-Seeming, who has much to say in the matter, and numerous others. Thanks to
these agencies, it takes that unhappy youth some 23,000 verses before he attains to
plucking the object of his affections. Yet it was this reading which inspired the earliest efforts
of our Chaucer, and which, in his first stage, he fell to imitating. It was this literature which
the cultivated Norman English delighted hugely to hear. Allegorical, also, and purely satirical
is the prolonged beast-fable known as the Roman de Renard (“the fox”), which enjoyed an
immense vogue throughout Europe, and provoked countless imitations.
One chief species of composition, and a highly important one, yet remains. This was the
fabliau, the amusing tale in verse, the one kind of writing to let us know that the world was
not wholly made of doughty knights and gentle damosels. The fabliau is a tale of real or
possible adventure in ordinary life, generally of a humorous kind. It is, in fact, a sort of
novelette. Chaucer’s Canterbury Tales are often fabliaux. From France the fabliau was
borrowed by the Novellieri of Italy. It was taken up and developed by Boccaccio, and both
directly from France and indirectly through Italy it made its way into the general stock of
European narrative material. Had it not been for fabliaux, one might have thought that in
those days there was nothing else for men to do but fight and love. Yet the great ordinary
mass of mankind existed all the time, doing its sober work in towns and country places. And
it was time for this great stratum to find recognition in bourgeois story.
Thus in northern France we have chiefly epics and romances of heroic adventures,
allegories more or less satirical and didactic, and amusing tales; in southern France we have
chiefly lyrics of love and chivalry. As time goes on, each half moves closer to the other,
although during the whole epoch of the northern romances the Provençal spirit had
combined with the Celtic to pervade them with a peculiar tone of chivalric sentiment. In 1249
the two geographical divisions became one kingdom. Before the middle of the fourteenth
century the Provençal tongue begins to die, and its literature to perish. The story of French
literature thenceforward is one and undivided.

While the troubadour and trouvère literatures were thus flourishing in the two halves of
France, the cultivated circles of Norman and Plantagenet England found those literatures
sufficiently adapted to their needs. The ordinary language of these circles was identical with
that of the trouvères, and at the same time the English possessions in Languedoc, including
the cultured centre Bordeaux (gained by the marriage of Henry II with Eleanor of Guienne
and Poitou), brought the Court into direct communication with the lyrics of the troubadours.
Henry’s son, Richard Cœur-de-Lion, was himself a troubadour and the friend of troubadours,
in particular of Bertran de Born. But better suited to the Anglo-Norman temperament, and, of
course, completely intelligible to the French-speaking barons and gentry, were the romances,
lais and fabliaux of the trouvères. The work of the Norman Wace (Geste des Bretons) in
1155 was as much intended for England as for France. So also was the Roman de Troie
(1160) of Benoît de Ste. More, which included the story of Troilus. The French Saint Grael
stories of Walter Map (1180) and the lais of Marie de France (1210) were produced in our
island, and were the common property of England and Norman France. The jongleurs
wandered from baron’s court to baron’s court, and the stories of Arthur,

Of Greece and Troy the strongë stryf,

There many a man lost his lyf;
Of Brute that baron bold of hand,
The first conqueroure of Engeland;
Of King Arthur, that was so riche,
Was not in his tyme him liche;
How Kyng Charles and Rowland fought,
With Sarazens nold they be caught;
Of Tristrem and Ysoude the swete,
How they with love first gan mete;
Of King John and of Isumbras,
Of Ydoine and of Amadas;
Stories of diverse thinges,
Of princes, prelates and kinges;
Many songes of divers rhyme,
As Engelish, French and Latyne.

Before French literature could make much further advance, it must pass, after that of Italy,
under pupilage to the Renaissance. As in England of the fifteenth century, there is first a
period of stagnation, and then one during which France is borrowing and assimilating to its
utmost lessons in thought and style, in form and substance, from the lately recovered
classical masterpieces of Greece and Rome, as well as from the Italian writers who first
enjoyed and exploited these treasures. During these stagnant and growing stages of French
literature it exercises comparatively little effect upon our own.
The Renaissance naturally reached France before it extended to England, and the
Renaissance meant in France what it had meant in Italy, and what it afterwards came to
mean in England, namely, the widening of the intellectual and moral horizon, broader
knowledge and broader views, a shaking off of old and dry traditions. And therewith it also
meant greater variety of subjects in literature and the reign of better models of thought and
expression. The effect of the Renaissance on French literature was to draw the thoughts of
authors away from the old monotonous round of romances and allegories, and at the same
time away from the old monotonous expressions and phrases; to make them attack all
interesting subjects of thought, and meanwhile to adapt and polish the instrument of
language which expressed them. This it was which the recovery of the Greek and Latin
classics accomplished for Italians, Frenchmen, and Englishmen alike, supplying them with
new range and scope, with new patterns and principles.
But as in England, so in France, this new birth and literary reformation did not exercise its
full effect immediately. In England it gradually culminated in the Elizabethan age, in France it
only attained its full development in the seventeenth century. That is to say, it was actually
slower of progress in France than in our own literature.
For a time, while the first influences of the Renaissance were being felt, the effect in
France was, as in England, a severance from the old subjects and methods, without a full
adoption of the classical subjects and principles. The classical influence acts as a solvent
before it comes to act as a crystallizing agent. There is, in fact, a transition period, during
which writing is left free to attempt various forms. If a man of natural genius arises in such
an epoch, he will give us his natural self, and so may create us prose or verse which, despite
a deficiency in knowledge, will be immortal through its own truth and strength. If on the
other hand at such a transition period men who write are lacking in native power, they will
write much worse when they follow no models and adhere to no principles. In England,
during the transition from the epoch of Chaucer to the time of Wyatt and Surrey, there
appeared no distinguished poetic genius, and, except among Scottish writers like Dunbar,
little more than tiresome production. In France, on the contrary, there were two
distinguished poets, Villon (fifteenth century) and Marot (early sixteenth). These were
stimulated by the new ideas, but were not yet dominated by the new classical models. They
were freed from the mediaeval shackles, and not yet fettered in the bonds of
misapprehended and misapplied “classical” principles.
François Villon wrote during the latter half of the fifteenth century, and is principally known
by his ballads, which were something quite new to French literature, and have, one may
venture to say, remained unique therein. From the old artificial romances and allegories he
breaks clean away. He is as original and independent as our poet Burns, whom, by the way,
he somewhat resembled in personal character. His merit, like that of the Greeks whom he did
not know, lies in his truth, in the candid expression of his own personal emotions, in his
naive confessions, in his sincere pathos. We all sympathize with emotions and confessions of
this nature, and therefore Villon, like Burns, possesses a permanent and a universal value.
And not only is he true in sentiment, he is clear and direct in his phrase, and musical in his
verse.
Clément Marot opens the sixteenth century. He has been called the father of modern
French poetry. If this means that he wrote with an ease and sprightliness, and a vein of
urbane satire, which are usually associated with the esprit gaulois, but which skim rather
along the surface of things, it is true. But if it means that he is the consummation of the
Renaissance, and that the critical principles of French poetry were established in his time, it
is without truth. For Marot, like Villon, is a poet and an artist without following the despotic
rules which afterwards came to prevail in France, and he furthermore sought his themes
rather in the old French subjects, the romances and the fabliaux, than in realms of classical
antiquity. The Italian influence, however, touches him and leads him into pastorals, which,
we must note, were known to Spenser.
Villon and Marot are both of the transition period, and not wholly of the Renaissance. They
both fell short in one great respect; they lacked depth and elevation. This is a vice to which
all French verse is prone, setting, as it does, so special a value on form; but it was the more
discoverable in these two poets, because the rich intellectual nutriment of antiquity had not
yet been assimilated by them, because their minds laboured under the intellectual poverty of
mediaevalism, because, in other words, they lacked the substance with which the best
ancient literature is crammed. Their poetry has many blossoms, but bears little fruit. Yet they
mark one great step in progress. They are emancipated from the old mediaeval artificiality.
 While Villon and Marot were thus emancipated, there were others during this transition
time who were not by any means so. On the one hand they allegorized, like the trouvères, to
the utmost; their subjects were obsolete and unreal; on the other, their language was trivial
and their contents uneven. Verse literature seemed to need bracing and correction in the
light of advancing study of the Greek and Latin masterpieces, and it is to the administering of
such correction that we come in what is known as the Pléiade.
The Pléiade, or constellation of the seven stars, was the term applied to seven men of
letters, who formed themselves into a coterie or league about the year 1550, with the
professed resolve of reforming the French language and French literary methods. The
conception is very French. This cool manner of looking at language and literary expression as
subject to definite laws of art, which may be codified by a league or academy, is contrary to
English notions. Not so with the French. They have no desire for impulsive and perhaps
erratic individuality. This is one of their clearest characteristics. Of the Pléiade the two
greatest names are Ronsard, the poet, and Du Bellay, who was both poet and manifesto-
writer. Their object, as stated by themselves, was to bring French literature nearer to the
classical models of Greece and Rome, and to create a nobler form and use of the language
for literary purposes. And while Du Bellay was to write his manifesto, Ronsard was to give a
practical illustration of the theory, by himself composing odes and sonnets in the proper
style. The attempt was bold, and it was successful. For fifty years all French literature
“Ronsardized.” Here are a few sentences of the manifesto concerning the Défense et
Illustration de la Langue Française. “Our ancestors have left us our language so poor and
bare, that it stands in need of the ornaments, and, so to speak, the features of other
people.... By what means can we hasten its development? By the imitation of the ancients....
Translating is not a sufficient means of elevating our vulgar tongue to an equality with the
most famous. What must we do? Imitate! Imitate the Romans as they did the Greeks!... We
must digest the best authors and convert them into blood and nutriment.... You that mean to
be a poet, read and re-read the Greek and Latin models. Then leave all those old French
rondeaux, ballades, virelais, chants royaux, chansons and other such vulgarities (épiceries),
which corrupt the taste of our tongue, and only serve to testify to our ignorance. Throw
yourself on those witty epigrams in imitation of Martial!... Distil with a flowing style tender
elegies after the manner of Ovid and Tibullus!... Sing me some of those odes as yet
unknown to the French tongue ... and let there be nothing in which does not appear some
trace of rare and ancient learning.”
We need not agree with all this breezy advice. It is impossible to re-create a language all
at once. If there is not inspiration, there cannot be good poetry, though one may have
infinite good models to follow. Nevertheless the new school was a success for half a century,
and both Ronsard and Du Bellay, though often mechanical and often flat, have left a few
imperishable sonnets and other pieces. Our own Elizabethans not only read Marot and his
contemporary Saint-Gelais (who introduced the Petrarchan sonnet into France), they also
read Du Bellay, who finally established the sonnet and at the same time served as a pattern
for English writers. One writer of the Ronsardist school, Du Bartas, was a writer of real
religious conviction, and his Semaine or Week of Creation, translated by Sylvester, gained no
small currency in England.
What calls for particular notice in this connection is the deliberate way in which French
writers and critics can contemplate and formulate the principles and methods of good
literature. The English, to whom so much of French verse is cold and mechanical, may
perhaps think that it is this same formulating which has done incalculable harm to poetry, a
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

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

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookfinal.com