Vous êtes sur la page 1sur 31

Word

ISSN: 0043-7956 (Print) 2373-5112 (Online) Journal homepage: http://www.tandfonline.com/loi/rwrd20

Cleft sentences in English: A comprehensive view

Edward G Fichtner

To cite this article: Edward G Fichtner (1993) Cleft sentences in English: A comprehensive view,
Word, 44:1, 1-30, DOI: 10.1080/00437956.1993.11435891

To link to this article: https://doi.org/10.1080/00437956.1993.11435891

Published online: 16 Jun 2015.

Submit your article to this journal

Article views: 1597

Citing articles: 1 View citing articles

Full Terms & Conditions of access and use can be found at


http://www.tandfonline.com/action/journalInformation?journalCode=rwrd20
EDWARD G F I C H T N E R - - - - - - - - - - -

Cleft sentences in English:


A comprehensive view

Abstract. This article addresses four questions: first, what sentence


patterns may be considered 'cleft sentences'; second, how they are related
to one another; third, what function they serve within the larger structure
of English; and fourth, where within the structure of English they may
be situated. It is shown that there are three cleft sentence patterns which
are produced by a single set of operations on a normal English sentence;
each of the three-simple clefts, WH-clefts, and IT-clefts-is the output
of a different stage of these operations. Cleft sentences, like structures
introduced by it and there, belong to the group of Topic-Comment
Redistribution transformations. These allow the speaker or writer to shift
some of the component elements of a sentence to different positions, in
order to make those elements more prominent, and to make the sequence
of elements in the sentence conform more closely to the sequence of ideas
in the text of which that sentence is a part.

Most linguists who concern themselves with the structure of En-


glish would be familiar with the term 'cleft sentence', and would use
the term to designate expressions like:

(1) it was the butler served who the wine

Others might include in this category sentences like:

(2) the one who served the wine was the butler

and yet a few others might regard sentences such as:

(3) the butler was the one who served the wine

as cleft sentences, too. In this article, we will try to answer four


questions connected with these sentence patterns: first, what kinds of
sentence patterns may be considered cleft sentences, properly speak-
ing; second, whether these patterns can be related to one another in
any systematic way, and if so, how; third, what communicative func-

1
2 WORD, VOLUME 44, NUMBER I (APRIL, 1993)

tion they serve within the larger inventory of sentence-level structures


of English; and fourth, where within the syntactic structure of English
these sentence patterns may be situated. 1
The first linguist in this century to draw attention to these sentence
patterns was Otto Jespersen. He described the IT-clefts, e.g., sentence
(1) above, in some detail in his Modern English grammar (1965:88ff.)
To be sure, he did not call them "cleft sentences" there. However, in
1937, the term appears prominently in his Analytic syntax (1969:
73ff.), along with some speculations about how this construction
arose. As he saw it, the phrase it is plus the connective, if present, is
superimposed upon the simple sentence, which is the semantic sub-
stance of the cleft. Thus, the expression:

(4) it is the wife who decides

results from the insertion of the clefting elements it is . . . who into the
simple sentence:

(5) the wife decides

For Jespersen, the cleft sentence is regarded, not as a distinct syntactic


construction, but merely as the modification of a simpler sentence
pattern.
Lees (1963) takes Jespersen's work as his point of departure, as
many of his examples show. He also regards the IT-cleft as derived
from a simple sentence, but not directly. In his view, the sentence the
wife decides is reshaped by doubling the pivotal element the wife and
moving it to initial position, and then inserting a WH-morpheme before
the original phrase. In other words, the wife decides first becomes:

(6) the wife WH-the-wife decides

This sequence is inserted into the frame it is Cc, in which the 'Cc' is
the cleft complement; then the WH- element is reduced. This produces:

(7) it is + the wife WH- ... decides


As Lees sees it, the IT-cleft is still the modification of a simple sen-
tence, but the transformation by which that cleft is produced goes
through several stages.
Like Lees, Akmajian (1970) also regarded the 'genuine cleft' (his
term for the IT-cleft) as derived, not from a simple sentence, but what
FICHTNER: ENGLISH CLEFr SENTENCES 3

he called the 'pseudo-cleft' (his term for the WH-cleft). The IT-cleft
was derived from the latter by Chomskyan transformations in the style
of that time. Akmajian thus believed that sentences like:
(8) it was Agnew who Nixon chose
were derived from sentences of the pattern:
(9) the one who Nixon chose was Agnew
In addition, Akmajian regarded the verb be, which is characteristic of
both of the forms of cleft sentences which he considered, as present in
the underlying structure. In so doing, he detached these sentences from
the simple sentence pattern Nixon chose Agnew. Consistent with this
position, Akmajian makes no attempt to connect the pseudo-cleft with
any deeper layer of structure.
A number of studies of cleft sentences appeared during the sev-
enties in the wake of Akmajian's article. Higgins (1971) followed
Akmajian in deriving the IT-cleft from the WH-cleft, which he saw as
originating in deep structure. Hankamer ( 1974) and Pinkham &
Hankhamer (1974) proposed dual sources for them, one in which they
are derived from a nonclefted structure, another in which they are
base-generated. Gundel (1976, 1977), on the other hand, regarded the
IT-cleft as a reduced form of right-dislocated WH-clefts. More re-
cently, Knowles (1986) has taken a somewhat different view of these
sentence patterns. He regards clefts (IT-clefts) and pseudo-clefts (WH-
clefts) as derived from different underlying structures, although the
clefts overlap with both pseudo-clefts and with relative clauses. As in
his earlier studies, Knowles takes as his point of departure structures
which contain some form of be.
In the scholarship reviewed thus far, several notions about the
structure of cleft sentences can be distinguished. To begin with, there
is the question of the definition of a cleft sentence. While all who have
studied this phenomenon agree that the IT-cleft, i.e., sentence (1), is
indeed a cleft sentence, the use of the term 'pseudo-cleft' for the
WH-cleft, i.e., sentence (2), implies merely its similarity to, but not
actual inclusion in the category of cleft sentences. The simple cleft,
i.e., sentence (3), is usually ignored, although from time to time it is
referred to as somehow coming into comparison with the other two
clefts. Another question has to do with the origin of cleft sentences,
i.e., whether the they are derived from a simple underlying sentence,
from a pseudo-cleft (WH-cleft), or directly from deep structure.
4 WORD, VOLUME 44, NUMBER I (APRIL, 1993)

To begin with, we note in forming a working hypothesis that there


are three sentence patterns which are so closely related in meaning,
use, and structure that it is reasonable to group them together as cleft
sentences, namely, sentences (1)-(3) above. Moreover, almost any
native speaker of English would acknowledge that the semantic sub-
stance of those sentences is expressed by the simple sentence:

(10) the butler served the wine.

Not only do those three sentences exhibit an affinity of meaning with


the latter sentence, but there are many other sentence patterns which do
so, too. For example, this simple sentence could be so modified as to
yield three more cleft sentences:

(11) the wine was what the butler served

(12) what the butler served was the wine

(13) it was the wine that the butler served

Furthermore, if the simple underlying sentence is recast in the passive


voice, as many as six additional cleft sentences can be formed:

( 14) the wine was what was served by the butler

(15) what was served by the butler was the wine

(16) it was the wine that was served by the butler

(17) the butler was the one whom the wine was served by

(18) the one whom the wine was served by was the butler

(19) it was the butler whom the wine was served by

And finally, the following additional clefts are obviously variants of


the last three preceding patterns:

(20) the butler was the one by whom the wine was served

(21) the one by whom the wine was served was the butler

(22) it was the butler by whom the wine was served


FICHTNER: ENGLISH CLEFf SENTENCES 5

If we accept these fifteen sentences, i.e., sentences (1)-(3) and


(11)-(22), as all belonging to the category of cleft sentences, the
question arises as to whether they are produced by systematic modi-
fication of the simple sentence, the butler served the wine, or whether
both this sentence and the clefts are derived by different processes from
the same or different levels of the deep structure of English. The fact
that nine of these sentences have undergone the passive transforma-
tion, i.e., sentences (14)-(22), suggests that, whatever the specific
operations in the process of cleft formation, it takes place subsequent
to Passivization, the change in the form of the verb and the accompa-
nying adjustments in the surface roles of subject and object. This in
turn points to a stage in the generation of a sentence at or fairly close
to surface structure as the level on which the formation of cleft sen-
tences takes place.
If the formation of a cleft sentence takes place by various operations
carried out on a sentence already fully formed, then we must try to
define these operations. A clue as to what form these might take is
offered by Jacobs and Rosenbaum (1968:39-40). To be sure, they were
primarily interested in devising a test to show whether a phrase in a
given sentence was a noun phrase. In effect, the procedure they adopted
involved the creation of a cleft sentence. First, they placed the element
WHAT at the beginning of the sentence containing the phrase to be
tested; second, they placed the appropriate form of the verb be at the
end of the sentence; and third, they moved the phrase to be tested to the
end of the sentence. If after these changes the sentence was acceptable,
then the phrase in question was deemed to be a noun phrase.
Let us assume that we wish to determine whether the phrase the
wine in the sentence which we have been using as our example is a
noun phrase. If we apply the operations devised by Jacobs and Rosen-
baum to that sentence, we observe the following changes:

Underlying sentence: the butler served the wine


Operation 1: WHAT the butler served the wine
Operation 2: WHAT the butler served the wine WAS
Operation 3: WHAT the butler served ... WAS the wine

The test demonstrated to their satisfaction that the phrase the wine is
indeed a noun phrase. But it also shows that a cleft sentence can be
formed by a short sequence of superficial modifications of the structure
of a simple sentence, i.e., the insertion of a pronominal element and a
form of the verb be into a simple English sentence, along with a shift
6 WORD, VOLUME 44, NUMBER I (APRIL, 1993)

in the position of one element in the original sentence. It so happens


that Jacobs and Rosenbaum's procedures will generate only WH-clefts.
Nevertheless, a set of very similar operations can be formulated which
will generate all of the cleft sentence patterns listed above in both the
active and passive voice, as well as many others.
The point of departure for the generation of these cleft sentences
is the underlying sentence (US), the fully formed declarative sentence:

(23) the butler served the wine.

First, we designate a noun (or pronoun or adverb) as the 'Focus' of the


cleft. In this sentence, both the butler and the wine are possible foci;
for the purpose of this demonstration, we choose the latter as Focus.
Then, by a procedure called 'Cleftization' (CLFfZ), three elements
are attached to the Focus, namely, the verb be, a Referent (RFNT), and
a Relative Pronoun or Adverb (RLTV). The Focus plus the forms
attached to it become the 'Cleft Phrase' (CLFfPHR). These operations
will be represented in the discussion below in the following way:
US: the butler served the wine
CLFfZ: the butler served the wine BE RFNT RLTV
The next step is to topicalize the Cleft Phrase:
T/CLFfPHR: the wine BE RFNT RLTV + the butler served
The abbreviation means that the operation of Topicalization (T/ ... ),
sometimes called 'left dislocation', has been applied to the element
following the virgule-here, the Cleft Phrase (CLFfPHR). The plus
sign indicates the junction between sequences of words whose positions
have been changed. The sequence of elements at this point is that of a
simple cleft. Now, the various dummy elements, BE, RFNT, and
RLTV, must be given the value which they normally have in surface
structure. First, the verb be must be changed to agree in person and
number with the surface grammatical subject, which by the transpo-
sition just described is now the wine. Moreover, it normally assumes
the tense of the lexical verb in the base sentence. These changes produce
the sequence:

(24) the wine was RFNT RLTV the butler served.

The dummy elements RFNT and RLTV are realized according to


the semantic properties of the Focus, i.e., whether it is + 1- Human,
FICHTNER: ENGLISH CLEFf SENTENCES 7

+I- Specific, and +I- Plural. These combinations are listed in the
table below:

Resolution of RFNT and RLTV in surface structure 2


Focus RFNT + RLTV Reduced form
Nominals
+Hum +Spec +1-Plur the one(s) + who(m) who(m)
-Hum +Spec +/-Plur the one(s) + that that
-Hum + Abstract the thing + that what
-Hum -Spec -Plur that + which (~that) what
-Hum -Spec + Plur those + that that
Adverbials
+Time the time + when/that when
+Place the place + where/that where
+Manner the way + that how
+Cause the reason + why/that why

Since the Focus phrase the wine is -Human, -Specific (it is a mass
noun), and -Plural, the dummy elements RFNT and RLTV are real-
ized as that which. As the table indicates, the combinations of Referent
and Relative also have reduced forms, which may replace the full
sequences except when these have the feature + Specific and stand in
initial position. By these changes and the adjustment of the form of the
verb be, the following simple cleft has been generated:

SOP: "The wine was what the butler served."


The abbreviation 'SOP' stands for 'standard orthography and punctu-
ation'. Presenting the sentence in this way, with the customary visual
attributes of normal English printed text, makes it easier to evaluate its
acceptability.
The other two cleft sentence patterns, the WH-cleft and the IT-
cleft, are formed by further operations on the output of the topicaliza-
tion of the Cleft Phrase, which was:
T/CLFTPHR: the wine BE RFNT RLTV the butler served
The next change in this sequence is the topicalization of be:
T/BE: BE + the wine ... RFNT RLTV the butler served
8 WORD, VOLUME 44, NUMBER 1 (APRIL, 1993)

A WH-cleft is then generated by the subsequent topicalization of the


Referent and all forms which follow it (RFNT&FF):
T/RFNT&FF: RFNT RLTV the butler served + BE the wine
Again, the verb be is changed to agree with the person and number of
the grammatical subject, taking its tense markers from the verb served,
and the elements RFNT and RLTV are modified according to the
semantic properties of the their antecedent. The full form of these
elements produces the sentence:
SOP: "That which the butler served was the wine."
With the reduced form, the sentence finally becomes:
SOP: "What the butler served was the wine."
The third of the three cleft patterns, the IT-cleft, is, in our view,
not derived from the WH-cleft, but is, in a sense, an alternative to it.
Like the WH-cleft, the IT-cleft is formed by a modification of the
pattern:
T/BE: BE the wine .. RFNT RLTV the butler served
However, instead of topicalizing the Referent and forms following
it (RFNT&FF) at this point, the dummy topic it is inserted (II . . .) at
the beginning of the phrase:
1/IT: IT + BE the wine RFNT RLTV the butler served
With the further change in the form of be, the IT-cleft is fully gener-
ated:
SOP: "It was the wine that the butler served."
In resolving the abstract elements RFNT and RLTV, we delete the
Referent in accord with a contact rule which stipulates that a RFNT
which comes to stand after the Focus, with which it is after all coref-
erential, is redundant and is lost.
To summarize, then, all three kinds of cleft sentences can be
derived from the simple underlying sentence by the insertion of various
dummy elements, by their subsequent repositioning by Topicalization,
and by their resolution into the surface forms required by the semantic
features of the Focus. To demonstrate that these procedures are not
simply linguistic accidents which by chance produce the three cleft
sentences above, but but represent actual syntactic properties of the
FICHTNER: ENGLISH CLEFf SENTENCES 9

language, let us apply them to several other expressions-to begin


with, to the other noun in the simple sentence. The operations applied
above are repeated, with the exception that the clefted element, the
Focus, is not the wine, but the butler:
US: the butler served the wine
CLFTZ: the butler BE RFNT RLTV served the wine
To be sure, the topicalization of the Cleft Phrase is redundant here,
since it is already at the beginning. For the sake of consistency, how-
ever, we repeat it, though it has no overt effect:
T/CLFTPHR: the butler BE RFNT RLTV + served the wine
With the adjustments to BE, RFNT, and RLTV applied above, we
have formed the simple cleft:
SOP: "The butler was the one who served the wine."
The remaining clefts presuppose the Topicalization of be:
T/BE: BE + the butler ... RFNT RLTV served the wine
To form a WH-cleft, the element RFNT and forms following it must be
topicalized:
T/RFNT&FF: RFNT RLTV served the wine + BE the butler
As the table above indicates, the realization of RFNT RLTV when the
Focus is +Hum and +Spec is the one who or the one that:
SOP: "The one who served the wine was the butler."
The generation of an IT-cleft requires the insertion of IT into the
sequence produced by the Topicalization of be:
lilT: IT + BE the butler RFNT RLTV served the wine
As before, when RFNT immediately follows the focus element, the
former is deleted:
SOP: "It was the butler who served the wine."
If these procedures are applied to the test sentence in its passive form,
the additional patterns presented earlier can also be generated. In the
first example, the wine is the Focus:
US, PASVZ: the wine was served by the butler
CLFTZ: the wine BE RFNT RLTV was served by the butler
10 WORD, VOLUME 44, NUMBER I (APRIL, 1993)

T/CLFTPHR: the wine BE RFNT RLTV was served by the butler


SOP: "The wine was what was served by the butler."
T/BE: BE + the wine ... RFNT RLTV was served by the butler
T/RFNT&FF: RFNT RLTV was served by the butler+ BE the wine
SOP: "What was served by the butler was the wine."
1/IT: IT + BE the wine RFNT RLTV was served by the butler
SOP: "It was the wine that was served by the butler."
The remaining three cleft sentences are generated when the Focus is
the agent prepositional phrase introduced by Passivization (PASVZ).
However, it is possible in English to make either the entire preposi-
tional phrase by the butler the Focus, or merely the object of the
preposition, the butler. In the following example, the Focus is simply
the butler:
IS, PASVZ: the wine was served by the butler
CLFTZ: the wine was served by the butler BE RFNT RLTV
T/CLFTPHR: the butler BE RFNT RLTV + the wine was served by
SOP: "The butler was the one (whom) the wine was served by."
T/BE: BE + the butler ... RFNT RLTV the wine was served by
T/RFNT&FF: RFNT RLTV the wine was served by + was the
butler
SOP: "The one (whom) the wine was served by was the butler."
1/IT: IT + BE the butler RFNT RLTV the wine was served by
SOP: "It was the butler (whom) the wine was served by. " 3
In the variant of the sentences above, i.e., those in which the Cleft
Phrase includes the agent preposition by, an additional step must be
added to the clefting procedures. While the object of the preposition
by, i.e., the butler, may successfully imitate the grammatical subject
in sentences like:

(25) the butler was the one whom the wine was served by

the rigid SVO structure of English does not admit of prepositional


phrases which occupy that position. Because the entire prepositional
phrase by the butler has been made the Focus of the cleft, the proce-
dures for the rearrangement of various syntactic elements which have
been used in previous examples would now generate an unacceptable
sentence, namely:

(26) ??? by the butler was the one who(m) the wine was served
FICHTNER: ENGLISH CLEFT SENTENCES 11

Hence, an additional operation is required where the Focus includes


elements like the preposition by. The following procedures yield cor-
rectly formed cleft sentences:
US, PASVZ: the wine was served by the butler
CLFTZ: the wine was served by the butler BE RFNT RLTV
T/CLFTPHR: by the butler BE RFNT RLTV + the wine was served
This sequence of elements, as sentence (26) shows, does not constitute
an acceptable English sentence. In order to correct this deficiency in
our set of rules, it is necessary to introduce an additional step in those
cases where the Focus cannot be construed as a subject because of the
presence of the preposition by. In this step, the noun in the Focus,
together with the verb be and the Referent, are topicalized immediately
after the topicalization of the Cleft Phrase:
T/CLFTPHR: by the butler BE RFNT RLTV + the wine was served
T/FOCUS&BE&RFNT: the butler BE RFNT + by ... RLTV the
wine was served
With this addition, the derivation of the various kinds of cleft sentences
can proceed as before. The first to be generated is the simple cleft:
SOP: ! "The butler was the one by whom the wine was served."
The topicalization of be is required for the generation of the other two
clefts:
T/BE: BE + the butler ... RFNT by RLTV the wine was served
When the Referent and all forms following it are topicalized, the
WH-cleft is produced:
T/RFNT&FF: RFNT by RLTV the wine was served + BE the butler
SOP: ! ''The one by whom the wine was served was the butler.''
And finally, by the insertion of IT after the topicalization of be, the
IT-cleft is formed:
1/IT: IT + BE the butler RFNT by RLTV the wine was served
SOP: ! "It was the butler by whom the wine was served."
To summarize, then, we have generated all fifteen cleft sentences,
(1)-(3) and (11)-(22), by a series of changes applied to the underlying
sentence the butler served the wine. These are, first, Passivization
(PASVZ), the normal optional transformation of a sentence containing
12 WORD, VOLUME 44, NUMBER 1 (APRIL, 1993)

a transitive verb from the active to the passive voice. In Cleftization


(CLFfZ), a Focus element is selected, and the dummy elements be +
Referent + Relative (BE RFNT RLTV) are attached to it. The Focus
plus these elements constitutes the Cleft Phrase (CLFfPHR). This
Cleft Phrase must then be topicalized. If the Cleft Phrase is headed by
an element not part of the noun phrase, i.e., the agent preposition by,
the sequence of elements Focus + be + RFNT must then be topical-
ized, too. At this point, a simple cleft has been produced. The remain-
ing cleft sentence patterns require the further topicalization of be. Two
changes can now be made in the sequence thus produced. First, the
Referent and all of the forms which follow it (RFNT&FF) can be
topicalized, producing a WH-cleft. Or, second, the dummy subject/T
can be inserted before be, which generates an IT-cleft.
These procedures are sufficient to generate all of the cleft sen-
tences which can correctly be formed from the underlying sentence
which was our example, and only the correct ones. However, our rules
would be inadequate if some features of that underlying sentence were
changed. If, for example, the object noun the wine were plural instead
of singular, adjustments would have to be made in some cases in the
form of the verb be in surface structure. To illustrate this complication,
let us apply the procedures described above to the sentence:

US: the butler served the drinks


CLFfZ: the butler served the drinks BE RFNT RLTV
T/CLFfPHR: the drinks BE RFNT RLTV + the butler served
SOP: "The drinks were (!) what the butler served."

Although the Focus the drinks is the object of the verb in the under-
lying sentence, it assumes the syntactic role of the subject in the simple
cleft, forcing the verb be to adjust to its number, just as if it were the
grammatical subject of be. The derivation continues as above:

T/BE: BE + the drinks ... RFNT RLTV the butler served


T/RFNT&FF: RFNT RLTV the butler served + BE the drinks
SOP: "What the butler served were (!) the drinks."

Here again, the Focus the drinks is felt to be the grammatical subject.
And finally:

1/IT: IT + BE the drinks RFNT RLTV the butler served


SOP: "It was (!) the drinks that the butler served."
FICHTNER: ENGLISH CLEFT SENTENCES 13

In this case, however, the dummy pronoun IT precedes the finite verb,
and this word, not the Focus, is felt to be its subject. Since IT is
singular in form, the verb was is also in the singular.
A similar adjustment is required in sentences containing reflexive
objects, e.g., I cut myself. In the following demonstration, the reflex-
ive pronoun object myself is the Focus of the cleft:
US: I cut myself
CLFfZ: I cut myself BE RFNT RLTV
T/CLFfPHR: myself BE RFNT RLTV + I cut
SOP: "I (myself) was the one whom I cut."
Here again, the basic SVO structure of English requires an adjustment
in the form of the Focus such that it must "look like" a normal subject
when it precedes the verb. The reflexive pronoun myself must either be
preceded by, or replaced by, the subject pronoun I. The derivation
continues:
T/BE: BE + myself RFNT RLTV I cut
T/RFNT&FF: RFNT RLTV I cut BE myself
SOP: "The one whom I cut was myself."
1/IT: IT + BE myself RFNT RLTV I cut
SOP: "It was (I) myself (whom) I cut."
The possibility of introducing the pronoun I is probably motivated by
the similarity of this sentence to sentences formed in answer to ques-
tions like: (Who was it?) It was I. Here, as elsewhere, it is possible to
delete a relative pronoun object which immediately follows its ante-
cedent.
However, if the subject I becomes the Focus element, then the
person of the reflexive pronoun has to be changed to conform to the
person, number, and gender of RLTV, which is now the subject of cut.
This appears in the following derivation:
US: I cut myself
CLFfZ: I BE RFNT RLTV cut myself
T/CLFfPHR: I BE RFNT RLTV + cut myself
SOP: ''I was the one who cut himself/herself.''
Since Cleftization changes the subject of the verb of the underlying
sentence into the relative pronoun represented by RLTV, the reflexive
pronoun must be of the same person as that of the relative pronoun
subject, i.e., the myself of I cut myself must change into the himself!
herself of (the one) who cut himself/herself. Both the change in the
14 WORD, VOLUME 44, NUMBER I (APRIL, 1993)

number of the verb described earlier and this change in the form of the
object reflexive pronoun represent adaptations which occur in the cleft
sentence to maintain at least a superficial similarity to the morpholog-
ical features of simple sentences in English. 4
T/BE: BE + I . . . RFNT RLTV cut myself
TIRFNT &FF: RFNT RLTV cut myself + BE I
SOP: "The one who cut himself/herself was lime."
Not only does the reflexive pronoun change to the third person mas-
culine or feminine to agree with the subject of the verb cut, i.e., who,
but the case ofthe pronoun I after the verb was changes, too, reflecting
the tendency in English to change any pronoun following a verb into
the object case. The same problems arise in the case of the IT-cleft:
IIIT: IT + BE I RFNT RLTV cut myself
SOP: "It was lime who cut himself/herself."
In the examples of the clefting transformations which have been
examined so far, it was found necessary to adjust the number of the
verb be introduced by Cleftization so that it agrees in number with its
surface subject. The tense of the verb be, on the other hand, has been
modified without comment to correspond to that of the verb in the
underlying sentence. Thus, the sentence the butler served the wine was
transformed into the simple cleft, the butler WAS the one who served
the wine, without any explanation for the change of be into was. In
fact, there are quite specific constraints which govern the assignment
of tense markers to be in the Cleftization process. One of these can be
observed in the difference between the clefted forms of the following
underlying sentences:

(27) the butler IS serving the wine ~


the butler IS the one who IS serving the wine
the one who IS serving the wine IS the butler
it IS the butler who IS serving the wine

(28) the butler WAS serving the win~


the butler WAS the one who WAS serving the wine
the one who WAS serving the wine WAS the butler
it WAS the butler who WAS serving the wine

The principle which governs the extension of the tense markers from
the verb in the underlying sentence to the verb be inserted by Clefti-
FICHTNER: ENGLISH CLEFr SENTENCES 15

zation is very similar to the one which governs the adjustment of the
form of the verb in the reported statement in indirect discourse. Leaving
aside a few exceptions which do not contradict the rule (for which see
QGLS para. 14.32-34, pp. 1026-1033), we can say that a Past Tense
marker of the verb of saying is extended to the verb in the reported
statement; on the other hand, if there is no Past Tense marker in the verb
of saying, then the verb in the reported statement is unchanged. This
is exemplified by the following statements and reports of them:

(29) Mary: "Ann is working at Macy's. "~


Mary says (-Past) that Ann is (-Past) working at Macy' s.
Mary said (+Past) that Ann was (+Past) working at Macy's.

(30) Mary: "Ann has been studying French."~


Mary says (-Past) that Ann has (-Past) been studying
French.
Mary said (+Past) that Ann had (+Past) been studying
French.

In much the same way, a Past Tense marker in the verb in the
sentence underlying a cleft is extended to the verb be, a process which
has been applied in all of the examples presented above.
However, there are other variations in the form of the verb in the
cleft sentence which arise from certain peculiarities of modal auxilia-
ries in English. At first glance, the following two sentences seem
identical in structure:
(31) the butler should have served the wine
(32) the butler must have served the wine
The former means that the butler was under an obligation, presumably
occupational, to serve the wine, an obligation which for some reason
he did not fulfill. While this has the subjective elements of any judge-
ment, it is relatively objective in that it refers to an observable situa-
tion, one not the product of the speaker's mental processes. The latter,
on the other hand, posits no obligation, but is rather the subjective
process of inferring a conclusion from an observed fact, i.e., that since
the wine has been served, it is the butler who is the one most likely to
have performed this service. The use of the modal in the former sen-
tence is sometimes called 'deontic', because, as its Greek root implies,
it expresses a 'binding', a duty or an obligation, which in this case
16 WORD, VOLUME 44, NUMBER I (APRIL, 1993)

devolves upon the butler. The use of the modal in the latter example is
frequently called 'episternic', since the sentence in question reflects
the knowledge of the situation which the person who utters it has
acquired by observation and reasoning.
This distinction between the deontic and epistemic uses of modal
auxiliaries is reflected in the ways in which the various tense markers
of the verbs in underlying sentences are extended to the verb be in-
troduced by Cleftization. In the case of the deontic meaning, only the
Past Tense marker is extended from the verb in the underlying sentence
to be. 5 The following derivation differs in no way from the examples
already presented:
US: the butler should have served the wine
CLFfZ: the butler BE RFNT RLTV should have served the wine
T/CLFfPHR: the butler BE RFNT RLTV + should have served the
wine
SOP: ''The butler was (+Past) the one who should (+Past) have
served the wine.''
T/BE: BE + the butler ... RFNT RLTV should have served the wine
T/RFNT&FF: RFNT RLTV should have served the wine + BE the
butler
SOP: "The one who should (+Past) have served the wine was
(+Past) the butler."
1/IT: IT + BE the butler RFNT RLTV should have served the wine
SOP: "It was (+Past) the butler who should (+Past) have served the
wine."
However, in sentences having an epistemic meaning, not only is the Past
Tense marker extended to the verb be, but all other verb modifications,
including the modal auxiliary, are transferred to it as well. Thus, the
verb forms of the underlying sentence containing the epistemic modal
auxiliary are treated very differently when that sentence is transformed
into the various cleft sentence types, as this derivation shows:
US: the butler must have served the wine
CLFfZ: the butler BE RFNT RLTV must have served the wine
T/CLFfPHR: the butler BE RFNT RLTV + must have served the wine
SOP: ? "The butler was the one who must have served the wine."
Here, only the Past Tense marker has been copied from must to be. Yet
·the sentence sounds awkward. Only when the other verb modifica-
tions, Modal and the Perfect Tense, are transferred to the verb be does
the sentence seem unobjectionable:
FICHTNER: ENGLISH CLEFf SENTENCES 17

SOP: ! ! "The butler must have been the one who served the wine."
Similar changes are required in the case of the WH- and IT-clefts:
T/BE: BE + the butler ... RFNT RLTV must have served the wine
T/RFNT&FF: RFNT RLTV must have served the wine + BE the
butler
SOP: ?? "The one who must have served the wine was the butler."
SOP: ! ! ''The one who served the wine must have been the butler.''
1/IT: IT + BE the butler RFNT RLTV must have served the wine
SOP: ?? ''It was the butler who must have served the wine.''
SOP: ! ! "It must have been the butler who served the wine. " 6
While modal auxiliaries, by their very function, always express a
modal meaning, there are other sentences which do not contain modals
which may also have a modal meaning. When sentences with an
epistemic meaning are subjected to Cleftization, all of the various tense
and mood markers are shifted to the verb be introduced by the clefting
process. The following is just such a case:
US: the butler has by tradition served the wine here
The phrase by tradition suggests here that the speaker is, from his or
her knowledge and experience, affirming a custom of long standing, so
that the sentence is epistemic in character. This is manifested in the
verb forms which appear upon Cleftization:
CLFfZ: the butler BE RFNT RLTV has by tradition served the wine
here
T/CLFfPHR: the butler BE RFNT RLTV + has by tradition served
the wine here
SOP: ? "The butler is the one who has by tradition served the wine
here."
SOP: ! "The butler has been the one who by tradition serves the wine
here.''
Indeed, the sentence would be even more idiomatic if the adverb by
tradition were moved into the verb complex of be along with the verb
modifications, e.g.:
SOP: ! ! "The butler has by tradition been the one who serves the wine
here."
On the other hand, if the adverb by tradition were replaced with,
e.g., always, so that the subjective knowledge implied by the former
18 WORD, VOLUME 44, NUMBER 1 (APRIL, 1993)

were replaced by a simple item of objective information, the sentences


would manifest just the opposite degrees of acceptability:
SOP: ! "The butler is the one who has always served the wine here."
SOP: ? "The butler has been the one who always serves the wine
here."
The same contrasts are apparent in the other cleft sentence patterns, the
WH- and IT-clefts:

SOP: ? ''The one who has by tradition served the wine here is the
butler.''
SOP: ! "The one who serves the wine here has by tradition been the
butler."
SOP: ? "It is the butler who has by tradition served the wine here."
SOP: ! "It has by tradition been the butler who serves the wine here."
Up to this point, the elements which have served as the Focus of
cleft sentences in the examples analyzed have all been noun phrases or
pronouns. Even the agent prepositional phrase by the butler was re-
solved into a simple noun phrase through an additional step in the
series of operations comprising the clefting process. However, there
are other kinds of clause elements which can serve as the Focus of a
cleft sentence, too, such as adverbs, adjectives, infinitives, and de-
pendent finite nominal clauses. The following examples give some
idea of the range of possibilities available to the speaker of English in
selecting a clause element as Focus of a cleft sentence.
The following sentence shows that an adverbial phrase of time can
become the Focus of a cleft:
US: I heard the news this morning
CLFTZ: I heard the news this morning BE RFNT RLTV
T/CLFTPHR: this morning BE RFNT RLTV + I heard the news
SOP: "This morning was (the time) when I heard the news."
The table of surface equivalents above contains an entry showing that
these elements are to be resolved as '' (the time) when'' when the Focus
is an adverb of time. 7
T/BE: BE + this morning ... RFNT RLTV I heard the news
The WH-cleft requires the Topicalization of the Referent and all
forms following it:
FICHTNER: ENGLISH CLEFr SENTENCES 19

T/RFNT&FF: RFNT RLTV I heard the news + BE this morning ...


SOP: "(The time) when I heard the news was this morning."
The IT-cleft is generated as above:
1/IT: IT + BE this morning RFNT RLTV I heard the news
SOP: "It was this morning when I heard the news."
Here, too, the Referent is deleted when it comes to stand immediately
after its antecedent.
An adverb of manner may also serve as the Focus of a cleft,
though the changing syntactic relationships which appear on the sur-
face in the course of the clefting process necessitate some adjustments
in the form of those words. A good example is the one cited by
Bolinger ( 1972:99):
US: he spoke offensively to her
CLFTZ: he spoke offensively BE RFNT RLTV to her
T/CLFTPHR: offensively BE RFNT RLTV + he spoke to her
SOP: ! "Offensively was the way (that) he spoke to her."
T/BE: BE + offensively ... RFNT RLTV he spoke to her
T/RFNT&FF: RFNT RLTV he spoke to her + BE offensively
SOP: ? "The way (that) he spoke to her was offensively."
Even though we know that the word offensively was an adverb in the
underlying sentence, it looks like a predicate adjective here, so that the
adverbial suffix seems out of place. Moreover, even though a strong
case has been made for the notion that the expression the way is not a
noun, but part of a compound relative expression (Bolinger 1972:105),
it nevertheless looks like a subject noun. Consequently, most speakers
of English would prefer the form:
SOP: ! "The way (that) he spoke to her was offensive."
The same considerations apply to the IT-cleft
1/IT: IT + BE offensively the way that he spoke to her
This sequence of elements could be realized in several ways:
SOP 1: ??? "It was offensively the way that he spoke to her."
SOP 2: ? "It was offensively that he spoke to her."
SOP 3: ! "It was offensive the way that he spoke to her."
SOP 4: !? "It was offensive that he spoke to her."
20 WORD, VOLUME 44, NUMBER I (APRIL, 1993)

In the first two of these, the adverbial form seems out of place where
one would expect an adjective. The third example clearly expresses the
idea that his manner of address was offensive. But in the fourth in-
stance, the loss of the noun phrase the way changes both the structure
and the meaning of the sentence to suggest that the mere act of his
speaking to her was the cause of offense. Though a Referent is nor-
mally deleted when it immediately follows the Focus, the phrase the
way is necessary in this case to ensure that the word offensive is
understood as an adverb of manner, and not as an adjective modifying
the noun phrase that he spoke to her.
Not only adverbs, but also adjectives, may serve as the Focus of
a cleft sentence, as in this example: 8
US: Sally is amazing
CLFTZ: Sally is amazing BE RFNT RLTV
T/CLFTPHR: amazing BE RFNT RLTV + Sally is
SOP: "Amazing is what Sally is."
T/BE: BE + amazing .. RFNT RLTV Sally is
T/RFNT&FF: RFNT RLTV Sally is + BE amazing
SOP: ''What Sally is is amazing.''
1/IT: IT + BE amazing RFNT RLTV Sally is
SOP: "It is amazing that Sally is!"
When an adjective is the Focus of a cleft, there is usually a strong
connotation of contrast which is especially noticeable in the case of the
IT-cleft.
In an underlying sentence containing a predicate adjective, the
subject noun can also be the Focus of a cleft, as in the case of the
sentence above:
US: Sally is amazing
CLFTZ: Sally BE RFNT RLTV is amazing
T/CLFTPHR: Sally BE RFNT RLTV + is amazing
SOP: "Sally is the one who is amazing."
T/BE: BE + Sally ... RFNT RLTV is amazing
T/RFNT&FF: RFNT RLTV is amazing + is Sally
SOP: "The one who is amazing is Sally."
1/IT: IT + BE Sally . . . RFNT RLTV is amazing
SOP: "It is Sally who is amazing."
The infinitive which forms part of the verb complex may be de-
tached and serve as the Focus of a cleft, as in this example:
FICHTNER: ENGLISH CLEFr SENTENCES 21

US: she always wanted to visit Rome


CLFfZ: she always wanted to visit Rome BE RFNT RLTV
Here, the infinitive, which is essentially a complex nominal, is the
Focus of the cleft.
T/CLFfPHR: to visit Rome BE RFNT RLTV + she always wanted
SOP: "To visit Rome was what she always wanted (to do)."
The expression to do, which is a surrogate for the topicalized infinitive
clause to visit Rome, is optional in these cases.
T/BE: BE + to visit Rome . . . RFNT RLTV she always wanted
T/RFNT&FF: RFNT RLTV she always wanted + BE to visit Rome
SOP: "What she always wanted (to do) was to visit Rome."
1/IT: IT + BE to visit Rome . . . RFNT RLTV she always wanted
SOP: "It was to visit Rome that she always wanted."
Even a dependent finite nominal clauses can serve as the Focus of
a cleft, although its presence in the clause limits the capacity for other
elements to serve as the Focus of the cleft. When the clause is the
Focus, the derivation is as follows:
US: that she sang off pitch annoyed Tom
CLFfZ: that she sang off pitch BE RFNT RLTV annoyed Tom
Here, the clause that she sang off pitch is in fact a complex noun, and
as such can serve as the Focus of a cleft.
T/CLFfPHR: that she sang off pitch BE RFNT RLTV + annoyed Tom
SOP: "That she sang off pitch was the thing that (~ what) annoyed
Tom."
T/BE: BE + that she sang off pitch ... RFNT RLTV annoyed Tom
T/RFNT&FF: RFNT RLTV annoyed Tom + BE that she sang off
pitch
SOP: "The thing that (~ what) annoyed Tom was that she sang off
pitch."
1/IT: IT + BE that she sang off pitch . . . RFNT RLTV annoyed Tom
SOP: "It was that she sang off pitch that annoyed Tom."
Unlike the examples presented earlier, in which both a subject
noun or object, or a predicate noun or adjective may serve as the Focus
of a cleft, the object noun or pronoun may not do so when the subject
is a dependent finite nominal clause. This condition appears in the
22 WORD, VOLUME 44, NUMBER I (APRIL, 1993)

following derivation, which uses the same underlying sentence as in


the example above:

US: that she sang off pitch annoyed Tom


CLFTZ: that she sang off pitch annoyed Tom BE RFNT RLTV
T/CLFTPHR: Tom BE RFNT RLTV + that she sang off pitch an-
noyed
SOP: ?? ''Tom was the one whom that she sang off pitch annoyed.''
T/BE: BE + Tom .. RFNT RLTV that she sang off pitch annoyed
T/RFNT&FF: RFNT RLTV that she sang off pitch annoyed + BE
Tom
SOP: ?? ''The one whom that she sang off pitch annoyed was Tom.''
1/IT: IT + BE Tom ... RFNT RLTV that she sang off pitch annoyed
SOP: ?? ''It was Tom whom that she sang off pitch annoyed.''

The reason for this apparent exception lies not in the clefting process
itself, but in a constraint in English on dependent finite clauses serving
as the subjects of other dependent finite clauses. This observation is
supported by the fact that the clefting procedures may be applied to this
sentence when Tom is the Focus, provided that the underlying sentence
has first been passivized, as in the following derivation:

US: that she sang off pitch annoyed Tom


PASVZ: Tom was annoyed by (the fact) that she sang off pitch

Without the insertion of the phrase the fact, the derivation leads to
three unacceptable sentences. However, this is again, not a feature of
the clefting process, but a constraint coming from another area of
English structure, namely, the restriction that a dependent finite clause
introduced by that may not function as the object of a preposition, such
as by in this case, but must rather follow a dummy noun like the fact,
so that the finite clause may be construed as standing in apposition to
it. On the other hand, the entire prepositional phrase by the fact may
be omitted, a change which results in three completely idiomatic cleft
sentences:

CLFTZ: Tom BE RFNT RLTV was annoyed (by the fact) that she sang
off pitch BE RFNT RLTV
T/CLFTPHR: Tom BE RFNT RLTV + was annoyed (by the fact) that
she sang off pitch
SOP: ! "Tom was the one who was annoyed (by the fact) that she
sang off pitch."
FICHTNER: ENGLISH CLEF!' SENTENCES 23

T/BE: BE + Tom .. RFNT RLTV was annoyed (by the fact) that she
sang off pitch
T/RFNT&FF: RFNT RLTV was annoyed (by the fact) that she sang
off pitch + BE Tom
SOP: ! "The one who was annoyed (by the fact) that she sang off
pitch was Tom.''
1/IT: IT + BE Tom .. RFNT RLTV was annoyed (by the fact) that she
sang off pitch
SOP: ! "It was Tom who was annoyed (by the fact) that she sang off
pitch."
As these examples show, nouns, adjectives, adverbial expressions,
finite clauses, and infinitives can all serve as the Focus of a cleft
sentence.
We turn now to the third of the three questions raised at the
beginning of this article, namely, the communicative function of these
sentence patterns. This has been the subject of several studies in the
recent past, notably Prince (1978) and Declerck (1983 and 1984). Both
of these linguists approach the problem of the way in which clefts are
used by examining a number of passages in which these sentence
patterns appear, and then trying to infer from those citations what
semantic features of the context motivated their authors to use cleft
sentences, and hence what specific contribution the cleft in question
made to those sentences.
One matter must be settled at the start: what sentence patterns are
to be included in the category of cleft sentences. Prince recognizes
only WH-clefts and IT-clefts, as does Declerck, although he refers in
passing at one point to simple clefts as "inverted WH-clefts" (De-
clerck 1984:259). At the beginning ofthis study, we placed all three of
these sentence patterns in the category of cleft sentences, a classifica-
tion which is supported by the fact that all three have been found to be
the output of the same set of operations, though emerging at different
stages of those operations. This being the case, it might be more useful
to approach the question of their meaning and function, first by noting
their formal properties, and then proceeding to a discussion of their
function in various contexts. Just as each has a distinctive form, it also
has a distinctive mode of predication by virtue of by that form. The
three classes which our procedures generated are:

(33) simple clefts (definitions)


the wine was what the butler served
24 WORD, VOLUME 44, NUMBER I (APRIL, 1993)

the wine was what was served by the butler


the butler was the one who/that served the wine
the butler was the one by whom the wine was served

(34) WH-clefts (specifications)


what the butler served was the wine
what was served by the butler was the wine
the one who served the wine was the butler
the one by whom the wine was served was the butler

(35) IT-clefts (affirmations)


it was the wine that the butler served
it was the wine that was served by the butler
it was the butler who/that served the wine
it was the butler by whom the wine was served

As the captions in the table above suggest, each kind of cleft has,
by virtue of its form, a potential for a certain mode of expression.
Simple clefts are essentially a species of definition, in that they provide
a description of what a term stands for in a given context, e.g., wine
is obviously a fermented beverage made from grapes, but in the hy-
pothetical context suggested by the example, it is what was served by
the butler. Similarly, the task of the WH-cleft is to identify the person
who or thing which fulfills the specifications set forth in the WH-
clause, e.g., the butler may have many duties, but the beverage he
served on this occasion was the wine. In form, the two clefts are mirror
images of one another, in that the single term in the simple cleft is
defined by a clause following the form of the verb be, whereas the
WH-clause of the WH-cleft is followed by a form of be and the term to
which the clause applies. In terms of form, the IT-cleft is in a class by
itself, particularly in that the verb be does not serve as the fulcrum of
a balance between the term to be defined or specified, on the one hand,
and the defining and specifying clause on the other, but along with the
empty pronoun it provides a neutral introduction to the Focus of the
cleft, its most important term, for which reason we label the mode of
predication of the IT-cleft as affirmation.
These modes of predication may have varying discourse functions
in the texts in which they appear. As Prince (1978) and Bromser (1984)
have pointed out, the specific functions which the various clefts may
have depend upon the distribution of old and new information. As
Jones and Jones have recently shown, one of the main discourse func-
FICHTNER: ENGLISH CLEFf SENTENCES 25

tions for which cleft sentences are used is to highlight the theme of a
given segment of discourse (1985:10f.). Cleft sentences have this ef-
fect, because the clefting process adds a feature to the underlying
sentence which that sentence lacks, i.e., polarity.
What the clefting process does is to express this feature overtly in
the surface structure of the sentence, not merely by means of intonation
patterns, which can, of course, be inferred from the context, but per-
ceived only in actual speech. Not only does the clefting process allow
the speaker or writer to place certain words or phrases at the crucial
points at the beginning or end of the sentence, thus overcoming the
limitations imposed by the rigid SVO word order of English, but it also
divides the sentence into two zones, which makes it impossible for the
reader to interpret it as a simple declaration of fact in which the various
nouns and adverbs are understood merely as participating in the action
represented by the verb. 9 The polarization of the sentence-its sepa-
ration into two zones-implies opposition, and opposition, particu-
larly between elements in the same structure, frequently suggests con-
trast. Moreover, the designation of a specific element as Focus of the
cleft draws attention to that element, and increases the hearer's aware-
ness of any difference between the information presented in the cleft
sentence and that in or suggested by the context.
Having seen how cleft sentences make it possible for the speaker
or writer to escape the limitations of the rigid SVO structure of En-
glish, it is easy to offer an answer to the fourth question raised at the
beginning of this article, namely, where cleft sentences may be situ-
ated in the larger set of syntactic structures of English. Most of those
who have studied these sentence patterns since about 1960-Lees
(1963), Higgins (1971), Gundel (1976 and 1977), Wirth (1978),
Delahunty (1981), Bromser (1984), and Knowles (1986)-regard the
cleft sentence as originating on some level of deep or underlying
structure. The remainder either take no position as to where cleft
sentences should be placed, or, like Jespersen (1965, 1969), treat them
as modifications of existing or possible English sentences.
Several considerations compel us to return to Jespersen's assump-
tion, if not to his explanation. For one thing, it has been amply dem-
onstrated that it is at least possible to generate all of the possible
correctly formed cleft sentences, and only the correct ones, by the
series of operations which we have performed on an underlying struc-
ture which is itself an acceptable English sentence. Thus, there is no
need to refer to deep structure, however that may be understood, in
order to explain how cleft sentences are formed. Moreover, it was
26 WORD, VOLUME 44, NUMBER I (APRIL, 1993)

noted that Cleftization could be applied to underlying sentences in the


Passive voice. While it is true that Passivization is a change which
takes place below the level of surface structure, it is also true that it,
like all verb modifications, is fairly close to surface structure. In any
case, Passivization is clearly prior to Cleftization.
It is also important to note that cleft sentences are not the only
sentence types which have arisen in English in order to introduce some
flexibility into the limited range of possibilities afforded by its SVO
structure. There is a whole series of what have been called Topic-
Comment Redistribution (TCR) transformations (Fichtner 1979:
164ff.), which also operate on existing sentences. While these are not
always placed in the same category, all will be recognized by speakers
of English in the examples which follow:

(36) specific TCR-transformations:


with one-word forms of be: the uses of adversity are
sweet ~ sweet are the uses of adversity
with one-word forms of lexical verbs: the ship went
down ~ down went the ship
with verb forms containing one or more auxiliary verbs: a man
in a blue suit was standing in the doorway ~ in the doorway was
standing a man in a blue suit I standing in the doorway was a man
in a blue suit

All of these sentences can be produced by the same kinds of changes


applied above to simple underlying sentences to form cleft sentences.
In addition, there are a number of sentence patterns which can be
formed by adjustments in the position of certain elements, to which are
then added the word THERE. The following sentence pairs illustrate
this capacity of English:

(37) THERE-transformations:
many prominent people will be at the reception ~
THERE will be many prominent people at the reception
a large number of cases of AIDS have been reported this year
~ THERE have been a large number of cases of AIDS
reported this year I THERE have been reported a large number
of cases of AIDS this year
FICHTNER: ENGLISH CLEFT SENTENCES 27

Another group of transformations involves the Commentization of a


subject clause-a dependent finite clause or an infinitive-and the
insertion of the element IT:

(38) IT-transformations:
that this could happen is unbelievable ~ IT is unbelievable
that this could happen
to refute this theory is difficult ~ IT is difficult to refute this
theory

Like cleft sentences, all of the TCR-, THERE-, and IT-transformation


procedures take simple underlying sentences as the point of departure
in generating them.
And finally, all of these sentence patterns, including cleft sen-
tences, are subject to the modifications which can be applied to simple
sentences, such as the formation of yes-no and information questions,
relative clauses, etc. For example, the last sentence above:

(39) IT is difficult to refute this theory

could be transformed into:

(40) a yes-no question: is IT ... difficult to refute this theory?

(41) an information question: WHAT is ... difficult to refute?

(42) a relative clause: (that is a theory) WHICH IT is difficult to


refute ...

In like fashion, the cleft sentences generated earlier can all undergo
these transformations, too, e.g., into:

(43) a yes-no question: was the butler ... the one who served the
wine?

(44) an information question: WHO was the one who served the
wine?

(45) a relative clause: (the butler,) who was the one who served the
wine, (has retired)
28 WORD, VOLUME 44, NUMBER 1 (APRIL, 1993)

All of these considerations suggest that the clefting process is, in


effect, an additional kind of TCR-transformation, one which forms its
sentences by adjustments in the form and position of certain elements
in simple underlying sentences. Cleft sentences share with specific
TCR-, THERE-, and IT-transformations the potential for reshaping the
underlying sentence so as to enable it to conform more closely to the
sequence of ideas in the text in which they appear. They form a sharply
delineated group of three sentence patterns, each produced at a specific
point in the series of operations, so that it is entirely appropriate to
regard the three-simple clefts, WH-clefts, and IT-clefts-as compris-
ing the category of cleft sentences in English.
Department of Germanic, Slavic, and East European Languages
Queens College, City University of New York
Flushing NY 11367-0904 I USA

ENDNOTES

1
This article is a revised and expanded version of a lecture presented at a monthly meeting
of the International Linguistic Association held on 9 May 1987.
2
This table is adapted from Bolinger (1972:105).
3
The optional omission of the object whom here is another instance of the possibility in
English of omitting relative pronouns which immediately follow their antecedent, e.g., that was
the man (whom) I met at the conference.
4
Akmajian (1970) considers at some length the possibility of leaving the reflexive pronoun
in the first person, i.e., a sentence like I was the one who cut myself It is important to keep in
mind that a language is a system composed of many subsystems, and even though features such
as subject-verb agreement and subject-reflexive agreement are certainly relevant to the produc-
tion of correctly formed cleft sentences, these are not integral parts of the clefting process, as
Akmajian seems to imply. Sentences like this are probably syntactic conflations of pairs like I cut
myself and I was the one who cut himself, where the speaker's knowledge about the persons on
whom the cut has been inflicted, on the one hand, and his/her sense that a third person reflexive
object is required with a third person subject, on the other, are brought into conflict by the shift
of person introduced by the clefting process.
5
Following Fichtner (1979: 115), the form should is regarded as formed by the combination
of the verb modifications MODAL plus PAST plus SUBJ (subjunctive) plus PSNM (person and
number).
6
Although must is a defective verb in English, it is interpreted in this sentence as being in
the Past Tense on the basis of the equivalent expression, i.e., the butler had to have served the
wine. The general distinction has arisen in English today that have to is deontic, whereas must is
epistemic. When must is used in the Present Tense in American English, it is usually epistemic,
whereas in British English it is often deontic. This can lead to alternative cleft sentence patterns,
e.g., the sentence the butler must serve the wine could form either a deontic cleft, the butler is
the one who must, i.e., has to, serve the wine or an epistemic one, the butler is the one who
habitually serves the wine. On the other hand, the sentence, the butler was the one who had to
serve the wine is incontestably deontic, whereas its counterpart, the butler had to have been the
one who served the wine would be interpreted by most speakers as epistemic.
FICHTNER: ENGLISH CLEFf SENTENCES 29
7
For adverbs of place, the corresponding forms would be "(the place) where", and for
adverbs of manner, "the way in which", of which the reduced form would be "how".
8
Poldauf (1969:20) claims that a 'depictive attribute' cannot serve as the Focus of a cleft, and
cites the following sentence to support his view:
? it was black that he drank his coffee
While this sentence is admittedly questionable, the facts are not so simple. If we take as the
underlying sentence:
he drank his coffee black
and subject it to the clefting procedures used heretofore in this article, we find that an acceptable
simple cleft and WH-cleft can in fact be generated:
CLFfZ: he drank his coffee black BE RFNT RLTV
T/CLFfPHR: black BE RFNT RLTV + he drank his coffee
SOP: ! "Black was the way (that) he drank his coffee."
T/BE: BE + black ... RFNT RLTV he drank his coffee
T/RFNT&FF: RFNT RLTV he drank his coffee + BE black
SOP: ! "The way (that) he drank his coffee was black."
It is true, however, that the IT-cleft is not entirely acceptable:
IIIT: IT + BE black .. RFNT RLTV he drank his coffee
SOP: ? "It was black that he drank his coffee."
What has happened here is an overlap of two structures of the language. The word black in the
cleft sentences is not an adjective, but an adverb of manner-it tells HOW he drank his coffee.
In the simple and WH -cleft, its adverbial character is clear from the presence of the phrase the
way, i.e., black was the way ... in the simple cleft, and the way ... was black in the WH-cleft.
However, in the phrase IT was black, the most likely interpretation of the syntactic role of black
by the typical speaker of English is as a predicate adjective, and it is this forced re-interpretation
of the syntactic role of the word as predicate adjective that makes the IT-cleft seem ill-formed.
9
0n the SVO word order of English, cf. the recent work by Giv6n (1984-1990:190).

REFERENCES

Akmajian, Adrian. 1970. "On deriving cleft sentences from pseudo-cleft sentences." Linguistic
inquiry 1:149-168.
Bolinger, Dwight. 1972. "A look at equations and cleft sentences." In: Studies for Einar
Haugen. Ed. Evelyn S. Firchow eta!. Janua linguarum: Series maior, 59. The Hague, Paris:
Mouton. Pp. 96-114.
Briimser, B. 1984. "Towards a functional description of cleft constructions." Lingua 62:325-
48.
Declerck, Renaat. 1983. "Predicational clefts." Lingua 61:9-45.
- - - . 1984. "The pragmatics of it-clefts and WH-clefts." Lingua 64:251-89.
Delahunty, Gerald Patrick. 1984. "The analysis of cleft sentences." Linguistic analysis 13:
63-113.
Fichtner, Edward G. 1979. English and German syntax: A contrastive analysis on generative-
tagmemic principles. Miinchen: Wilhelm Fink Verlag.
Giv6n, Talmy. 1984-1990. Syntax: Afunctional-typological introduction. 2 vols. Amsterdam,
Philadelphia: Benjarnins.
Gundel, Jeanette K. 1976. On the source of "it" in cleft sentences. Bloomington IN: Indiana
Univ Linguistics Club.
- - - . 1977. "Where do cleft sentences come from?" Language 53:543-59.
30 WORD, VOLUME 44, NUMBER 1 (APRIL, 1993)

Hankamer, Jorge. 1974. "On the non-cyclical nature of WH-clefting." Proceedings of the lOth
regional meeting, Chicago linguistic society. Pp. 221-233.
Higgins, F. R. 1971. "The pseudo-cleft construction in English." Diss., Massachusetts Institute
of Technology.
Jacobs, Roderick J. & PeterS. Rosenbaum. 1968. English transformational grammar. Waltham
MA: Blaisdell.
Jespersen, Otto. 1965. A modern English grammar on historical principles. Vol III: Syntax.
London: Allen & Unwin: Copenhagen: Munksgaard. (Rpt. of ed. of 1927.)
- - - . 1969. Analytic syntax. Transatlantic series in linguistics. New York: Holt, Rinehart &
Winston. (Rpt. of ed. of 1937.)
Jones, Larry B. & Linda K. Jones. 1985. "Discourse functions of five English sentence types."
WORD 36:1-21.
Knowles, John. 1986. "The cleft sentence: A base-generated perspective." Lingua 69:295-317.
Lees, R. B. 1963. "Analysis of the 'cleft sentence' in English." Zeitschrift fUr Phonetik,
Sprachwissenschaft und Kommunikationsforschung 16:371-388.
Pinkham, Jessie, & Jorge Hankamer. 1975. "Deep and shallow clefts." Proceedings of the lith
regional meeting, Chicago linguistic society. Pp. 429-450.
Poldauf, Ivan. 1969. "The so-called medio-passive in English." Prague studies in English
13:15-34.
Prince, Ellen F. 1978. "A comparison of WH-clefts and it-clefts in discourse." Language
54:883-906.
QGLS = Randolph Quirk, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. 1985. A
comprehensive grammar of the English language. London, New York: Longman.
Wirth, Jessica. 1978. "The derivation of cleft sentences in English." Glossa 12:58-82.

Vous aimerez peut-être aussi