1

Introduction

1.1 Introduction: a general view on the passive voice and

its description

The subject matter of this book is the semantic and pragmatic analysis
of the English passive voice in diachronic perspective. The passive voice
has received much attention over the past several decades. Grammati-
cal voice itself is complex and there are a number of properties yet to
be analysed. A large number of previous publications are dedicated to
issues of grammatical voice per se (e.g. Siewierska 1984; Keenan 1985;
Kemmer 1993; Geniušienė 1987; Kleiman 1991; just to name a few) or
to the interrelationship within the voice systems, sometime known as
the voice continuum (e.g. Croft 1994, 2001: 283–319; Givón 1990: 563–
644; Palmer 1994: 142–75; Shibatani 1985, 1998), which reveal that the
active, passive and middle voices are somehow related to each other and
that there are certain patterns among them, as we will examine in detail
in Chapter 3.
The system of grammatical voice has been presented in the gram-
mar books of numerous languages and it is safe to say that it is almost
always mentioned in some way. However, what is treated as the passive
may vary from book to book. This means that the actual languages are
described according to a scholar’s own discipline, belief, intuition, etc.
Thus, there is a danger of misinterpreting the data. There certainly was,
and there remains to a certain degree, a trend of having an anglocentric
view of the description of grammar in modern linguistics. The gram-
mar of various languages was described on the basis of constructions
in English alone. The passive is a good example of such cases: in some
languages, a periphrastic construction similar to ‘copula + main verb in
past participle’, based on its surface structure, is blindly named passive

J. Toyota, Diachronic Change in the English Passive

© Junichi Toyota 2008
2 Diachronic Change in the English Passive

and if there is no matching construction, the passive meaning is often

rendered by alternative constructions, such as indefinite pronouns. The
English structure has unconsciously been influencing the description of
other languages (Andersen 1990: 142ff.).
Constructions and their properties vary from language to language.
So the property of the passive in one language is different from that
in another. This causes various problems for the learner of a foreign
language (see, for example, Fredriksson 2001 for Swedish and English
passive constructions and Swan and Smith 1987 for various languages
in comparison with English) or in the area of translation (Filipović
2002 lists some cases of different interpretation in Serbo-Croatian and
English motion verbs and particles). These problematic features are
all synchronic matters and once they are considered diachronically, a
translation from earlier languages, say Old English or Old Japanese,
to Present-day English or Japanese, may create various challenges for
translators, although the languages are basically English or Japanese
of some sort. By looking at a construction in various languages, its
grammatical properties become clearer. There are a number of historical
works on the English passive which are predominantly syntax-oriented.
However, there is little work on the passive in the area which empha-
sises the interaction with other similar ‘passive-like’ constructions, both
syntactically and semantically (except for some previous work such as
Givón and Yang 1994, Haegeman 1985, Toyota 2007, for the relation-
ship between get-passive and reflexive-causative). Also, in previous work,
the definition of passive is rather unchallenged, i.e. the construction
‘auxiliary be and past participle’ is considered to be either a verbal or
adjectival passive and get+past participle is automatically taken as a
type of passive. As a result, some very interesting interactions of the
voice system in English may be overlooked, especially at the diachronic
level.

1.2 Various approaches to the voice system

The passive voice is often associated with its active counterpart because
of the syntactic correspondence between the subject and the object, i.e.
the active object corresponds to the passive subject, the active subject to
the oblique agent phrase in the passive. Thus, a boundary between the
active and the passive is often assumed. This syntactic property seems
to override differences in terms of semantics and pragmatics in the pas-
sive, such as topicality change and impersonalisation. This syntax-based
 Introduction 3

analysis has an enormous influence on the grammatical approaches

illustrated below, and both traditional and modern approaches assume
this influence from the syntactic relationship.
Broadly speaking, traditional approaches to the passive are purely
descriptive, while modern approaches can be usefully divided into three
major types (for present purposes): structural, relational and functional.
The first two may jointly be called formal approaches. The characteris-
tic of these approaches is the use of a fully explicit device to reveal the
internal grammatical system. In addition, they take full advantage of
mathematical or logical methods for analysis. They also generally posit
a strict distinction between syntax and semantics. Within these formal
approaches, accounts have varied between more syntactic and more lex-
ical views of the passive. Functional approaches are distinguished by
their concern for explanations of the influence from context and various
attempts to unify pragmatics and semantics with syntax. This division
can be schematised in Figure 1.1.
In addition, there are some semantic and pragmatic approaches
(among others, Shibatani 1985; Givón 1983, 1994; Foley and Van Valin
1984), which normally involve dealing with a large amount of natural
occurring data.
What is common among all approaches is the assumption of a categor-
ical boundary, such as NOUN, VERB, ADJECTIVE, etc. These distinctions
are normally made according to syntactic behaviour. A similar distinc-
tion can be applied to the voice system, i.e. ACTIVE, PASSIVE and

Approaches to the passive voice

Traditional Formal Functional

Structural Relational

syntactic lexical syntactic lexical

• Descriptive • Transformational • LFG • Relational Grammar • Functional grammar

grammar • GPSG • Optimality theory • Cognitive grammar
• Principle and Parameter • HPSG • Construction grammar
theory • Optimality • Role and reference
• Government and Binding theory grammar
theory
• Minimalism
• Optimality theory

Figure 1.1 Various approaches to the passive voice

4 Diachronic Change in the English Passive

MIDDLE are rigidly distinguished, although, we may note, less so in

functional approaches, where some fuzzy boundaries are often observed.

1.3 Aim of the study

The general objective of this book is to improve our understanding

of the voice continuum in English, focusing especially on the pas-
sive voice. I analyse the significance of the category PASSIVE within
grammatical voice and how this category changes over time. I include
those ‘passive-like’ constructions (syntactically similar constructions)
and other non-passive constructions which produce the same reading
or effect as the passive (semantically and pragmatically similar construc-
tions). In addition I compare changes in the passive and in those related
constructions. As we will see, there are numerous cases where there is no
clear division among different types of constructions. Thus my view is
that the voice system is best treated as an example of gradience, where
the three different voice types are treated in a network or continuum of
semantic characteristics.
The main aim is to reveal the conceptual development of the English
passive and describe its historical changes. It is not intended to explore
the various grammatical voice constructions in a particular grammatical
framework, as this is basically a descriptive work, although it is strongly
biased towards the functional approaches shown in Figure 1.1 above.
Thus, some of the treatments used in these approaches will be shown
on various occasions. This study mainly focuses on the analysis of the
English voice system, which has been rather unchallenged for purposes
of description. However, I often look at the English voice system from
the typological point of view, which leads to an unconventional set of
definitions.

1.4 Method

A variety of data sources will be employed for the analysis: corpora

are used for statistical analysis (see 1.4.2 below). It should be noted
that some examples were collected manually, since certain construc-
tions or phrases may happen not to appear in corpora. Also, a limited
amount of typological data, for example, from other Indo-European lan-
guages (henceforth IE languages), are incorporated in order to highlight
a particular construction or its change in English. On the diachronic
level, I sometimes refer to some reconstructed ancient languages such as
 Introduction 5

Proto-Indo-European (henceforth PIE). Several points necessary for the

analysis are described below.

1.4.1 Diachronic classification of the English language

There are various conventional divisions of the English language into
time periods. The following broad division is generally agreed upon
among scholars and therefore widely used: Old English (OE) (700–
1100), Middle English (ME) (1100–1500), Modern English (ModE)
(1500–present), and Present-day English (PDE). However, this is some-
times broken down into finer divisions which divide ME into Early
Middle English (eME) (1100–1350) and Late Middle English (lME) (1350–
1500), and ModE into Early Modern English (eModE) (1500–1700) and
Late Modern English (lModE) (1700–present). This classification can be
schematised in Figure 1.2. In the present work, both broad and finer
divisions are adopted and applied according to the importance of the
specificity of the period.

700 1100 1350 1500 1700 present

eME lME eModE lModE
OE ME ModE PDE

Figure 1.2 Diachronic classification of periods in the English language

1.4.2 Data
A linguistic study ideally involves analysis of both spoken and written
data. The crucial aspect in historical work is that access to spoken data is
unavailable. Therefore, we have to bear in mind the effect of its absence.
Fortunately there are several electronic databases for English data from
OE to PDE. The advantage of such databases is that the registers are
well mixed, making the result more representative. For this study I have
drawn on the corpora listed in Table 1.1 (on page 6).
In addition, HC usefully divides each period so that it includes even
finer periods (finer than the distinction shown in Figure 1.2 above), as
shown in the Table 1.2 (details taken from Kytö 1996: 233–48). When
the data taken from HC are mentioned, we refer to the finer period. So
for example, when an example is taken from the OE text Beowulf, it is
indicated as (HC OE3 cobeowulf).
Since dialectal difference is not of interest here, British English alone
will be analysed. The electronic database is useful for statistical purposes,
but when it comes to analysing a particular verb phrase or construction,
6 Diachronic Change in the English Passive

Table 1.1 Size of corpora used in the study1

Corpus Period Size in words

• Helsinki (HC) OE (700–1150) 413,300

ME (1150–1500) 608,570
eModE (1500–1710) 551,000
• ARCHER lModE (1710–1960) 606,634
• London-Lund (LL) PDE (spoken) 623,784
• Lancaster-Oslo/ Bergen (LOB) PDE (written) 1,214,752

Table 1.2 Further period distinctions in HC

General distinction Finer distinction Periods

OE OE1 –850
OE2 850–950
OE3 950–1050
OE4 1050–1150
ME ME1 1150–1250
ME2 1250–1350
ME3 1350–1420
ME4 1420–1500
eModE E1 1500–1570
E2 1570–1640
E3 1640–1710

it may not contain a useful example. Thus, these electronic data are com-
bined with some secondary sources: Visser (1963–73), Denison (1993),
Mustanoja (1960), Mitchell (1985), the Oxford English Dictionary (OED),
the Middle English Dictionary (MED) and The Anglo-Saxon Poetic Records
(ASPR).

1.5 Organisation of the Study

This book starts with the analysis of be-passive, then moves on to differ-
ent constructions, such as get-passive, This TV needs fixing construction,
the use of indefinite pronouns, etc. Chapter 2 focuses on the overview
of the passive voice, including a taxonomic system of the passive, both
form and function, in addition to the aspectual issues concerning the
English be-passive. I will deal with the development of the perfective
aspect in relation to the passive, as well as some specific forms, such as
 Introduction 7

progressive passive and perfective passive, which are considered impor-

tant indicators of grammaticalisation. This chapter also introduces the
general history of the English passive.
Chapter 3 deals with detailed morphosemantic and syntactic analy-
sis of each component of the English be-passive, i.e. auxiliary and past
participle. The auxiliary is analysed in an historical context implying
gradience, and the past participle is dissected into small parts, such as
the prefix ge-, the suffix -ed, specific participles including stative verbs,
prepositional phrases, etc. Then the relationship between the auxil-
iary and past participle is analysed from functional perspectives, which
supplement various characteristics discussed earlier in this chapter.
Chapter 4 introduces various functional aspects of the passive voice.
Various basic functions, such as topicality change or impersonalisation,
are explored, but at the same time, some marginal cases which violate
the basic functional characteristics. The gradient nature of functions will
also be highlighted in this chapter. Chapter 5 focuses on the historical
changes in the functions of the English passive. In particular, I discuss
word order change, which is crucial in the formation of the passive func-
tionally. This chapter also introduces the notion of voice continuum,
which will be crucial in later chapters.
Chapters 6–8 analyse various passive constructions in English. Chap-
ter 6 is dedicated to the analysis of the get-passive. I analyse its semantic
characteristics first and based on this, I argue for a certain type of his-
torical development. In Chapter 7 we are concerned with constructions
which have an undergoer subject without the overt marking as pas-
sive. Constructions analysed here show the voice continuum with other
grammatical voice systems. This chapter also contains the analysis of
constructions related to modality. Chapter 8 is about constructions with-
out overt passive marking, but which share the same meanings produced
by the canonical passive. This chapter contains analysis of inversion and
the use of indefinite pronouns.