Download the full version of the ebook at

https://ebookultra.com

An Introduction to Transformational Syntax

Roger Fowler

https://ebookultra.com/download/an-introduction-
to-transformational-syntax-roger-fowler/

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

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Non Transformational Syntax 1st Edition Robert Borsley

https://ebookultra.com/download/non-transformational-syntax-1st-
edition-robert-borsley/

ebookultra.com

A transformational analysis of Turkish syntax Robert H.

Meskill

https://ebookultra.com/download/a-transformational-analysis-of-
turkish-syntax-robert-h-meskill/

ebookultra.com

An Introduction to Criminological Theory Third Edition

Roger Hopkins Burke

https://ebookultra.com/download/an-introduction-to-criminological-
theory-third-edition-roger-hopkins-burke/

ebookultra.com

Mass Media Research An Introduction 9th Edition Roger D.

Wimmer

https://ebookultra.com/download/mass-media-research-an-
introduction-9th-edition-roger-d-wimmer/

ebookultra.com
Criminal Justice Theory An Introduction 1st Edition Roger
Hopkins Burke

https://ebookultra.com/download/criminal-justice-theory-an-
introduction-1st-edition-roger-hopkins-burke/

ebookultra.com

Thai Syntax An Outline Udom Warotamasikkhadit

https://ebookultra.com/download/thai-syntax-an-outline-udom-
warotamasikkhadit/

ebookultra.com

Syntax A Minimalist Introduction 1st Edition Andrew

Radford

https://ebookultra.com/download/syntax-a-minimalist-introduction-1st-
edition-andrew-radford/

ebookultra.com

Syntax Workbook A Companion to Carnie s Syntax Andrew

Carnie

https://ebookultra.com/download/syntax-workbook-a-companion-to-carnie-
s-syntax-andrew-carnie/

ebookultra.com

An Introduction to Teaching Gill Nicholls

https://ebookultra.com/download/an-introduction-to-teaching-gill-
nicholls/

ebookultra.com
An Introduction to Transformational Syntax Roger
Fowler Digital Instant Download
Author(s): Roger Fowler
ISBN(s): 9781138207639, 1138207632
Edition: Reprint
File Details: PDF, 17.58 MB
Year: 2016
Language: english
 ROUTLEDGE LIBRARY EDITION:
SYNTAX

Volume 9

AN INTRODUCTION TO
TRANSFORMATIONAL SYNTAX
C\
~-
Taylor & Francis
Taylor & Francis Group
http:// tayl ora ndfra nci s.com
 AN INTRODUCTION TO
TRANSFORMATIONAL SYNTAX

ROGER FOWLER
First published in 1971 by Routledge & Kegan Paul Ltd.
This edition first published in 2017
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
711 Third Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 1971 Roger Fowler
All rights reserved. No part of this book may be reprinted or reproduced or utilised
in any form or by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying and recording, or in any information
storage or retrieval system, without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks or registered
trademarks, and are used only for identification and explanation without intent to
infringe.
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library

ISBN: 978-1-138-21859-8 (Set)

ISBN: 978-1-315-43729-3 (Set) (ebk)
ISBN: 978-1-138-20763-9 (Volume 9) (hbk)
ISBN: 978-1-315-46149-6 (Volume 9) (ebk)

Publisher’s Note
The publisher has gone to great lengths to ensure the quality of this reprint but
points out that some imperfections in the original copies may be apparent.
Disclaimer
The publisher has made every effort to trace copyright holders and would welcome
correspondence from those they have been unable to trace.
 Roger
Roger Fowler
Fowler

AnAn introduction
introduction to to
TransformationalSyntax
Transformational Syntax

Routledge & Kegan Paul London

Routledge & Kegan Paul London
 First published in Great Britain 1971
by Routledge and Kegan Paul Ltd.
Broadway House
68-74 Carter Lane
London, EC4V SEL
Printed in Great Britain by
The Camelot Press Ltd
London & Southampton
and set in Monotype Times New Roman,
10 point, 1 point leaded
© Roger Fowler, 1971
No part of this book may be reproduced in
any form without permission from the
publisher, except for the quotation of
brief passages in criticism

ISBN 0 7100 6975 8 (c)

ISBN 0 7100 6976 6 (p)
 Contents

Preface vii
one What is a Grammar? 1
two Deep and Surface Structure 10
three Coostituent Structure: 8yntaetic FunctioDS 21
four CODStitaent Structure: Categories and DerivatioDS 35
five Lexical InterpretatioD 49
six Deixis: Det and Au 61
seven DerivatioD of a Simple 8eDtence 77
eight Negatives, Passives, QaestiODS and Similar Structures 85
nine Pro-forms 100
ten Complex Senteaees 115
eleven Some NomiaaHzatioDS 129
twelve Relative Clauses and Adjectives Relisited 139
thirteen CoDjoiDing Tnmsformatioos ISO
fourteen Postscript: LiDgaistic Universals and the Nature
of Grammatical DescriptioDS 164
Selected Reading 174
Index 177
Taylor & Francis
Taylor & Francis Group
http://taylorandfrancis.com
 Preface

This is a textbook on transformational-generative syntax, a mode of

grammatical description proposed by the American linguist Noam
Chomsky in a little book called Syntactic Structures published in
1957. Since that time, TG, as it has come to be called, has undergone
massive growth and change. In the first place, its general framework
has become accepted by the majority of Western linguists as provid-
ing the most reliable and revealing version of linguistic analysis. This
fact has to be acknowledged despite intractable opposition from a
few representatives of older schools of linguistics or of more insular
traditions; and despite many disagreements about details of the
proposed analysis. Second, TG has benefited from very substantial
and useful revisions over the years.
The net consequence of these developments is that, although there
is an increasingly wide demand for information on TG, current
writings in the field are forbiddingly specialized and somewhat dis-
putatious, and the older books have become somewhat out of date.
Available elementary textbooks fall into two categories: there are
those which were published in the early and mid-1960s, and project
a version of TG which is not entirely consonant with more recent
statements of the approach; and there are newer books~ an increas-
ing number - which embody fragments of contemporary revisions in
what is sometimes a puzzling way. I have attempted to provide a
'compromise' account. The primary intention is to describe a
transformational model of syntax which is more up to date than the
classic textbooks, based as they are on Syntactic Structures, can
provide. In essence, this means incorporating the general changes
announced in Chomsky's Aspects of the Theory of Syntax (1965) and
foreshadowed in Katz and Postal's Integrated Theory of Linguistic
vii
 Preface
Descriptions (1964). At the same time, I have tried to avoid making
Syntactic Structures 'unreadable' through uncompromising para-
phrase of Aspects: I assume that any student who uses the present
book as a way of gaining access to contemporary syntactic theory
will be interested enough to read Syntactic Structures, which, though
now superseded in many respects, remains the most succinct,
powerful and attractive argument for a transformational approach
to syntax.
Readers who are familiar with the history of TG will soon realize
that the present book is not a faithful paraphrase of Aspects. I
would claim that it builds on the basic framework of that account -
although even that claim may be controversial. Prematurely, I feel,
the whole position of Aspects is under attack from some quarters.
What I have tried to do is tidy up such contradictions and omissions
as appear in Aspects without, in my opinion, invalidating the overall
position. In an attempt to reflect contemporary work I have gone
beyond the letter of Aspects (hence my reading of it could be called
inaccurate) in several respects: a more extensive use of feature
analysis in syntax, and, in particular, a new treatment of Det and Aux
which is not envisaged in Aspects. Beyond this up-dating of Aspects, I
have tried to indicate directions of subsequent enquiry by other
grammarians: for this reason, my treatment of pronouns, relative
and appositive clauses, and conjoining has been worded in a tentative
and open-ended way - these are current preoccupations in syntactic
research and I want to suggest that further rethinking in these areas
may bring important and radical revisions to the very basis of the
grammar.
The grammar presented here, then, is by no stretch of the imagina-
tion 'final': it is a provisional grammar designed to help students
read both classic and contemporary writings in TG. Certain obvious
limitations of the present model of syntax make it clear that it is quite
provisional: I would point especially to the difficulty of explaining
adverbials and phrasal conjunction in this version of transforma-
tional syntax. A student who realizes just what these particular
difficulties are will be well equipped to evaluate both older and newer
solutions to such problems.
A word on how this book is to be read. It is a textbook and an
instrument, to be used rather than consulted. The material is
presented sequentially, with modifications en route, and so it should
be read slowly, from beginning to end. It is not a reference book, and,
to discourage its use as such, the index is minimal. I assume that the
book will normally be used in a taught course, in conjunction with
other reading materials. At the end of the book there is a short
reading list: this reflects a range of important books and articles
viii
Taylor & Francis
Taylor & Francis Group
http://taylorandfrancis.com
 one

What is a Grammar?

The word 'grammar' in present-day linguistics has at least two

important meanings. On the one hand, we say that a speaker knows
the grammar of his language. He usually does not know it con-
sciously -unless he has some special training in linguistics, he cannot
talk confidently about the nature of his grammar. A grammar in this
first sense comprises the linguistic knowledge speakers possess which
enables them to communicate in their language. 'Grammar' here is a
psychological, mentalistic, concept. The second sense relates to the
linguist, not to the speaker: the linguist is said to write a grammar
of the language. This grammar is a formal, explicit, description of the
language.
Now these two usages must be kept apart. One look at a printed
grammar is enough to convince us that it is extremely unlikely that
the speaker knows his grammar as an object of the shape the linguist
provides when he writes his grammar. If we could magically 'tap'
the speaker's hidden linguistic knowledge - by hypnosis, drugs or
whatever other implausible technique - so that he could tell us
directly what it is that he knows which we refer to as 'his grammar',
he would not simply dictate Jespersen's Modern English Grammar or
Chomsky's Syntactic Structures to us. The speaker docs not store
his linguistic knowledge in the format which the linguist adopts for
explanatory purposes; nor, when he produces sentences, does he
follow step-by-step the processes which the linguist spells out as he
constructs derivations for sentences. This latter point is most
important, and I will return to it: a linguist's grammar generates
sentences; a speaker produces (and understands) sentences- the two
processes are quite independent.
Although the two senses of 'grammar' must be dissociated, we can
1
 What is a Grammar?
learn a lot about how to write a grammar, and what to put in it,
by speculating on the nature of the grammatical knowledge of
speakers. We can profitably ask: what must a speaker-hearer know
in order to communicate in his language? If we observe linguistic
behaviour from a number of angles, we can begin to make observa-
tions which encourage us to predict certain necessary components of
grammatical knowledge. First, native speakers know that, of the
following three sentences, (I) is not a sentence of English, (2) is an
ungrammatical sentence of English, (3) is a grammatical sentence of
English:
(l) Quel est !'objet ala fois integral et concret de Ia linguistique?
(2) Three tons are weighed by this truck.
(3) This truck weighs three tons.
To go into more detail, they know more about ungrammatical
sentences; for example, that (4), (5), (6) and (7) are progressively
more deviant:
(4) This circle is square.
(5) John alarmed an apple.
(6) John alarmed a the.
(7) Alarmed a the John.
More relevantly, perhaps, they know an enormous amount about
grammatical sentences of English. For example, they know that
(8) and (9) are similar in meaning, as are (10), (11) and (12) and, in
a different way, (13) and (14):
(8) Her frankness astonished him.
(9) He was astonished by her frankness.
(1 0) The carpet was brown.
(11) The brown carpet ...
(12) The carpet which was brown ...
(13) He mounted his proud horse.
(14) He mounted his proud steed.
It goes without saying, of course, that speakers know which sentences
are different, as well as which ones are alike. That is, they can tell
sentences apart. This observation needs no illustration at this point,
since the book as a whole is a discourse upon it.
Another area of linguistic knowledge concerns ambiguous sen-
tences. Consider the following two examples:
(15) The chicken is·ready to eat.
(16) I saw her in the street.
(15) can be associated with either 'X eats the chicken' or 'the chicken
2
 What is a Grammar?
eats X'. (16) means either 'I saw her when I was in the street' or 'I saw
her when she was in the street'. A mature speaker of English knows
enough about the structure of (15) and (16) to retrieve either (or, as
alternatives, both) of the meanings for each of these sentences.
The linguist attempts to find a way of explaining these facts about
speaker-hearers' linguistic capacities. He has to account for the
structure of English sentences in a way which takes cognizance
of speakers' intuitions of deviance, similarity, distinctness and
ambiguity in their experience of English sentences. For instance, no
analysis of (15) is adequate unless it assigns two alternative structural
descriptions to that sentence, in recognition of the fact that speakers
attach two different meanings to it. In this case, the grammarian will
probably say that the chicken is the Object of the verb in one inter-
pretation ('X eats the chicken'), the Subject of the sentence in the
other ('The chicken eats X'). 'Subject' and 'Object' are descriptive
concepts which the linguist proposes as a way of explaining certain
structural facts about English. Notice that, while the motivation for
these concepts comes from an enquiry into 'what the speaker
knows' - here, the speaker's perception of ambiguity - they are no
more than theoretical terms, aids to expressing a hypothesis about
linguistic knowledge. It is not necessary to assume that English
speakers' brains contain two compartments labelled 'Subject' and
'Object'.
A linguist writes a grammar in an attempt to expose the structure
of the sentences of a language. His structural analysis is well-
motivated to the extent that he bears in mind that this set of sentences
relates to a shared linguistic competence in speakers of the language
under description. The problem 'What do speakers know?' has an
immense bearing on our more directly relevant question 'How shall
I present the structure of the sentences by which speakers com-
municate?'
Briefly, a language L is a set of sentences. The linguist must
account for all and only the grammatical sentetlC@S of L. ('L' is a
standard abbreviation for 'any natural language'.) This obligation
follows from my comments on sentences (1)-(3) above: the mature
speaker-hearer can distinguish between grammatical sentences of L,
ungrammatical sentences of L, and sentences which are not of L. If
the set described did not have limits, the grammar produced would
be utterly unprincipled: it would fail to divide off English from
French sentences, and, since it would omit to separate off un-
grammatical and grammatical sentences of L, it would be structurally
anarchic. I will assume that we have procedures for discounting
sentences which are not of Land sentences which are not grammatical
sentences of L. (Actually, these procedures are not yet properly
3
 What is a Grammar?
established, but the problems are too complex to be discussed here.)
If we can thus recognize grammatical sentences of L, we must go on
to ask 'How many of them are there?' The answer to this question is
known: the set L contains an infinite number of grammatical sen-
tences. Almost every sentence we hear, or produce, is new to us.
We are not normally conscious of the inventiveness of natural
language; do not realize that few of our sentences are exact repeti-
tions of already-used utterances. Of course, every society has a stock
of routine utterances like 'Good morning', 'Dear Sir', 'Thank you',
'No Smoking', 'I love you', 'Any other business?' and so on. These
utterances, which are frequently used, invariant, and tied to ritualized
communication situations, are quite untypical of normal linguistic
performance, which is diversified apparently without limit. One
might object that this observation is either unprovable, or, if prov-
able, irrelevant, since, because of human mortality, we cannot
actually experience an infinite set of sentences. However, we need
this assumption, because we must account for the creativity of
language - we are interested in the newness of sentences, even if we
cannot be concerned with their infiniteness. And there is, as it
happens, a demonstration of the notion 'infinite set of sentences'
which is not vulnerable to the embarrassing death of the grammarian
before he finishes counting sentences. What we can show is that
there is no longest sentence in a natural language, and therefore by
implication that there are an infinite number of sentences. (This is
not to say that there can be a sentence of infinite length, as has
sometimes been claimed, quite erroneously.) For every sentence of
the type (17), a longer sentence (18) is possible:
(17) John eats meat and vegetables.
(18) John eats meat, vegetables and fruit.
And for every sentence (18), a longer sentence can be constructed by
adding one more item. I will give two more examples of construc-
tions with this property; there are in fact several syntactic devices
available for extending sentences indefinitely:
(19) John believed that Mary claimed that Peter maintained that
Clive said that ...
(20) This hot, sunny, lazy, predictable ... climate suits me very well.
As the sentences of a language are infinite in number, the set which
the linguist must describe cannot be coextensive with any finite
corpus of sentences which, by observation and recording, he might
collect. There is a second reason why the task of writing a grammar
cannot be accomplished by merely cataloguing the structures found
in an observed corpus of sentences. The fact is that the actual utter-
4
 What is a Grammar?
ances of speakers do not adequately reflect speakers' real compe-
tence in L. Actual speech, as any unprejudiced observation will con-
firm, is riddled with grammatical mistakes of all kinds: incomplete
sentences, false concords, arbitrary changes of structure in mid-
sentence, illicit conjoinings of constituents which ought not to be
linked together - or at least not in the manner that they are - and
so on. (I am not appealing to 'prescriptive' standards. By 'ungram-
matical' here I don't mean structures which, in the manner of the
eighteenth-century purifiers or the edicts of the French Academy,
have been decreed to be unacceptable; but structures which native
speakers, if they could be reliably consulted, would agree are ill-
formed from the standpoint of their grammatical knowledge.) These
errors stem from various kinds of psychological and situational
'interference': distraction, lapses of memory, shifts of attention,
hesitation, etc. To describe such deviant sentences as these which
occur in a corpus would be to describe linguistically irrelevant
psychological factors as well as the linguistically relevant structural
knowledge of speakers.
Thus a corpus of utterances is not the true subject-matter of
linguistic description: it is only data-a set of observations from
which, with caution, the linguist must draw his grammatical state-
ments. In view of what has just been said, it is clear that the linguist's
use of his primary data must involve two adaptations. First, some
'idealization' is necessary so that the grammar does not take account
of the deviant sentences which occur in the corpus. Second, the
linguist must devise rules which project from his finite, observed
materials to an infinite set of sentences. That is to say, the grammar
must have predictive power.
All this adds up to the fact that a grammar is not a simple reflection
of linguistic usage. A few years ago, linguists used to be attacked,
for instance in the editorials of educational journals, for abandoning
all standards and saying that 'anything goes': in fact, linguists until
quite recently believed that any sentence which was produced ought
to be described by a grammar. But now a major reorientation has
taken place - it has been realized that speakers' actual linguistic
performance is not a very accurate indication of their underlying
linguistic competence. Many features of linguistic performance, many
aspects of texts and utterances, have to be discounted when writing
a grammar. At this stage, I might mention just one other character-
istic of discourse which a grammar does not seek to represent. It is
well known that sbme words, and some constructions, occur more
frequently than others: e.g. words like the and and are much more
frequent than discourse or dog; complex sentences more frequent
than simple sentences. Furthermore, the types of sentences which
5
 What is a Grammar?
occur in discourse correlate broadly with the circumstances in which
discourse is used- there are typical styles for advertising, informal
conversation, political rhetoric, scientific writing, etc. But as far as
the grammar is concerned, no one sentence, or type of sentence, is
more predictable than any other. Grammar does not take account of
probabilities. If a sentence occurs in a text or discourse, the grammar
will describe its structure; it will not exp1ain why that sentence
rather than some other was selected. The explanation of why sen-
tences occur in discourse is the task of stylis tics and sociolinguistics,
not of grammar.
A grammar which meets the requirements outlined above is called
a generative grammar. Such a grammar is predictive or projective
in the sense that, given a finite body of data (including, a collection
of observed sentences), it offers a system of rules so framed as to
account for an infinite set of potential sentences. In this way a
grammar 'generates' or 'enumerates' or 'describes' or 'defines' the
set of sentences which makes up the language. In an explicit and
formal manner, the grammar assigns at least one structural descrip-
tion to each sentence in the language (allowing that many sentences
are ambiguous and must therefore receive two or more structural
descriptions). We can test individual sentences - 'Is this sentence
generated by the grammar of English?' - by retracing a formal
derivation: by working through a series of rule-applications by
which the sentence is derived. (For the notion of 'derivation', see
below, pp. 45-7.) A generative grammar allows each structural
description to be associated unambiguously with one derivation.
Remember that a derivation is not an account of how a speaker
produces a sentence. As we will see when we have looked at some
derivations, such a proposal would be completely nonsensical. Early
critics of transformational-generative grammar believed, quite
mistakenly, that 'generate' meant 'produce' -that such a grammar
focused on the speaker's end of the communicative process. Actually,
a generative grammar is quite neutral with respect te speaker or
hearer: it makes no claims to explain how a sentence in actual
linguistic performance is either produced or comprehended.
One further clarification of terminology is necessary. A generative
grammar doe~ not have to be a transformational grammar. 'Trans-
formation' refers to a particular kind of rule, and a generative
grammar may or may not utilize transformational rules. In practice,
most modem generative grammars happen also to be transforma-
tional. But in principle a generative grammar without transforma-
tional rules could be written. We may note also that transformations
are not restricted to syntax: there are transformational rules in
phonology, also. Note that the present book is about transforma-
6
 What is a Grammar?
tional syntax, and this is not the same as transformational grammar
because grammar includes more than syntax.
I have said that a generative grammar 'assigns structural descrip-
tions to sentences' and that in this way the linguist accounts for their
structure in a manner which is consistent with what he can deduce
about speaker-hearers' linguistic knowledge. We must now ask
'What do structural descriptions [SDs] tell us about sentences?'
Given any one of the infinite set of sentences of L, all fully competent
speakers of that language will agree, within reasonable limits, on its
meaning. Equally, discounting peculiarities of accent and personal
voice quality, speakers agree on what it sounds like. To put it
another way, speakers are able to correctly associate a semantic
interpretation with a phonetic representation for each of an infinite
set of sentences of L. It would seem reasonable to expect a structural
description to reveal those qualities which speakers attribute to
sentences as they achieve sound-meaning associations. Let us con-
sider a simple sentence:
(21) The cat sat on the mat
is readily interpretable somewhat as follows: it concerns a cat (known
to be a certain kind of animal), particularized as one cat (rather than
as more than one) and as a specific cat rather than any old cat (the,
not a); identifiable behaviour (sitting) is attributed to the cat; a loca-
tion is specified; this location is identified as a particular kind of
inanimate object; the position of the cat relative to this object is
given ('on'); the whole semantic complex- cat-sitting-location-mat-
is set in past time. All this is roughly what the sentence means, to
any unprejudiced English speaker. He possesses conventions for
constructing this meaning, and he is also able to give these conven-
tions realization in sound or script. These conventions of meaning
and sound are community property: for every sentence, all speakers
in the community agree on the mechanisms by which meanings are
built up and associated with sounds.
Generative linguists, like traditional grammarians in general, deal
with these facts by setting up three interrelated levels of description:
a semantic level, a syntactic level and a phonological level. Alterna-
tively, we could say that a grammar has three 'components', calling
the components by their traditional names. (Note that 'grammar' is
often used as an equivalent to 'syntax'; but ·our usage of the term
'grammar' is, perhaps untraditionally, more inclusive.) The semantic
component is responsible, first, for assigning meanings to lexical
items: it must incorporate a dictionary. Like ordinary dictionaries,
this one must attempt to distinguish each lexical item from all
others, by stating exactly what senses mature speakers attribute to
BTs 7
 What is a Grammar?
each item in the language's vocabulary. It must also try to set out
the structure of the lexicon: the semantic relations (synonymy,
antonymy, superordination, etc.) which exist between lexical items.
A thesaurus aims to show these relationships, but conventional
dictionaries do not usually attempt to define such relationships
systematically. Second, the semantic component of a grammar
should account for the fact that the meanings of individual words
are, in sentences, amalgamated so that more complex meanings are
formed. Since these 'larger' meanings are built up under syntactic
constraints, the semantic component has to be arranged so that it can
make reference to appropriate syntactic properties of sentences. The
general design of the syntactic component of a grammar will be
indicated in some detail in the next chapter. To put it in rather
impressionistic terms, syntax distributes lexical items - and non-
lexical formatives - in patterns, patterns which are spread out
'left-to-right' in time or space. Syntax lays the basis for translating
an abstract meaning-complex into a piece of sequential behaviour.
It does so by generating a linear string of words arranged in a
regular pattern. This string constitutes the input to the phonological
component of the grammar. For every word, and every string of
words, there is an agreed realization in sound, a phonetic shape.
The phonological component specifies what phonetic contour is to
be attached to each of the infinite number of strings of words that
the semantic and syntactic parts of the grammar produce between
them. It is a set of instructions for pronunciation. Since many
languages use a written, as well as spoken, medium, there is also a
graphological equivalent to the phonological section of the grammar.
This book is about syntax. It will therefore have little to iay about
the details of phonological and semantic structure. But we must
remember that the three. components interlock, that none of them
functions independently of the others. I have already mentioned, for
example, that semantics must make reference to syntax to guide the
formation of sentence-meanings out of the sub-sentence elements
provided by the dictionary. Likewise, the phonological component
cannot work unless it has a very precise analysis of the syntactic
structure of the sentences for which it has to design a phonetic
representation. Syntax is very definitely not autonomous, and so
during the course of this book I will do my best to clarify the points
at which it makes contact with the other components.

Exercises and topics for discussion

1. Discuss the distinction between 'prescriptive' and 'descriptive'
grammar.
8
 What is a Grammar?
2. InCh. 1, some of the reasons why a grammar cannot be simply a
description of a finite corpus were given. Work out the arguments
against 'corpus-bound' grammar in more detail.
3. Investigate the notion of 'grammaticalness' in the writings of some-
modem transformational grammarians.
4. Make a critical review of Noam Chomsky's expositions of the
distinction between 'linguistic competence' and 'linguistic per-
formance'. You may wish to consider also Ferdinand de Saussure's
distinction between langue and parole.

9
 two

Deep and Surface Structure

As it happens, the most important relationship between syntax,

semantics and phonology can be presented straight away. The
linguistic levels of 'meaning' and 'sound' both have to be invoked to
help us define the central distinction in syntax - and with it the idea
of a transformational grammar itself. Consider the following
sentences:
(22) He took off his hat.
(23) He took his hat off.
These sentences have the same meaning; but they are different
arrangements of words. Since the difference between (22) and (23)
is immediately apparent at first glance, 'on the surface', as it were,
let us say that (22) and (23) exhibit different surface structures (or
superficial structures). To continue the metaphor, we will explain the
synonymy of (22) and (23) by saying that they have the same deep
structure (or underlying structure). Deep structure relates to meaning;
surface structure relates to order of elements, and hence to sound,
for in effect the surface structure determines the sequence of sounds
which occurs in a phonetic realization of a sentence. Surface struc-
ture is a dimension with physical associations, since it is the point at
which a sentence impinges on space and time. Deep structure, how-
ever, is an abstraction, a complex of meanings which is 'unpronounce-
able' unless it is rendered as a surface structure. Before we attempt
to say more about the theoretical status of deep and surface struc-
tures, let us look at some more examples.
(22) and (23) illustrate the situation in which one deep structure is
realized as two different surface structures. Another type of example
of this same relationship is (24), (25) and (26):
10
 Deep and Surface Structure
(24) The black cat sat on the mat.
(25) The cat, which was black, sat on the mat.
(26) The cat sat on the mat. The cat was black.
All three of the sentences - in the case of (26), pair of sentences -
express the same structure of meanings. In (24)-(26) a cat is said to
have been black and the same cat is said to have sat on a mat.
Although (25) is longer than (24), it adds no meaning; although (26)
splits the statement into two distinct sentences, it does not alter the
meaning-relationships within the statement - it would be an
extremely perverse interpretation to claim that two different cats are
mentioned in (26). So, as in (22) and (23), a set of utterances with
quite dissimilar surface structures is found to contain the same deep
structure.
The reverse situation may be found too: one surface structure may
hide the presence of two or more deep structures. I refer to the class
of syntactically ambiguous sentences, already represented above:
(15) The chicken is ready to eat.
(16) I saw her in the street.
We cannot pronounce (15) one way to show that the chicken is the
Object of eat, another way to emphasize that it is the Subject. In
(16), in the street goes with I in one interpretation, her in another-
the two interpretations sound just the same: one surface structure,
but undeniably two meanings. A favourite example of transforma-
tionalists is the following pair:
(27) John is easy to please.
(28) John is eager to please.
These two sentences appear to be identical except for the simple
lexical contrast easy/eager. But there is a fundamental difference: in
(27) John stands in an Object-verb relation to please; in (28) John is
in a Subject-verb relation to please. Compare It is easy to please
John with (27); but It is eager to please John is not in a meaningful
relation to (28). (Notice that the two different functions of ready
in the sentence about the chicken are represented separately in
eaJy and eager, neither of which is the syntactic chameleon that
ready is.)
Obviously, a grammarian who paid attention only to surface
structure would fail to notice some vitally important distinctions -
between (27) and (28), between the two meanings of (15) and
between the two meanings of (16). The familiar operation of 'parsing'
would miss the point entirely. Parsing entails assigning each word
in a sentence to a part of speech or word-class, and then representing
11
 Deep and Surface Structure
the sentence as a sequence of these word-classes. On this analysis,
(15), (27) and (28) would all come out as
Noun+ Copula+ Adjective+ Particle+ Verb
Or, bracketing together items which go together to form one unit,
(Noun+ (Copula+ (Adjective +(Particle +Verb))))
Alternatively, we might use notions like 'Subject' and 'Object'; but
that way John in (27) and till! chicken on one reading of (15) would
be erroneously labelled, since we would certainly be tempted to call
John and the chicken 'Subject' in all cases. Similarly, He in (9) would
be wrongly called 'Subject':
(9) He was astonished by her frankness.
But in (9), despite appearances, her frankness is the 'real' Subject, as
it is in (8), in which it is also the superficial Subject:
(8) Her frankness astonished him.
To return to our original example (22) versus (23), an analysis by
parsing would again misrepresent the syntactic facts. (22) would be
Pronoun+ Verb +Particle +Possessive Pronoun+ Noun
and (23):
Pronoun+ Verb +Possessive Pronoun+ Noun+ Particle
Two analyses, but one meaning. Parsing shows how (22) and (23)
differ, but not how they are the same. What is wrong with the
analysis of (23), of course, is that it obscures the fact that took and
off form, semantically, one unitary constituent (cf. remove) like put
away (store), run up (accumulate), put down (suppress) and many
other similar English verbs. In the case of (23), we clearly Deed two
different levels of representation. On one level, we must show that
certain words occur in a certain order-if we do not show this, we
cannot tell how (23) is to be spoken or written. On another level, we
must show certain meanings, and meaning-relations, which are
present in the sentence - this involves associating took and off
semantically, whereas on the other level they must be dissociated
positionally. The positional level of representation is a representa-
tion of (23) at the level of surface structure; the semantic level (which
can ignore positional facts if necessary) is a representation at the
level of deep structure. To complete our analysis, we must make sure
that the two levels of representation are related in some rational way,
and that the analysis of (23) is referable to whatever an~ysis is pro-
posed for (22): specifically, that the deep structure analysis of (23)
12
 Deep and Surface Structure
is the same as the deep structure analysis of (22), for, as we have
seen, they have the same meaning.
Before describing the technique which fulfils these analytic require-
ments, I will comment briefly on the problems which it has to solve
for the other sentences discussed in this chapter. For (24)-(26), we
must show that all the sentences express the same meaning, and we
must show how the surface structure differences among the members
of this trio arise: how from one underlying representation are
derived different word-orders, different words, and different lengths
and apparent complexities of sentences. And we must demonstrate
how all these changes occur without variation in the meaning of the
statement. With (15), (16) and (27), (28) we have a mirror-image of
the same problems. We must show how the same (or similar) surface
structure is derived from different deep structures: how do these
words, in these orderst stem from the given meanings? In addition,
we need some formal way of representing the underlying meanings -
as of course we do for (22)-(26); here, however, we need to refer
explicitly to the meanings of the sentences in order to express the
contrasts involved. Obviously, deep ,structure is determined in part
by the meanings of the lexical items which occur, so we need a
method of stating formally dictionary meanings. But as our examina-
tion of the sentences concerned has shown, certain syntactic functions
and relations also contribute to the establishment of meaning in
deep structure. The notions 'Subject' and 'Object', it appears, are
critical to distinguishing (27) and (28), and the two meanings of (15).
Evidently, any adequate theory of the nature of deep structure must
be prepared to give sense to these concepts and to identify them in
actual sentences. Moreover, it must distinguish deep structure
functions (e.g. Subject) from surface structure constituents which
appear to-have these same functions: for example, it must allow one
to decide that John is not tbe underlying Subject of (27), nor He that
of (9). (The conventional distinction between 'logical subject' and
'grammatical subject' is relevant to these cases: for 'logical subject'
read 'Subject in deep structure'; for 'grammatical', 'Subject in
surface structure'.) To give this criterion a more general and more
popular formulation, we might say that a theory of deep structure
should allow us to decide what aspects of overt syntactic ordering are
significant and what are insignificant or misleading. In (23) the
separation of took and off is a red herring, as far as meaning is con-
cerned. (22), in which took and off are adjacent, reflects the unity
of the verb much more directly. As a final example, a sufficient theory
of deep structure will tell us that black and cat in (24)-{26) are in a
semantically relevant relationship, despite the fact that in surface
structure they relate to each other in three different ways, and that
13
 Deep and Surface Structure
this diversity of physical arrangement is immaterial to the meaning.
By the same token, a sufficient theory of surface structure should
allow us to derive these various arrangements from the underlying
structure in an ordered fashion. Mter all, surface structure may be
opaque, but it is not haphazard: if it did not have principles of its
own, we would hardly be able to recover deep structure from it;
sentences would be uninterpretable.
Let us now look at a fresh example. In Latin, permutation of lexical
items in surface structure does not affect deep structure. All of the
following sentences mean 'Claudius loves the queen':
(29) Claudius reginam amat.
(30) Claudius amat reginam.
(31) Reginam Claudius amat.
(32) Reginam amat Claudius.
(33) Amat reginam Claudius.
(34) Amat Claudius reginam.
The agreement of meaning among these sentences is, of course,
marked by the suffixes attached to each of the words, which establish
Claudius as Subject and reginam as Object, regardless of the positions
which the words occupy in surface structure. Since the verb amat is
in the active rather than passive voice, the superficial Subject is
known to be identical to the Subject in deep structure. To account
for all the structural characteristics of these sentences, we clearly need
two kinds of rules:
(i) a set of rules which explains the syntactic relations and
semantic relations and content common to all the sentences;
(ii) a set of rules which explains the diversity of superficial
word-orders distinguishing the sentences.
Rules of type (i) are called by various authors phrase-structure rules,
constituent-structure rules, rewriting rules or branching rules. Type (ii)
rules are (one form of) transformational rules. Sentences (29)-(34)
dramatically illustrate the fact that one important property of
transformational rules ('T-rules•) is that they can rearrange con-
stituents without altering meaning.
Example (23) differs from (22) in that it has undergone a reposition-
ing transformation whiCh has not affected (22). We may represent
the process very informally, introducing the double arrow '=>'which
conventioruilly indicates that aT-rule is applied:
He took off his hat => He took his hat off
(deep structure) (surface structure)
This presentation symbolizes a fact that has been implicit in the
whole of my discussion so far: that T-rules are applied 'between'
14
 Deep and Surface Structure
(in some sense, not necessarily a temporal sense) deep and surface
structure. We must, however, disperse this metaphor 'between' by
discovering (a) how to represent deep structures as abstract objects;
(b) how to state T-rules as formal instructions; (c) how to represent
surface structures.
Both deep and surface structures can be represented as what we
call phrase-markers. Transformational rules apply to underlying
phrase-markers to give derived or superficial phrase-markers. (Where
more than one T-rule is applied, there may be intermediate phrase-
markers also.) In the informal notation of the previous paragraph,
these phrase-markers were shown as fully-fledged sentences, but this
was a mere convenience. Transformations do not derive sentences
from sentences, but P-markers from P-markers. Here is a more
appropriate notation (simplified for this preliminary discussion) for
the underlying phrase-marker of (23); the mode of representation is
called a tree-diagram or branching diagram:
s

~ NP PredP

~A
Det N Aux VP
(he) (Past) ~

V NP

MV
1\1\ Part Det N
(f41ee) (off) (his) (hat)

(This diagram shows the underlying phrase-marker for (22) as well

as (23), since the two sentences have the same meaning. I will return
to this point.) We can see in the diagram the structure of an under-
lying string obtained by listing left-to-right the items from the
lowest point of each branch:
(a) he -Past- take- off- his- hat
The T-rule accepts (a) as input and shifts the fourth item to the right-
most position, giving a derived string (b):
15
 Deep and Surface Structure
(b) he- Past- take- his- hat- off
So aT-rule is an instruction which says: 'Take a string of the form
(a) and perform such-and-such an operation on it, deriving a string
of the form (b).' String (a) may be either underlying or intermediate;
the only condition is that it must be associated with an explicit
phrase-marker, for in order to produce the derived string (b) the
transformational rule. must be able to refer to the structure of (a).
We know that the T-rule which derives ·P-marker (b) from P-
marker (a) -not, it must be stressed, 'sentence (23) from sentence
(22)' - accounts for many other sentences besides (23). Compare:
(35) The librarian put the book away.
(36) The truck knocked the old man down.
(37) I saw her off.
(38) I take the clock apart.
(39) Let's move the table over.
For economy's sake, we do not wish to reformulate a T-rule for
every particular underlying string to which it applies, so the input (a)
can be stated in such a way as to cover any number of pertinent
cases. We can achieve this economy by referring to branching-points
higher up in the tree which models the underlying phrase-marker.
Thus (at) will serve as input to the T-rule which is behind (23), (35),
(36) and (37):
(a1) NP - Past - MY - Part - NP
(38) shows that the T-rule applies whether the sentence is in the past
tense or not, so a more general representation still is possible:
(a2) NP - Aux- MY - Part - NP
Finally, it appears from (39) that from the point of view of this
transformation it is immaterial what precedes the main verb (MV);
so a very general rule can be given which embraces all the cases so
far examined:
X - MY - Part - NP => X - MY - NP - Part
(X = 'any or no formative(s)')
This is a rule of English called the particle-shift transformation.
This rule does not appear in the derivation of (22) He took off his
hat. The tree-diagram on p. 15 portrayed the underlying phrase-
marker for (22) as well as (23), since the sentences have the same
meaning - a common deep structure. Note that take and off are
adjacent in the diagram as well as in the surface structure of (22).
16
 Deep and Surface Structure
Perhaps the deep and surface structures of (22) are identical. Actually,
this is not the case. As we will see, there is no sentence in whose
derivation there are noT-rules. The transformations which bridge
underlying and superficial structures in (22) are rather too complex to
show at this early stage, but a simpler sentence will illustrate the point:
(40) I cleaned it.
At one level of generality, the underlying string is
NP- Aux - V- NP
Aux is Past, ultimately the affix (Af) -ed. In surface structure, -ed
follows clean. So we propose a general affix-shifting rule
NP-Af- V- NP=> NP- V -Af- NP
(alternatively, X - Af- V - Y => X - V - Af- Y, where X and Y
stand for anything occurring in those positions, as did X in the
particle-shift rule). This rule by definition applies also in the case
of (22), although, typically, its application is not obvious in surface
structure.
The claim that 'deep structure determines meaning' assumes that
all sentences display a distinction between underlying and super-
ficial syntax. It would be misleading and erroneous to interpret the
relation between (22) and (23) as demonstrating that (23) has both
· deep and surface structure whereas (22), because it does not display
the transformation placing his hat within took off, has only one
layer of structure; or, by a similar argument, that Her frankness
astonished him has no distinction between deep and surface structure
because it has undergone no transformation of the kind which
forces us to draw a line between deep and surface structure in the
case of He was astonished by her frankness. It would be wrong to
suggest that a language consists of a set of one-layered sentences such
as He took off his hat, Her frankness astonished him, plus a set of
two-layered sentences derived from them by transformation.
Distinguishing deep and surface structure is not a matter of mere
analytic convenience - ease of handling sentences like He took his
hat off, for example, through the medium of transformation. The
crucial motivation, which demands that the distinction apply to all
sentences, is that it provides a natural explanation for the fact that
sound and meaning are not directly related; that the phonetic surface
of linguistic symbols incorporates no intrinsic semantic properties.
The separation of underlying and superficial syntax recognizes this
fact by proposing that surface structure embodies all information
relevant to the phonetic representation of sentences, and deep struc-
ture contains all information necessary to the semantic interpretation
17
 Deep and Surface Structure
of sentences. Thus the syntactic (transformational) route between
deep and surface structure can be considered the means whereby
the 'pairing' of sounds and meanings is achieved.
Transformations are responsible for the arrangement of words
(or, better, morphemes, which are syntactic units smaller than the
word) in surface structure. My discussion of (23) and of the Latin
examples (29)-(34) may have suggested that the job transformations
do consists of repositioning or permuting elements. In fact, trans-
formational rules perform many different kinds of operation besides
permutation. They can delete elements: for instance, it is a trans-
formation which dispenses with the 'understood' you (in deep
structure) of imperatives. They can insert elements, for example con-
junctions and relative pronouns. They are responsible for combining
units in various ways - for example, the whole apparatus of 'sub-
ordination' familiar in traditional grammar is transformationally
managed: subordinate and main clauses are whole underlying P-
markers linked in different ways by transformational rules. Some-
thing of the range of types ofT-rule can be seen in examples (24)-
(26):
(24) The black cat sat on the mat.
(25) The cat, which was black, sat on the mat.
(26) The cat sat on the mat. The cat was black.
Each of these statements seems to be based on two propositions:
'cat - sitting (on the mat)' and 'cat - black'. In this analysis, (26)
appears to be the most direct expression of the underlying structure:
it represents in two separate derived strings the meaning-complexes
which are distinct in deep structure. We may therefore take (26) to
be more 'basic' than (24) or (25): for the sake of exposition - but
subject to the caution voiced above that sentences are not derived
from sentences - transformations may be regarded as operating on a
pair of strings something like (26) to derive (24) and (25). In (26)
the Noun Phrase the cat is repeated for each of the two statements
about the cat; in (24) and (25), where the two separate propositions
are united, one occurrence of the cat is deleted: it is o~ duty of
transformational rules to reduce redundancy in complex sentences
by removing needless repetitions of the same item. This deletion
happens as the second process in a sequential derivation which has
roughly the following milestones:
(l) The cat sat on the mat. The cat was black.
(2) The cat - the cat was black - sat on the mat.
(T-rule places the second string in the appropriate position
within the first.)
18
 Deep and Surface Structure
(3) The cat - was black - sat on the mat.
(One redundant occurrence of NP the cat deleted.)
(4) The cat - which was black - sat on the mat.
(Relative pronoun inserted in the right place.)
(5) The cat, which was black, sat on the mat.
('Commas' inserted: actually, this T-rule provides an
instruction for pronouncing the sentence with maintained
or rising intonation on cat and black, thus distinguishing
(25) from this, which is a different construction: (41) The
cat which was black sat on the mat, implying that some cat
which was not black did not sit on the mat. (25) is said to
contain a non-restrictive relative clause, (41) a restrictive
relative clause.)
The derivation for (24) demands further reduction and rearrange-
ment. We may assume that (24), like (25), goes through stage (2)
above - an assumption which has the added advantage of formally
connecting pre-posed adjectives and their corresponding relative
clauses. Next, everything except black in the 'embedded' string is
deleted:
(3a) The cat - black - sat on the mat.
Finally, the adjective is shifted to the position in front of the noun
which it modifies:
(4a) The black cat sat on the mat.
(In French, where the majority of adjectives follow nouns, this last
transformational stage is unnecessary; so in French the trans-
formational relationship between relative clauses and corresponding
adjectives is much more obvious in superficial syntax.)
In this chapter we have already seen quite a range of formally
distinct types of transformational rule. In later chapters, the range
will be increased even more. Because of this variety in the kinds of
operations performed by T-rules, it would not be very illuminating
to define a transformational grammar by simply listing types of
transformation. Instead, we may employ this more general definition:
a transformationalgrammar is a grammar which explicitly formalizes
the distinction between deep and surface structure in syntax. The
transformational section of the syntax receives deep structures as
its input and allows us to derive surface structures as output.

Exercises and topics for discussion

1. Collect some further examples of (a) pairs of sentences which
differ in surface structure but have the same underlying structure;
19
 Deep and Surface Structure
(b) sentences which have two or more deep structures underlying one
surface structure. Try to illustrate as wide a range of syntactic types
as you can.
2. Read the arguments in Chomsky's Syntactic Structures in favour
of a transformational level of analysis in syntax. He makes no
reference to the concepts of deep and surface structure; how does
this early exposition of his relate to the distinction in question?
3. It is a fundamental tenet of linguistics that sound and meaning
are not directly related. Explore as many branches of this argument
as you can.
4. A grammar should generate all of the sentences of L which are
grammatical, and none of the ungrammatical ones. Reconsider the
particle-shift T-rule in the light of the following sentences and of
other relevant ones which may occur to you:
(a) He took off his hat.
(b) He took his hat off.
(c) The truck knocked the old man down.
(d) The truck knocked down the old man.
(e) I saw her off.
(f) *I saw off her.
(g) Can you put me up?
(h) *Can you put up me? (Me is not strongly stressed.)
5. Examples (29)-{34) show that Latin enjoys very free privileges for
surface structure word-order; it is also a very richly inflected lan-
guage. Conversely, word-order in English, a language which now
has few inflexions, is quite rigidly fixed. Work out a general explana-
tion for these facts.

An asterisk * marks deviant forms here and throughout this book.

20
 three

Constituent Structure: Syntactic

Functions

I have not yet explained what I mean by a phrase-marker. It is, of

course, a kind of syntactic organization, one which can be pictured
in a tree-diagram. One example of a tree-diagram was given on p. 15
above. This diagram shows two features of organization as yet un-
explained: the tree 'branches' over and over again from top to
bottom; and at each of the points (called nodes) at which the
branches divide, and at each of the terminal points of the tree, there
is an abbreviated label - S, NP, N, etc. In this and the following
chapter I will show what is meant by the branching convention and
by the labelling convention. When we have agreed on the motivation
for, and method of, drawing tree-diagrams, we can begin to use them
as a way of representing phrase-markers. Specifically, they will serve
to display the structure of underlying phrase-markers.
The first point to grasp is that, although a sentence is a sequence of
formatives, these formatives .are not joined to another one-by-one
in sequence like the links of a chain: B related to A, C related to B
but not to A, D to C but not to B or A, and so on. In physical
reality - spoken or written linguistic performance - words and
morphemes follow one another one at a time in sequence, but this is
not a reliable guide to their syntactic relationships. Syntactically, a
sentence is not a simple linear concatenation of formatives like this:
(42) The +student +s +like +the +new +library.
The sentence is rather an ordered construct in which small units are
progressively built up into large on regular structural principles. It is
apparent that, in sentence (42), some sub-sequences of items can be
21
 Constituent Structure: Syntactic Functions
extracted which form syntactic wholes, while others would be
meaningless, arbitrary, segmentations. The students and like the new
library are well-formed constituents of the sentence, but students
like and like the would not be. Nor would s llke; s is obviously
intimately connected with student but not at all directly related to
like. We can show which elements 'go together' by bracketing them
together:
((The (students)) (like (the (new library))))
Each pair of brackets except the outer-most pair defines a con-
stituent: a sequence of formatives which is a regular part o~,)he
structure of the next largest constituent. Thus, new and library· are
constituents of new library; new library and the are constituents of
the new library; like and the new library are constituents of like the
new library. Student and s are constituents of students, which in turn,
with the_, makes up the students. Finally, the students and like the
new library are the constituents of the whole sentertce, which itself is
not, of course, a constituent of anything. Notice that some sequences
(e.g. like the) are not bracketed together and so are not constituents;
also, that any constituent is only a constituent of the 'larger' unit
which immediately includes it. So, for instance, new library is not an
immediate constituent of like the new library, although it is related
to it indirectly.
We have just analysed, crudely, the constituent structure of
sentence (42). If the presentation is thought to be difficult to read,
we can give the same information in an unlabelled tree-diagram:

(i)

new library
Each intersection (node) gathers together two simple or complex
constituents to form one more complex unit.
Now compare the above tree-diagram with the following, which
22
 Constituent Structure: Syntactic Functions
reproduces the linear analysis with plus-signs which was offered when
(42) was first introduced:

(ii)

The student s like the new library

The advantage of (i) over (ii) is that it correctly acknowl~dges that

syntactic structure is hierarchical, not (as shown in (ii)) linear. The
concept of hierarchical structure has already been sufficiently illus-
trated: by it is meant that each part of a sentence joins forces with
some unit of equal rank to make up a higher-order unit which
dominates both of them.
Constituent structure analysis which reveals hierarchical structure
brings immediate benefit to the differentiation of sentences. Linear
analysis makes it impossible to tell many types of sentence apart,
beyond the fact that different lexical items occur, in different orders,
and in sentences of different lengths. A constituent structure analysis
expressed in a branching diagram conveys at a glance structural
differences among sentences; compare the shapes of the following
trees:

(43)

life

enjoy ed

Crs 23
 Constituent Structure: Syntactic Functions

(44)

However, it has been repeatedly demonstrated that constituent

analysis is not wholly adequate to the task of describing a natural
language. In the above diagrams, (43) and (44) are distinguished as
superficial structures: that is, the diagrams (similarly, the one for
(42)) obscure the fact that some parts of the structure are derived
transformationally, some not. Admittedly, it is a great convenience,
and unobjectionable, to draw tree-diagrams of surface structures,
since they offer instant, systematic, displays of the overt structural
differences between sentences, and so are useful for comparative
purposes. But grammars represent linguistic facts in the form of
rules, not diagrams; and when we come to enquire what rule-
applications these diagrams of derived phrase-markers imply, we
find, quite naturally, that they imply a mixed bag of rules.
A constituent structure diagram must not be read as if it were
based entirely on phrase-structure rules. Such rules, the nature of
which we shall explore directly, are used to generate underlying
phrase-markers. Derived phrase-markers must be seen as a result of
the application ofT-rules after the underlying structure has been
built up by using these phrase-structure rules. The technique of tree-
diagramming can become misleading unless we realize that it does
not distinguish between the kinds of rules used to generate the
hierarchical structures which it represents. Thus it is vital to dis-
tinguish between the rules which introduce underlying hierarchical
structure and those which produce superficial constituent structure.
A phrase structure rule is always of the form

A-+ X+Y

This rule-type imputes a very literal meaning to the notion of con-

stituent structure: 'A consists of X followed by Y'. Thus, a unit A,
say an NP, has as its constituents X and Y ('Article' and 'Noun') in
the order XY. Such a rule is reflected in a simple branching:
24
 Constituent Structure: Syntactic Functions

NP

Art
(tk)
/\ Noun
(library}
The rule for this bit of constituent structure (to. be revised shortly)
is
NP--+ Art+N
which is a general rule of English: according to this rule, every sub-
sentence structure which is an NP has as its constituents an Article
followed by a Noun. And, of course, the configuration Art+ N is an
elementary component in deep structure meaning. But how are we
to account for the new library? It also seems to be an NP, since it is
freely substitutable for the simpler construction: the students like
the library/the new library. So we might analyse the new library in
this way:

/\
NP

AI\
(d.)

(w) (library)
However, this analysis complicates the system of phrase structure
rules, which now has to include three specifications for NP:

NP--+ {~:~)
Adj+N
(The braces enclose alternative constructions.) Now we have very
severely weakened the analytic concept 'NP', particularly by blurring
the distinction between adjectives and articles. Note that adjectives
and articles must be assigned different distributional privileges, since
we can have
25
 Constituent Structure: Syntactic Functions
(45) The library was new.
but not
(46) *The library was the.
In fact, example (45) shows us the way to identifying what is wrong
with the above analysis of the phrase from (42). Although new is an
immediate constituent of new librqry and hence an indirect con-
stituent of the new library, it is a constituent which has a trans-
formational relation to library, and is thus different in type from the
constituents the and library (of the library) which arc proper com-
ponents of deep structure related by PS-rules. Compare black in
the black cat discussed in the preceding chapter. Adjectives in pre-
nominal position - (24), (42) - are surface-structure arrangements
derived from adjectives in predicate position in deep structure - (26),
(45).
We can put it another way. Our observations suggest that an
adjective occurring before a noun has the same relationship to the
noun as an adjective linked to a noun by the 'verb to be' has to that
noun. (Compare (24) and the second sentence in (26).) To describe
these two arrangements by means of two separate rules would be to
obscure the fact that one relationship underlies these two word-
orders. In a transformational grammar we would wish to treat one
of these arrangements as basic, the other as derived. As we have just
seen, the inefficiency of the rule for NP which includes Adj + N
suggests that this is not the basic form. In fact, the truth of the
situation seems to be that all adjectives, wherever they occur in
surface structure, appear 'first' (i.e. in deep structure) in a P-markcr
of the form
s

NP

1\
Art N Adj
It follows that the relationship between N and Adj diagrammed by
the above tree is (or is an instance of) one of the basic stock of
syntactic relationships defined by that part of the grammar which
generates underlying structures. Furthermore, it must be one of the
syntactic relationships implicitly noted on p. 8 above when I men-
26
 Constituent Structure: Syntactic Functions
tioned that the semantic component of a grammar makes reference
to 'appropriate syntactic properties of sentences' in building up the
meaning-complexes of deep structure.
Let us now collect together the general observations that have been
made so far and see what they imply for the design of the syntactic
component of a grammar. The deep structure of a sentence is a set of
simple P-markers containing either just one member (e.g. (8), (9),
(21), (22), (23), (29)-(34),' (45)) or more than one member (e.g. (16),
(24), (25), (42), (43), (44)). Each of these underlying P-markers is
generated by phrase-structure rules; every one is of a certain deter-
minate constituent structure (there is no ambiguity in deep structure)
and exemplifies ceftain fundamental syntactic properties and rela-
tions. Each of them is drawn from a small finite set of underlying
P-markers generated by what we shall call the base of the syntactic
component of the grammar, the base consisting largely of a simple
phrase-structure grammar (PSG) which defines syntactic relation-
ships and word-classes. Underlying strings are provided with lexical
items collocated according to principles laid down by the semantic
component. Such lexically endowed underlying P-markers are
mapped on to surface structures by the transformational apparatus
of the syntactic component. Transformations thus operate on
semantically interpreted underlying P-markers to produce derived
or superficial P-markers. These latter are surface strings of forma-
tives which, like the underlying string(s), have constituent structure.
But there are two defining differences from the structure of under-
lying strings: superficial strings may have apparently indeterminate
structure, that is to say they may be ambiguous; and their constituent
structure is accounted for by their transformational history as well as
by PS-rules. Surface ambiguity implies the existence of two or more
alternative transformational routes which lead 'backward' from one
and the same surface structure to different deep structures.
To return now to the syntactic relationships which determine the
structure of underlying P-markers. It will have been noticed that
the immediate constituent analyses presented thus far have usually
been based on a binary segmentation:
A

nor

B c B c D B
This division of one constituent into two - neither more nor less -
constituents has applied whether the object of analysis has been an
27
 Constituent Structure: Syntactic Functions
underlying or a superficial string. The decision to employ a principle
of bipartite segmentation is neither accidental nor arbitrary. The
first cut in the analysis of sentences is binary because every sentence
is fundamentally a relation of two elements, a Subject and a Predicate.
The analysis (47)(a) recognizes this fact; (47)(b) disguises it:
(47) (a) The children I ate the apples noisily.
{b) The children I ate I the apples noisily.
Tentatively, we can characterize the function of predication in this
way. Every sentence focuses attention, interest, enquiry, or what-
ever, on some 'topic' (object, event, process, concept, etc.), which is
'expressed as the Subject; the remainder of the sentence provides a
'comment' on this topic. In sentence (47) the Subject is 'the children';
what is predicated of the topic expressed in this Subject is that the
children 'ate the apples noisily'. In so far as ate the apples noisily
relates to the children, it must be considered as a unitary segment
constituting as a whole the Predicate of the sentence. In the under-
lying 'message' of (47), the whole complex of eating, eating apples,
eating apples noisily, forms one block of meaning which the syntax,
through the function of predication, attributes to the children.
In (47) the Predicate is quite complex: it consists of a verb followed
by an NP followed by an adverb. This particular complexity is not
essential to predication; simpler orderings of meaning may have this
function. To begin with, the adverb (noisily, in this case) is entirely
optional - as grammarians have long noticed. We can just as easily
express predication without an adverb:
(48) The children ate the apples.
{Transformationalists are uncertain about the status' of adverbs,
beyond knowing that, unlike the other kinds of element considered
in this chapter, they are optional. Perhaps adverbs do not appear in
deep structure, but are introduced by a range of transformations - a
possibility to which I will return in the next chapter.) Simpler
structures still - lacking the second NP - will also express the
function of predication:
(49) The children slept.
If we compare (49) with (47), (48) we notice a difference between two
types of Predicate, that containing an Object and that not containing
an Object. Both Subject and Object are NPs;· but they are readily
distinguished in diagrams of phrase-markers in terms of the
positions they occupy in trees. The distinction is obvious from the
following diagram even if I do not explain the meanings of all the
labels on nodes:
28
 Constituent Structure: Syntactic Functions
s

1\
1\
NP PrcdP

Aux VP

1\ V NP

The Subject is the NP on the left, below a branch dominated by the

node S; the Object is on the right, immediately dominated by the
node VP. Notice that, if trees are properly drawn, such syntactic
functions can be deduced from the shape of the tree: it is not neces-
sary to include the terms 'Subject' and 'Object' themselves in the
tree. However, one secondary test is required for ~Object', since
predication may also be expressed in sentences like the following -
(50) This truck weighs three tons.
(51) John became a politician.
- in which the right-most NPs are not Objects. The secondary test
entails submitting sentences to the passive transformation:
(52) The apples were eaten by the children.
(53) *Three tons are weighed by this truck.
(54) *A politician was become by John.
So the Object of a verb (within the Predicate) is an NP which is
immediately dominated by VP and which, if the passive transforma-
tion can be applied, becomes the superficial (or 'grammatical')
Subject of the derived sentence. NPs which do not fulfil both of these
conditions are not Objects.
Two more types of Predicate are encountered, illustrated by the
following:
(55) John was a politician.
(56) John was clever.
To make these Predicates consistent with those in sentences (47)-(51),
we might say that they are Predicates by virtue of containing the
29
 Constituent Structure: Syntactic Functions
'verb to be'. But this interpretation does not obviously reflect the true
situation. It would be quite reasonable to claim that predication in
(55) and (56) is expressed by the NP a politician and the adjective
clever, respectively. Was adds nothing to the meaning: it is semanti-
cally empty. Its function is to carry tense-information which, of
course, cannot be affixed directly to a politician or clever, since they
are not verbs. The underlying structure of (55) and (56) is roughly
as follows:
NP- Past- NP
NP - Past - Ad]
Because Past is in these circumstances, 'unaffixable', an obligatory
transformation is applied inserting be for Past to be attached to.
Other transformations like this one are found elsewhere in the
grammar: in particular, we may note the rule introducing do in
negatives and questions, constructions which separate Past from the
verb and so create a need for a surface structure morpheme to
attach it to:
(57) Did you like the soup?
Past-NP- V - NP
(58) You did n't like the soup.
NP - Past - Neg - V - NP
In summary, the Predicate of a sentence need not contain a verbal
element: its exponents may be a verb (49); a verb phrase containing
an Object (47, 48); a verb phrase containing an NP which is not an
Object (50, 51); an NP by itself (55); an adjective. These alternatives
may be captured in the following rules, in which the single arrow
'->-' may for the moment be regarded as meaning 'consists of':

(i) S -+ NP + PredP
(ii) PredP -+ Aux + {~p }
Adj
NP
(iii) NP -+ Det + N
(iv) VP-+ V +{NP}
Adj
Rule (i) states that every underlying phrase-marker consists of a
Noun Phrase followed by a Predicate Phrase. The Noun Phrase is the
Subject, a fact which does not need to be stated explicitly in the
rules because (a) the Subject must be an NP; and (b) the NP immedi-
30
 Constituent Structure: Syntactic Functions
ately dominated by S must be the Subject. (So no NP introduced by
(ii) or (iv) could be the Subject.)
(ii) divides PredP into an auxiliary and one of four other consti-
tuents. For the moment, we may regard Aux as comprising Tense
([+Past] or [-Past]) or Tense plus a 'modal auxiliary' such as can,
will, etc. Aux is separated out in this way because it applies to the
whole of the PredP, not just to the verb (if any): for instance, in (47)
the whole Predicate ate the apples noisily is characterized as past
tense, even though it is only the verb itself which is actually marked
[+Past] morphologically.
Reading from top to bottom, the bracketed choices in rule (ii)
allow for (49), (48), (56) and (55).
Rule (iii) breaks down NP into a Determiner plus a Noun. Read
in conjunction with the bottom line of rule (ii), this gives us (55)
John was a politician, in which there are two instances of NP. Note
that every NP - even a single-word NP such as John - has a Deter-
miner; but some Dets are not expressed in surface structure (see
Ch. 6).
(iv) expands VP, giving two more types of Predicate: V + NP as
in (48) and V +Adj as in
(59) John became rich.
Our simple PSG generates just five underlying phrase-markers, as
follows:
s

1\ NP PredP

1\1\
Det N Aux V

(E.g. The children slept by rules (i), (ii) first line, and (iii).)

31
 Constituent Structure: Syntactic Functions

1\
1\1\
NP PrcdP

1\
Det N Aux VP

V NP

(E.g. The children ate the apples by rules (i), (ii) second line, (iii)
twice, (iv) first line; or, by the same rules, This truck weighs three tons
where the VP-dominated NP is not the Object.)

1\
NP PrcdP

1\1\
Det N Aux VP

1\ V Adj

(E.g. John became rich by rules (i), (ii) second line, (iii), (iv) second
line.)

32
Exploring the Variety of Random
Documents with Different Content
 “Hullo, hullo, hullo!
It’s a different girl again!
Different hair, different clothes,
Different eyes, different nose....”

This affair had been kept a dead secret from press and public. It was
a “glorious stunt” which had for its amiable object the introduction of
all the prettiest girls of the theater world to all the smartest bloods
of the universities and clubs. It was entitled the Butterfly Ball.
Certainly there were some astoundingly beautiful girls at this
assembly, and not a few of them. The university boys were, for a
time abashed by so much loveliness. But they brightened up,
especially when the most famous sporting peer of England—Lord
Lonsdale—led off the dance with a little girl dressed, rather
naughtily, as a teetotum. By the time I left—a kind of Pierrot looking
on at the gayety of life—there was a terrific battle in progress
between groups of boys and girls, with little white rolls of bread as
their ammunition. Not commendable. Not strictly virtuous, nor highly
proper, but in its wildness there was the spirit of a youth which,
afterward, was heroic in self-sacrifice.... So things happened in
London before the war.
A series of articles appearing in The Daily Mail, by Robert Blatchford,
once a Socialist and still on the democratic side of political life,
disturbed the sense of security in the average mind by a slight
uneasiness. Not more than that, because the average mind had its
inherited faith in our island inviolability and the power of the British
Navy. There were articles entitled “Am Tag,” which is bad German,
and they professed to reveal a determination in the military and
naval castes of Germany to destroy the British fleet, invade England,
and smash the British Empire.
Some of the evidence brought forward seemed childish in its
absurdity. There were not many facts to a wealth of rhetoric. But
they created a newspaper sensation, and were pooh-poohed by the
government, as we now know, with utter insincerity—for there were
members of that government who knew far more than Blatchford
how deep and widespread was German hostility to Great Britain, and
how close Europe stood to a world war.
One fantastic little incident connected with those articles of
Blatchford’s amused me considerably at the time, though afterward I
thought of it as a strange prophecy.
I called on W. T. Stead one day in his office of The Review of
Reviews, which afterward I was to edit for a year. It was just before
lunch time, and Stead had an engagement with Spender of The
Westminster Gazette. But he grabbed me by the arm, in his genial
way, and said, “Listen to this for a minute, and tell me what you
think of it.”
It appeared that he had been rather upset by Blatchford’s articles.
He could not make up his mind whether they were all nonsense or
had some truth at the back of them. He decided to consult the spirit
world through “Julia,” his medium.
“We rang up old Bismarck, Von Moltke, and William II of Prussia.
‘Look here,’ I said, ‘Is there going to be war between Germany and
England?’”
The spirits of these distinguished Germans seemed uncertain.
Bismarck saw a red mist approaching the coast of England. Von
Moltke said the British fleet had better keep within certain degrees
of latitude and longitude—which was kind of him! One of the trio—I
forget which—said there would be war between Germany and
England. It would break out suddenly, without warning.
“When?” asked W. T. Stead.
A date was given. It was the month of August. The year was not
named.
I laughed heartily at Stead’s anecdote, especially when he told me
the effect this announcement had upon him. He was so disturbed
that he went round to the Admiralty, interviewed Lord Fisher, who
was a friend of his, and revealed the dread message that the
German fleet was going to attack in August. (It was then May,
1912).
Fisher leaned back in his chair, smiled grimly, and said, “No such
luck, my boy!”
In August of that year I was engaged in trouble which did not seem
connected with Germany, though I am inclined to think now that
German agents were watching it very closely—especially one
German baron who posed as a journalist and was always reporting
on industrial unrest in Great Britain, wherever it happened to break
out. I had met him at Tonypandy, in Wales, during the miners’ riots
down there, and I met him again in Liverpool, which was now in the
throes of a serious strike.
It was the nearest thing to civil war I have seen in any English city. I
have forgotten the origin of the strike—I think it began with the
dockers—but it spread until the whole of the transport service was at
a standstill, and the very scavengers left their work. The Mersey was
crowded for weeks with shipping from all the ports of the world,
laden with merchandise, some of it perishable, which no hands
would touch. No porters worked in the railway goods yards, so that
trains could not be unloaded. There was no fresh meat, and no milk
for babes. Not a wheel turned in Liverpool. It was like a besieged
city, and presently, in hot weather, began to stink in a pestilential
way, because of the refuse and muck left rotting in the streets and
squares.
This refuse, among which dead rats lay, was so filthy in one of the
best squares of Liverpool outside the hotel where I was staying, that
a number of journalists, and myself, borrowed brooms, sallied out,
swept up the rubbish heaps, and made bonfires of them, surrounded
by a crowd of angry men who called us “scabs” and “blacklegs,” and
threatened to “bash” us, if we did not stop work. We stuck to our
job, and were rewarded by a clapping of hands from ladies and
maidservants in the neighboring windows, so that our broomsticks
seemed as heroic as the lances of chivalry.
Some bad things happened in Liverpool. The troops were stoned by
mobs of men who were becoming sullen and savage. Shops were
looted. I saw no less than forty tramcars overturned and smashed
one afternoon in that sunny August, because they were being driven
by men who had refused to strike.
On that afternoon I saw something of mob violence, which I should
have thought incredible in England. A tramcar was going at a rapid
pace, driven by a man who was in terror of his life because of a mob
on each side of the road, threatening to stone him to death. Inside
the car were three women and a baby. A fusillade of stones suddenly
broke every window. Two of the women crouched below the window
frames, and the third woman, with the baby, utterly terrified, came
on to the platform outside, and prepared to jump. A stone struck her
on the head, and she dropped the baby into the roadway, where it
lay quite still. A gust of hoarse laughter rose from the mob, and not
one man stirred to pick up the baby. Terrible, but true. It was left
there until a woman ran out of a shop.... Wedged behind the men,
but a witness of all that happened, I was conscious then of a cruelty
lurking in the vicious elements of our great cities which, before, I
had not believed to exist in England of the twentieth century. If ever
there were revolution in England, it would not be made with rose
water.
The troops and police were patient and splendid in their discipline,
despite great provocation at times. Now and again, when the mob
started looting or stone throwing, the police made baton charges,
which scattered crowds of young hooligans like chaff before them,
and they thrashed those they caught without mercy. At such times I
had to run like a hare, for there is no discrimination in treatment of
the innocent.
One afternoon the troops were ordered to fire on a crowd which
made an attempt to attack an escort of prisoners, and there was a
small number of casualties. That night I had an exciting narrative to
dictate over the telephone to the office of The Daily Chronicle. But,
in the middle of it, the sub-editor, MacKenna, who was taking down
my message, said, “Cut it short, old man! Something is happening
to-night more important than a strike in Liverpool. The German fleet
is out in the North Sea, and the British fleet is cleared for action!”
When I put down the telephone receiver, I felt a shiver go down my
spine; and I thought of Stead’s preposterous story of war in August.
Had it happened?
There was nothing in next day’s papers. Some iron censorship closed
down on that story of the German fleet, true or false.... As we now
know, it was true. The German fleet did go out on that night in
August, but finding the British fleet prepared, they went back again.
It was in August of another year that Germany put all to the great
hazard.
The thoughts of the English people were not obsessed with the
German menace. For the most part they knew nothing about it,
apart from newspaper “scares,” which they pooh-poohed, and no
member of the government, getting anxious now in secret
conversations, took upon himself the duty of preparing the nation for
a dreadful ordeal.
England was excited by two subjects of sensational interest and
increasing passion—the mania of the militant suffragettes, and the
raising of armed forces in Ireland, under the leadership of Sir
Edward Carson, to resist Home Rule.
I saw a good deal of both those phases of political strife in England
and Ireland. The suffragette movement kept me in a continual state
of mental exasperation, owing to the excesses of the militant women
on one side, and the stupidity and brutality of the opponents of
women’s suffrage on the other. I became a convinced supporter of
“Votes for Women,” partly because of theoretical justice which
denied votes to women of intellect, education, and noble work, while
giving it to the lowest, most ignorant, and most brutal ruffians in the
country, partly because of a sporting admiration—in spite of
intellectual disapproval—of cultured women who went willingly to
prison for their faith, defied the police with all their muscular
strength, risked the brutality of angry mobs (which was a great risk),
and all with a gay, laughing courage which mocked at the
arguments, anger, and ridicule of the average man.
Many of the methods of the “militants” were outrageous, and
loosened, I think, some of the decent restraints of the social code,
for which we had to pay later in a kind of sexual wildness of modern
young women. But they were taunted into “direct action” by Cabinet
Ministers, and exasperated by the deliberate falsity and betrayal of
members of Parliament, who had pledged themselves at election
time to support the demand of women for the suffrage, by
constitutional methods.
A number of times I watched the endeavors of the “militants” to
present a petition to the Prime Minister or invade the Houses of
Parliament. Always it was the same scene. The deputation would
march from the Caxton Hall through a narrow lane in the midst of a
vast crowd, and then be scattered in a rough and tumble scrimmage
when mounted police rode among them.
Often I saw a friend of mine walking by the side of these
deputations, as a solitary bodyguard. It was H. W. Nevinson, the war
correspondent, with his fine ruddy face and silvered hair, a paladin of
woman suffrage as of all causes which took “liberty” for their
watchword. The crowd was less patient of men sympathizers of
militant women than with the women themselves, and Nevinson was
roughly handled. At a great demonstration at the Albert Hall, he
fought single-handed against a dozen men stewards who fell upon
him, when he knocked down a man who had struck a woman a
heavy blow. Nevinson, though over fifty at the time, could give a
good account of himself, and some of those stewards had a tough
time before they overpowered him and flung him out.
Round the Houses of Parliament, on those nights of attack, there
were strong bodies of police who played games of catch-as-catch-
can with little old ladies, frail young women, strong-armed and lithe-
limbed girls who tried to break through their cordon. One little old
cripple lady used to charge the police in a wheel chair. Others caught
hold of the policemen’s whistle chains, and would not let go until
they were escorted to the nearest police station. One night dozens
of women chained and padlocked themselves to the railings of the
House of Commons, and the police had to use axes to break their
chains.
There was a truly frightful scene, which made me shiver, one night,
when those “militants” refused to budge before the mounted police
and seized hold of their bridles and stirrup-leathers. The horses,
scared out of their wits by these clinging creatures, reared, and fell,
but nothing would release the grip of those determined and reckless
ladies, though some of them were bruised and bleeding.
The patience and good humor of the police were marvelous, but I
was sorry to see that they made class distinctions in their behavior.
They were certainly very brutal to a party of factory girls brought
down from the North of England. I saw them driven into a narrow
alley behind Westminster Hospital, and the police pulled their hair
down, wound it round their throats, and flung them about
unmercifully. It was not good to see.
I had several talks at the time with the two dominant leaders of the
militant section, Mrs. Pankhurst and her daughter Christabel, and I
was present at their trial, when they were indicted for conspiracy to
incite a riot. Mrs. Pankhurst’s defense was one of the most
remarkable speeches I have ever heard in a court of law, most
eloquent, most moving, most emotional. Even the magistrate was
moved to tears, but that did not prevent him from setting aside an
unrepealed statute of Charles II (which allowed a deputation of not
more than thirteen to present a petition, without let or hindrance, to
the King’s ministers) and sentencing Mrs. Pankhurst and her
daughter to two years’ imprisonment.
I saw Christabel Pankhurst during the course of the trial, and she
asked me whether I thought she would be condemned. I told her
“Yes,” believing that she had the strength to hear the truth, and
afterward, when she asked me how much I thought she would get, I
said “Two years.” I had an idea from her previous record that she
was ready for martyrdom at any cost, but to my surprise and
dismay, she burst into tears. Her defense and cross-examination of
witnesses were also marred by continual tears, so that it was painful
to listen to her. Her spirit seemed quite broken, and she never took
part again in any militant demonstrations, although she was
liberated a short time after the beginning of her imprisonment. She
worked quietly at propaganda in Paris.
One nation watched the mania of the “wild, wild women” with a
growing belief in England’s decadence, as it was watching the Irish
affairs, and industrial unrest. German agents found plenty to write
home about.
 XVII
One day in 1913, I was asked by Robert Donald to call on a
Canadian professor who had been engaged in “a statistical survey of
Europe,” whatever that may mean, and might have some interesting
information to give.
When he received me, I found him a little, mild-eyed man, with
gold-rimmed spectacles, behind which I presently discovered the
look of one obsessed by a knowledge of some terrific secret. That
was after he had surprised me by declining to talk about statistics,
and asking abruptly whether I was an honest young man and a good
patriot. Upon my assuring him that I was regarded as respectable by
my friends and was no traitor, he bade me shut the door and listen
to something which he believed it to be his duty to tell, for England’s
sake.
What he told me was decidedly alarming. In pursuit of his “statistical
survey of Europe” on behalf of the Canadian and American
governments, he had spent two years or so in Germany. He had
been received in a courteous way by German professors, civil
servants, and government officials, at whose dinner tables he had
met German celebrities, and high officers of the German army. They
had talked freely before him after some time, and there was
revealed to him, among all these people, a bitter, instinctive,
relentless, and jealous hatred of England. They made no secret that
the dominant thought in their souls was the necessity and
inevitability of a conflict with Great Britain, in order to destroy the
nation which stood athwart their own destiny as their greatest
commercial competitor, and as the one rival of their own sea power,
upon which the future of Germany was based. For that conflict they
were preparing the mind of their own people by intensive
propaganda and “speeding up” the output of their naval and military
armament. “England,” said my little informant, “is menaced by the
most fearful struggle in history, but seems utterly ignorant of this
peril, which is coming close. Is there no one to warn her people, no
one to open their eyes to this ghastly hatred across the North Sea,
preparing stealthily for their destruction? Will you not tell the truth in
your paper, as I now tell it to you?”
I told him it would be difficult to get such things published, and still
more difficult to get them believed. I had considerable doubt myself
whether he had not exaggerated the intensity of hatred in Germany,
and, in any case, the possibility of their daring to challenge Great
Britain, as long as our fleet maintained its strength and traditions.
But I was disturbed. The little man’s words coincided with other
warnings I had heard, from Lord Roberts, from visitors to Germany,
from Robert Blatchford—to say nothing of W. T. Stead and his
German “spooks.” ... Robert Donald, of The Daily Chronicle, laughed
at my report of the conversation. “Utter rubbish!” was his opinion,
and he refused to print a word.
“Go to Germany yourself,” he said, “and write a series of articles
likely to promote friendship between our two peoples and undo the
harm created by newspaper hate-doctors and jingoes. Find out what
the mass of the German people think about this liar talk.”
So I went to Germany, with a number of introductions to prominent
people and friends of England.
It was not the first time I had visited Germany, because the previous
year, I think, I had been to Hamburg with a party of journalists, and
we were received like princes, fêted sumptuously, and treated with
an amazing display of public cordiality. There was private courtesy,
too, most kind and amiable, and I always remember a young poet
who took me to his house and introduced me to his beautiful young
wife who, when I said good-by, gathered some roses from her
garden, put them to her lips, and said, “Take these with my love to
England.”
But something had happened in the spirit of Germany since that
time. The first “friend of England” to whom I presented a letter of
introduction was a newspaper editor in Düsseldorf, a man of liberal
principles who had taken a great part in arranging an exchange of
visits between German and British business men. He knew many of
the Liberal politicians in England and could walk into the House of
Commons more easily than I could.
He seemed to be rather flustered when I called upon him and
explained the object of my visit, and he left me alone in his study for
a while, on pretext of speaking to his wife. I think he wanted me to
read his leading article, signed at the foot of the column, in a paper
which he laid deliberately on his desk before me. I puzzled through
its complicated argument in involved German, and through its fog of
rhetoric there emerged a violent tirade against England.
When he came back, I tackled him on the subject.
“I understood that you were an advocate of friendly relations
between our two peoples? That article doesn’t seem to me very
friendly or helpful.”
He flushed a hot color, and said, “My views have undergone a
change. England has behaved abominably.”
The particular abomination which he resented most deeply was the
warning delivered by Lloyd George—of all people in England!—that
Great Britain would support French interests in Morocco, and would
not tolerate German aggression in that region. That was at the time
of the Agadir incident. The British attitude in that affair, said the
Düsseldorf editor, was a clear sign that Great Britain challenged the
right of Germany to develop and expand. That challenge could not
be left unanswered. Either Germany must surrender her liberty and
deny her imperial destiny, at the dictation of Britain, or show that
her power was equal to her aspirations. That, anyhow, was the line
of his argument, which we pursued at great length over pots of lager
beer, in a restaurant where we dined together.
I encountered the same argument, and more violent hostility, from a
high ecclesiastic in Berlin, who was a great friend of the Kaiser’s and
formerly a professed lover of England. He was a tall, thin, handsome
man, who spoke English perfectly, but was not very civil to me.
Presently, as we talked of the relations between our two nations, he
paced up and down the room with evident emotion, with suppressed
rage, indeed, which broke at last through his restraint.
“English policy,” he said, “cuts directly across our legitimate German
rights. England is trying to hem in Germany, to hamper her at every
turn, to humiliate her in every part of the world, and to prevent her
economic development. During recent days she has not hesitated to
affront us very deeply and deliberately. It is intolerable!”
He spoke of an “inevitable war” with startling candor, and when I
said something about the duty of all Christian men, especially of a
priest like himself, to prevent such an unbelievable horror, he asked
harshly whether I had come to insult him, and touched the bell for
my dismissal.
Such conversations were alarming. Yet I did not believe that they
represented the general opinion of the great mass of German
people. I was only able to get glimpses here and there in Düsseldorf
and Frankfort, Hanover, Leipzig, Berlin, and Dresden of middle-class
and working-class thought, but wherever I was able to test it in
casual conversation with business men, railway porters, laborers,
hotel waiters, and so on, with whom I exchanged ideas in my very
crude German, or their remarkably good English (in the case of
commercial men and waiters), I found utter incredulity regarding the
possibility of war between England and Germany, and a contempt of
the sword-rattling and “shining armor” of the Kaiser and the military
caste.
I was, for instance, in a company of commercial men at Abendessen
in a hotel at Leipzig, when the topic of conversation was the Zabern
affair, in which Lieutenant von Förstner had drawn his sword upon
civilians—and a cripple—who had jeered at him for swaggering down
the sidewalk like a popinjay. The Crown Prince had sent him a
telegram of approbation for his defense of his uniform and caste.
But, one and all, the commercial men with whom I sat expressed
their loathing of this military arrogance, and were indignant with
those who defended its absurdity. I remember the German who sat
next to me had been a designer in a porcelain factory in the English
potteries for many years. With him I talked quietly of the chance of
war between England and Germany. “What is the real feeling of the
ordinary folk in Germany?” I asked. He answered with what I am
certain was absolute sincerity—though he was wrong, as history
proved. He told me that, outside the military caste, there was no war
feeling in Germany, and that the idea of a conflict with England was
abhorrent and unbelievable to the German people. “If there were to
be war with England,” he said, “we should weep at the greatest
tragedy that could befall mankind.”
There were many people I met who held that view, without
hypocrisy, and their sincerity at that time is not disproved because
when the tocsin of war was sounded, the fever of hate took
possession of them.
It was Edward Bernstein, the leader of the Socialists, who warned
me of the instability of the pacifist faith professed by German
democrats. “If war breaks out,” he said, “German Socialists will
march as one man against any enemy of the Fatherland. Although
theoretically they are against war, neither they nor any other
Socialists have reached a plane of development which would give
them the strength to resist loyalty to the Flag and the old code of
patriotism, when once their nation was involved, right or wrong.”
I tried to get the ideas of German youth on the subject of war with
England, and I had an excellent opportunity and an illuminating
conversation with the students of Leipzig University. A group of
these young men, who spoke excellent English, allowed me to
question them, and were highly amused and interested.
“Do you hate England?” I asked.
There was a rousing chorus of “Yes!”
“Why do you hate England?”
One young man acted as spokesman for the others, who signified
their assent from time to time. The first reason for hatred of
England, he said, was because when a German boy was shown the
map of the world and when he asked what all the red “splodges” on
it signified, he was told that all that territory belonged to England.
That aroused his natural envy. Later in life, said this young man, he
understood by historical reading that England had built up the British
Empire by a series of wars, explorations, and commercial adventures
which gave her a just claim to possession. They had no quarrel with
that. They recognized the strength and greatness of the English
people in the past. But now they saw that England was no longer
great. She was decadent and inefficient. Her day was done. They
hated her now as a worn-out old monster who still tried to grab and
hold, and prevent other races from developing their genius, but had
no military power with which to defend their possessions. England
was playing a game of bluff. Germany, conscious of her newborn
greatness, her immense industrial genius, her vital strength, needing
elbow room and free spaces of the earth, would not allow a
degenerate people to stand across her path. Germany hated England
for her arrogance, masking weakness, and her hypocritical
professions of friendship, which concealed envy and fear.
All this was said, at greater length, with admirable good humor and
no touch of personal discourtesy. But it made me thoughtful and
uneasy. The boy was doubtless exaggerating a point of view, but if
such talk were taking place in German universities, it boded no good
for the peace of the world.
I returned to England, perplexed, and not convinced, one way or the
other. As far as I could read the riddle of Germany, public opinion
was divided by two opposing views. The military caste, the old
Junker crowd, and their satellites, ecclesiastical and official, with,
probably the Civil Service, were beating up the spirit of aggression,
and playing for war. The great middle class, and the German people
in the mass, desired only to get on with their work, to develop their
commerce, and to enjoy a peaceful home life in increasing comfort.
The question of future peace or war lay with the view which would
prevail. I believed that, without unnecessary provocation on the part
of England, rather with generous and friendly relations, the peaceful
disposition of the German people would prevail over the military
caste and its intensive propaganda....
I was wrong, and the articles I wrote in an analytical but friendly
spirit were worse than useless, though I am still convinced that the
German people as a whole did not want war, until their rulers
persuaded them that the Fatherland was in danger, called to their
patriotism, and let loose all the primitive emotions, sentiments,
ideals, passions, and cruelties which stir the hearts of peoples, when
war is declared.
After that visit to Germany, I went several times to Ireland, and
although there seemed to be no link between these two missions, I
am certain now that in the mind of German agents, politicians, and
military strategists, the situation in Ireland was not left out of
account in their estimate of war chances. With labor “unrest” from
the Clyde to Tonypandy, with suffragette outrages revealing a
weakness and lack of virility (from the German point of view) in
English manhood, and with Ireland on the edge of civil war which
would involve great numbers of British troops, England was losing
her power of attack and defense. So as we know, German agents,
like the Baron von Zedlitz, were writing home in their reports.
Sir Edward Carson, afterward Lord Carson, with F. E. Smith,
afterward Lord Birkenhead (so does England reward her rebels!)
were arranging a bloody civil war in Ireland, which, but for a Great
War, would have spread to England, without let or hindrance from
the British government.
When the Home Rule Bill, under Asquith’s premiership, was nearing
its last stages, Carson raised an army of Ulstermen and invited every
Protestant and Unionist to take a solemn oath in a holy league and
covenant to resist Home Rule to the very death. I was an eyewitness
of many remarkable and historic scenes when “King Carson,” as he
was called in irony by Irish Home Rulers, inspected his troops, made
a triumphal progress through Ulster, stirring up old fires of racial and
religious hatred.
There was a good deal of play-acting about all this, and Carson was
melodramatic in all his speeches and gestures, with a touch of Irving
in the rendering of his pose as a grim and resolute patriot and leader
of Protestant forces, but there was real passion behind it all, and the
sincerity of fanaticism. If it came to the ordeal of battle, these young
farmers and shopkeepers who paraded in battalions before Carson
and his lieutenants, marching with good discipline, a strong and
sturdy type of manhood, would fight with the courage and
ruthlessness of men inspired by hatred and bigotry.
The British government pooh-poohed Carson’s “army” and described
it as an unarmed rabble. But a very brief inquiry convinced me that
large quantities of arms were being imported into Belfast and
distributed through Ulster. There was hardly a pretense at secrecy,
and the Great Western Railway authorities showed me boxes bearing
large red labels with the word “Firearms” boldly printed thereon. The
proprietor of one of the Belfast hotels led me down into his cellars
and showed me cases of rifles stacked as high as the ceiling. He told
me they came from Germany. I went round to the gunsmith shops,
and I was told that they were selling cheap revolvers “like hot
cakes.” There was hardly a man in Ulster who had not got a firearm
of some kind or other. “It’s good for business,” said one of the
gunsmiths, laughing candidly, “but one of these days the things will
go off, and there will be the devil to pay. Why the British government
allows it is beyond understanding.”
The British government did not acknowledge the truth of it. I made
a detailed report of my investigations to Robert Donald, who passed
it on to Winston Churchill, and his comment was the incredulous
remark, “Gibbs has had his leg pulled.” But it was Churchill’s leg that
was pulled, very badly, and he must have had a nasty shock when
there were full descriptive reports of a gun-running exploit, done
with perfect impunity, by the conspiracy of Ulster officers and
leaders, military advisers, and men of all classes, down to the
jarveys of the jaunting cars. Carson had armed his troops—with
German rifles and ammunition.
In view of later history, there must have been some gentlemen of
Ulster whose consciences were twinged by those dealings with
Germany, and by allusions made in the heat of political speeches to
their preference for the German Emperor rather than a Home-rule
House of Parliament in Dublin.
Religious fanaticism was at the back of it all in the minds of the rank
and file. Catholic laborers were chased out of the shipyards by their
Protestant fellow workers, and hardly a day passed without brutal
assaults on them, as was proved by the list of patients in the
hospitals suffering from bashed heads and bruised bodies. I saw
with my own eyes gangs of Ulster Protestants fall upon Catholic
citizens and kick them senseless. Needless to say, there was
retaliation when the chance came, and woe betide any Ulsterman
who ventured alone through the Catholic quarter.
The mediæval malignancy of this vendetta was revealed to me
among a thousand other proofs by a draper’s assistant in a shop
down the Royal Avenue. I was buying a collar stud or something,
and recognizing me as an Englishman, he began to talk politics.
“If they try to put Home Rule over us,” he said, “I shall fight. I’m a
pretty good shot, and if a Catholic shows his head, I’ll plug him.”
He pulled out a rifle, which he kept concealed behind some bundles
of linen, and told me he spent his Saturday afternoons in target
practice.
“What do you think of this? Good shooting, eh?”
He pulled out a handful of pennies and showed me how at so many
paces (I forget the range) he had plugged the head of His Majesty,
King George V. It seemed to me a queer way of proving his loyalty to
the British crown and Constitution.
Carson’s way of loyalty was no less strange. By what method of logic
this great lawyer could justify, as a proof of loyalty and patriotism,
his raising of armed forces to resist an Act of Parliament passed by
the King with the consent of the people, passes my simple
understanding. I can understand rebellion against the law and the
Crown, for Liberty’s sake, or for passion’s sake, or for the destruction
of civilization, or for the enforcement of any kind of villainy. But I
cannot understand rebellion against the law and the Crown in order
to prove one’s passionate loyalty to the law, and one’s ardent
devotion to the King.
Nor can I understand how those who condemn the “direct action” of
Labor in the way of general strikes and other methods of demanding
“rights” (as Lords Carson and Birkenhead and Londonderry
condemned such revolutionary threats), can uphold as splendid
heroism the menace of bloody civil war by a minority which refused
to accept the verdict of the government and peoples of Great Britain
and Ireland.
Sir Edward Carson was an honest man, a great gentleman in his
manner, a great lawyer in repute, but his blind bigotry, some dark
passion in him, made him adopt a line of action which has caused
much blood to flow in Ireland and made one of the blackest chapters
in modern history. For it was the raising of the Ulster Volunteers
which led to the raising of the Irish Republican Army, and the armed
resistance to Home Rule which led to Sinn Fein, and a thousand
murders. It might have led, and very nearly led to civil war in
England as well as in Ireland. When the British Officers in the
Curragh Camp refused to lead their troops to disarm Ulster, and
resigned their commissions rather than fulfil such an order, the
shadow of civil war crept rather close, and there were politicians in
England who were ready to risk it, as when Winston Churchill raised
the cry, “The Army versus the People.”
But another shadow was creeping over Europe, and fell with a chill
horror upon the heart of England, when, as it were out of the blue
sky of a summer in 1914, there came the menace of a war which
would call many great nations to arms, and deluge the fields of
Europe in the blood of youth. Ireland—suffragettes—industrial
unrest, how trivial and foolish even were such internal squabbles
when civilization itself was challenged by this abomination!
In June of 1914—June!—there was a great banquet given in London
to the editors of German newspapers, where I renewed
acquaintance with a number of men whom I had met the previous
year in Germany. Lord Burnham, of The Daily Telegraph, presided
over the gathering, and made an eloquent speech, affirming the
unbreakable ties of friendship between our two peoples. There were
many eloquent speeches by other British journalists, expressing their
admiration for German character, science, art, and social progress. A
distinguished dramatic critic was emotional at the thought of the old
kinship of the German and English peoples. The German editors
responded with equal cordiality, with surpassing eloquence of
admiration for English liberty, literature, and life. There was much
handshaking, raising of glasses, drinking of toasts.... It was two
months before August of 1914.
 XVIII
Fleet Street in the days before the declaration of war was like the
nerve center of the nation’s psychology, and throbbed with all the
emotions of fear, hysteria, incredulity, and patriotic fever, deadened
at times by a kind of intellectual stupor, which took possession of her
people.
It was self-convicted of stupendous ignorance. None of those leader
writers, who for years had written with an immense assumption of
knowledge, had revealed this imminence of the world conflict. Some
of them had played a game of party politics with “the German
menace,” and had used it as a stick for their political opponents. The
Daily Mail, favoring a big navy, and more capital ships, had led the
chorus of “We want eight and we won’t wait.” The Daily News,
favoring disarmament, had denied the existence of any aggressive
spirit in Germany. According to the political color of the newspapers,
Liberal or Tory, the question of German relations had been written
up by the leader writers and news had been carefully selected by the
foreign news editors. But the public had never been given any clear
or authoritative guidance; they had never been warned by the press
as a whole, rising above the political game, that the very life of the
nation was in jeopardy, and that all they had and were would be
challenged to the death. Murder trials, suffragette raids, divorce
court news, the social whirligig, the passionate folly in Ireland, had
been the stuff with which the press had fed the public mind to the
very eve of this crash into the abyss of horror.
Even now, when war was certain, the press said, “It is impossible!”
as indeed the nation did, in its little homes, because their
imagination refused to admit the possibility of that monstrous
cataclysm. And when war was declared, the press said, “It will be
over in three months.” Indeed, men I knew in Fleet Street, old
colleagues of mine, said, “It will be over in three weeks!” Their
theory seemed to be that Germany had gone mad and that with
England, France, and Russia attacking on all sides, she would
collapse like a pricked bladder.
Looking back on that time, I find a little painful amusement in the
thought of our immeasurable ignorance as to the meaning of
modern warfare. We knew just nothing about its methods or
machinery, nor about its immensity of range and destruction.
After the first shock and stupor, news editors began to get busy, as
though this war were going to be like the South-African affair,
remote, picturesque, and romantic. They appointed a number of
correspondents to “cover” the various fronts. They engaged press
photographers and cinema men. War correspondents of the old
school, like Bennett Burleigh, H. W. Nevinson, and Frederick Villiers,
called at the War Office for their credentials, collected their kit, and
took riding exercise in the Park, believing that they would need
horses in this war on the western front, as great generals—dear
simple souls—believed that cavalry could ride through German
trenches.
The War Office kept a little group of distinguished old-time war
correspondents kicking their heels in waiting rooms of Whitehall,
week after week, and month after month, always with the promise
that wonderful arrangements would be made for them “shortly.”
Meanwhile, and at the very outbreak of war, a score of younger
journalists, without waiting for War Office credentials, and
disobeying War Office orders, dashed over to France and Belgium,
and plunged into the swirl and backwash of this frightful drama.
Some of them had astounding and perilous adventures, in sheer
ignorance, at first, of the hazards they took, but it was not long
before they understood and knew, with a shock that changed their
youthful levity of adventure into the gravity of men who have looked
into the flames of hell, and the torture chamber of human agony.
Henceforth, between them and those who had not seen, there was
an impassable gulf of understanding....
Owing to the rigid refusal of the War Office, under Lord Kitchener’s
orders, to give any official credentials to correspondents, the British
press, as hungry for news as the British public whose little
professional army had disappeared behind a deathlike silence,
printed any scrap of description, any glimmer of truth, any wild
statement, rumor, fairy tale, or deliberate lie, which reached them
from France or Belgium; and it must be admitted that the liars had a
great time.
A vast amount of lying was done by newspaper men who accepted
the official statements of French Ministers, hiding the frightful truth
of the German advance. It was an elaboration of the French
communiqués which in the first weeks of the war were devoid of
truth. But a great deal of imaginative lying was accomplished by
young journalists, who at Boulogne, Calais, Dunkirk, Ghent, or Paris,
invented marvelous adventures of their own, exaggerated affairs of
outposts into stupendous battles, and defeated the Germans time
and time again in verbal victories, while the German war machine
was driving like a knife into the hearts of Belgium and France.
Reading the English newspapers in those early days of the war, with
their stories of starving Germany, their atrocity-mongering, their wild
perversions of truth, a journalist proud of his profession must blush
for shame at its degradation and insanity. Its excuse and defense lie
in the psychological storm that the war created in the soul of
humanity, from which Fleet Street itself—very human—did not
escape; in the natural agony of desire to find some reason for
hopefulness; in the patriotic necessity of preventing despair from
overwhelming popular opinion in the first shock of the enemy’s
advance; and in the desperate anxiety of all men and women whose
heritage and liberties were at stake, to get some glimpse behind the
heavy shutters of secrecy that had been slammed down by military
censorship.
I was one of those who did not wait for official permits, and plunged
straightway into the vortex of the war game. In self-defense I must
plead that I was not one of the liars! I did not manufacture
atrocities, and had some temperamental difficulty in believing those
that were true, because I believed in the decency of the common
man, even in the decency of the German common man. I did not
invent imaginary adventures, but found tragedy enough, and drama
enough, in the things I saw, and the truth that I found. As I had two
companions most of the time in those early days, whose honor is
acknowledged by all who know them—H. M. Tomlinson and W. M.
Massey—their evidence supported my own articles which, like theirs,
revealed something to our people of the enormous history that was
happening.
Strangely, as it now seems to me, I was appointed artist
correspondent to The Graphic, as I had been in the Bulgarian war,
and I actually made some sketches of French mobilization and
preparations for war, which were redrawn and published. But my old
paper, The Daily Chronicle, desired my services and I changed over
to them, and abandoned the pencil for the pen, with The Graphic’s
consent, a few days after the declaration of war.
I had crossed over to Paris on the night the reservists had been
called to the colors in England, although so far war had not been
declared by England or France. But the fleet was cleared for action,
and ready, and that night destroyers were out in the English Channel
and their searchlights swept our packet boat, where groups of
Frenchmen who had been clerks, hairdressers, and shop assistants
in England were singing “The Marseillaise” with a kind of religious
ecstasy, while in the saloon a party of Lancashire lads were getting
fuddled and promising themselves “a good time” on a week-end trip
to Paris, utterly unconscious of war and its realities.
In The Daily Chronicle office in Paris, where I had done night duty so
often, my friend and colleague, Henri Bourdin, was white to the lips
with nervous emotion, and constantly answered telephonic inquiries
from French journalists: “Is England coming in? Nothing official, eh?
Is it certain England will come in? You think so? Name of God! why
doesn’t England say the word?”
It was the consuming thought in all French minds. They were
desperate for an answer to their questions. Because of the delay,
Paris was suspicious, angry, ready for an outbreak of passion against
the English tourists, who were besieging the railway stations, and
against English journalists, who were in a fever of anxiety.
I saw the unforgetable scenes of mobilization in Paris, which made
one’s very heart weep with the tragedy of those partings between
men and women, who clung to each other and kissed for the last
time—so many of them for the last time—and on the night of August
2nd I went with the first trainload of reservists to Belfort, Toul, and
Nancy. All through the night, at every station in which the train
stopped, there was the sound of marching men, and the song of
“The Marseillaise”:
“Formez vos bataillons!”
The youth of France was trooping from the fields and workshops,
not in ignorance of the sacrifice to which they were called, not light-
heartedly, but with a simple and splendid devotion to their country
which now, in remembrance, after the years of massacre and of
disillusion, still fills me with emotion....
I do not intend here to give a narrative of my own experiences of
war. I have written them elsewhere, and what do they matter,
anyhow, in those years when millions of men faced death daily and
passed through an adventure of life beyond all power of imagination
of civilized men? I will rather deal with the subject of the Press in
war, and with the peculiar difficulties and work of the
correspondents, especially in the early days.
For the first few months of the war we had no status whatever.
Indeed, to be quite plain, we were outlaws, subject to immediate
arrest (and often arrested) by any officer, French or British, who
discovered us in the war zone. Kitchener refused to sanction the
scheme, which had been fully prepared before the war, for the
appointment of a small body of war correspondents whose honor
and reputation were acknowledged, and gave orders that any
journalist found in the field of war should be instantly expelled and
have his passport canceled. The French were even more severe, and
sent out stern orders from their General Headquarters for the arrest
of any journalist found trespassing in the zone of war.
For some time, however, it was impossible to enforce these rules.
The German advance through Belgium and Northern France was
only a day or two, or an hour or two behind the stampede of vast
populations in flight from the enemy. The roads were filled with
these successive tides of refugees. The trains were stormed by
panic-stricken folk, and even the troop trains found room in the
corridors and on the roofs for swarms of civilians, men and women.
Dressed in civilian clothes, unshaved and unwashed, like any of
these people, how could a correspondent be distinguished or
arrested? Who was going to bother about him? Even the spy mania
which seized France very quickly and feverishly did not create, for
some time, a network of restriction close enough to catch us. I
traveled for weeks in the war zone on a pass stamped by French
headquarters, permitting me to receive the daily communiqué from
the War Office in Paris. I had dozens of other passes and permis de
séjour from local authorities and police, which enabled me to travel
with perfect facility, provided I was able to bluff the military guards
at the railway stations, who were generally satisfied with those
bunches of dirty passes and official-looking stamps. There was, too,
a dual control in France, and a divergence of views regarding war
correspondents. The civil authorities—prefects, mayors, and police—
favored our presence, desired to let us know the suffering and
heroism of their people, and welcomed us with every courtesy,
because we were English and their allies. Often they turned a blind
eye to military commands, or were ignorant of the orders against us.
Massey, Tomlinson, and I, working together in close comradeship, in
those first weeks of war, traveled in Northern France and Belgium
with what now seems to me an amazing freedom. We were caught
up in the tide of flight from French and Belgian cities. We saw the
retreat of the French army through Amiens, from which city we
escaped only a short time before the entry of Von Kluck’s columns.
We came into the midst of the British retreat at Creil, where Sir John
French had set up his headquarters; mingled with the crowds of
English and Scottish stragglers, French infantry and engineers, who
were falling back on Paris, before the spearheads of the German
invasion, with a world of tragedy behind them, yet with a faith in
victory that was mysterious and sublime. We had no knowledge of
the enemy’s whereabouts and set out in simple ignorance for towns
already in German hands, or alighted at stations threatened with
immediate capture. So it was at Beauvais, where we were the only
passengers in a train that pulled over a bridge where a cuirassier
stood by bags of dynamite ready to blow it up, and where the last of
the civilian population had trudged away from streets strewn with
broken glass. Only by a strange spell of luck did we escape capture
by the enemy, toward whose line we went, partly in ignorance of the
enormous danger, partly with foolhardy deliberation, and always
drugged with desire to see and know the worst or the best of this
frightful drama.
We were often exhausted with fatigue. On the day we came into a
deserted Paris, stricken with an agony of apprehension that the
Germans would enter, I had to be carried to bed by Tomlinson and
Massey, as helpless as a child. A few days later, Massey, a strong
man till then, but now ashen-faced and weak, could not drag one leg
after another. We had worn down our nervous strength to what
seemed like the last strand, yet we went on again, in the wagons of
troop trains, sleeping in corridors, the baggage rooms of railway
stations, or carriages crammed with French poilus, who told
narratives of war with a simplicity and realism that froze one’s blood.
We followed up the German retreat from the Marne, when the
bodies of the dead were being buried in heaps and the fields were
littered with the wreckage of battle, and then went north to Dunkirk,
bombed every day by German aëroplanes, but crowded with French
fusiliers, marins, Arabs, British aviators of the Royal Navy, and
Belgian refugees. Here I parted for a time with Massey and
Tomlinson, and in a brief experience as a stretcher bearer with an
ambulance column attached to the Belgian army, saw into the
flaming heart of war, at Dixmude, Nieuport, and other places, where
I became familiar with the sight of death, dirty with the blood of
wounded men, and sick with the agony of this human shambles—a
story which I have told in my book, The Soul of the War.
Other men, old friends of mine in Fleet Street, were having similar
adventures, taking the same, or greater, hazards, dodging the
military authorities with more or less luck. Hamilton Fyfe, then of
The Daily Mail and now editor of The Daily Herald, was caught in a
motor car by a patrol of German Uhlans, and only escaped becoming
a prisoner of war by an amazing freak of fortune. George Curnock,
also of The Daily Mail, was arrested by the French as a spy, and very
nearly shot. A little group of correspondents—among them Ashmead
Bartlett—were flung into the Cherche Midi prison and treated for a
time like common criminals. I happened to fall into conversation with
a French officer, who had actually arrested them. He was strongly
suspicious of me, and asked whether I knew these gentlemen, all of
whose names he had in his pocket book. I admitted that I had heard
of one or two of them by repute, and expected to be arrested on the
spot. But this officer had been French master at an English public
school and was anxious, for some reason, to get an uncensored
letter to the head master. I told him I was going to England, and
offered to take it.... I was not arrested that time.
Another adventurer was young Lucian Jones, son of the famous
playwright, Henry Arthur Jones. He made frequent trips to the
Belgian front and was one of the last to leave Antwerp after the
siege, which was not a pleasant adventure when heavy shells
smashed the houses on every side of him. As he made no disguise
whatever of his profession and purpose, he was sent back to
England and forbidden to show his face again. He took the next boat
back, and was again arrested and flung into a dirty prison. His editor,
who received word of his plight, sent a message to General Bridges,
asking for his release, and obtained the brusque answer, “Let the
fellow rot!”—only it was a stronger word than “fellow.”
One great difficulty we had in those days was to get our messages
back to our newspapers. Sometimes we intrusted them to any
chance acquaintance who was making his way to England. Several
times we had to get back to the coast, in those terrible refugee
trains, to bribe some purser on a cross-Channel steamer. When that
became too dangerous—because it was strictly forbidden by the
military and naval authorities—we made the journey to London,
handed in our messages, and hurried back again the same day to
France. The mental state of our newspaper colleagues exasperated
us. They seemed to have no understanding whatever of what was
happening on the other side, no conception of that world of agony.
“Had a good time?” asked a sub-editor, hurrying along the corridor
with proofs—and I wanted to choke him, because of his placid
unconsciousness of the things that had seared my eyes and soul.
I could not bear to talk with men who still said, “It will be over in
three months,” and who still believed that war was a rather jolly,
romantic adventure, and that our little professional army was more
than a match for the Germans who were arrant cowards and no
better than sheep. In Fleet Street, at that time, there was no vision
of what war meant to the women of France and Belgium, to the
children of the refugees, to the mothers and fathers of the fighting
men. It had not touched us closely in those first weeks of war.
My vexation was great one morning, after one of these journeys
home, when I missed the train to Dover, and my good comrades
Massey and Tomlinson—by just a minute. Perhaps I should never
see them again. They would be lost in the vortex.
“Take a special train,” said my wife.
The idea startled me, not having the mentality or resources of a
millionaire.
“It’s worth it,” said my wife, who is a woman of big ideas.
I turned to the station master, who was standing at the closed gates
of the continental platform.
“How long would it take you to provide a special train?”
He smiled.
“No longer than it would take you to pay over the money.”
“How much?”
“Twenty-two pounds.”
I consulted my wife again with raised eyebrows, and she nodded.
I went into a little office, half undressed, and pulled out of my belt a
pile of French gold pieces. By the time they had been counted and a
receipt given—no more than three minutes—there was a train with
an engine and three carriages, a driver and a guard, ready for me on
the line to Dover. My small boy (as he was then) gazed in awe and
admiration at the magic trick. I waved to him as the train went off
with me. I was signaled all down the line, and in the stations we
passed porters and officials stared and saluted as the train flashed
by. Doubtless they thought I was a great general going to win the
war! At Dover I was only one minute behind the express I had lost.
Massey and Tomlinson were pacing the platform disconsolately at
the loss of their comrade. They could not believe their eyes when I
walked up and said “Hello!” So we went back to a new series of
adventures.
I used with success, three times running, another method of getting
my “dispatches” to Fleet Street. After the third time some intuition
told me to change the plan. At that time, as all through the war, a
number of King’s messengers—mostly men of high rank and
reputation—traveled continually between British G.H.Q. and the War
Office, with private documents from the Commander-in-Chief. Three
times did I accost one of these officers—a different man each time—
in an easy and confidential manner.
“Are you going back to Whitehall, Sir?”
“Yes. What can I do for you?”
“I shall be much obliged if you will put this letter in your bag, and
deliver it at the War Office.”
“Certainly, my dear fellow!”
My letter was addressed to The Daily Chronicle, care of the War
Office, and, much to the surprise of my editor, was punctually
delivered, by a War-Office messenger. But my intuition was right.
After the third time the editor of The Daily Chronicle received word
from the War Office that if Gibbs sent any more of his articles by
King’s messenger, they would be destroyed.
The method of delivery became easier afterward, because the
newspapers organized a series of their own couriers between
England and France, and that system served until the whole courier
service was rounded up and forbidden to set foot in France again.
It was amazing that my articles, and those of my fellow
correspondents, were allowed to appear in the newspapers, in spite
of military prohibition. But the press censorship, which had been set
up by the government under the control of F. E. Smith, now Lord
Birkenhead, was not under direct military authority, and was much
more tolerant of correspondents who evaded military regulations. I
wrote scores of columns during the first few months of the war,
mostly of a descriptive character, and very few lines were blacked
out by the censors. So far from being in the black books of the press
censorship as established at that time, I was sent for by F. E. Smith,
who thanked me for my narratives and promised to give personal
attention to any future dispatches I might send. This was at the very
time when Kitchener himself gave orders for my arrest, after reading
a long article of mine from the Belgian front.
I was also received several times by Sir William Tyrrell, Secretary to
the Foreign Office, who questioned me about my knowledge of the