Comparative Studies in Brown, Frown, LOB and FLOB Corpora: the Changes in the Use of Should and Present Subjunctive in That-Clauses

Hiroshi Suga

Takahashi Technical High School

Abstract

This paper is intended to investigate that a usage which characterized AmE and BrE in the 1960s can be valid in the 1990s. It focuses on the use of should and present subjunctive in that-clauses after assertive verbs or adjectives. In the 1960s the preference of should in BrE and present subjunctive in AmE was remarkable in that-clauses. However, with the help of the four corpora: Brown, Frown, LOB and FLOB, we are able to show that this preference has been weakened in the 1990s. We also prove that present subjunctive which went out of fashion a long time ago has revived in BrE in the 1990s. Furthermore, we show that some main clause verbs strongly affect the choice of should or subjunctive in that-clauses.

1 Introduction

With the development of information technology, more and more communication is being done on the Internet. Amidst this social change, the English language has also become international, losing its local features. In other words, the clear differences once seen between some aspects of BrE and AmE have now become fuzzy.

In the 1960s the Brown Corpus was constructed collecting American English and its British counterpart, the LOB Corpus, was made in the 1970s. In the 1990s, both the Frown Corpus of American English and the FLOB Corpus of British English were built. These four corpora were made in the same structure in order to make comparison easier: informative prose of approximately 750,000 words and imaginative prose of around 250,000 words. These four corpora have enabled us to study how English has changed in the last 30 years.

30 years might not be long enough to see significant change in a language. However, some syntactic changes can occur. For instance, Hundt (1997) has analyzed these four corpora and revealed that the frequency of use of the relative pronoun which has decreased both in BrE and AmE. If we focus on a certain syntactic expression, we might find some unexpected facts. Even if they are small changes, they might become larger in a century.

What I wish to show in this paper is the change of the use of should and present subjunctive in that-clauses in BrE and AmE in the 1960s and 1990s. It has been said that when that-clauses are led by verbs or adjectives expressing the importance of action (e.g. insist, recommend, important, necessary), should is mainly used in BrE and present subjunctive in AmE. Our concern is to examine whether this syntactic feature is still marked in today’s English or not.

2 Present Subjunctive in BrE and AmE

Historically speaking, subjunctives were very common in Britain from OE to around 1600. With the appearance of the structure in which auxiliary verbs were used, however, subjunctive including present subjunctive gradually declined. In consequence, subjunctive was seldom seen from 1600 to 1900 in Britain. Mandative subjunctive was not required so much since the meaning of wish or demand was already included in main verbs. This might have aggravated the decline of subjunctive.

There are many cases in which expressions which have now fallen into desuetude in Britain have survived in the New Continent¹. Subjunctive had immigrated to America over the Atlantic Ocean before auxiliary verbs dominated, and thus has lived on.

Indeed in Modern English subjunctive is not as influential as it once was, but it is still used. Berk(1999:149) refers to the fact that since subjunctive is usually learned through school education or reading, those with an academic background use it most, and that the modern subjunctive expresses a variety of deontic meanings.

Quirk et al. (1985:1013) describes that putative should is far more common in BrE than in AmE, while the present subjunctive occurs more frequently in AmE than in BrE.

The employees have demanded that the manager

resign. <esp AmE>

should resign. <esp BrE>

resigns. <esp BrE> (ibd., p.157)

Other grammarians like Swan(1995) and Declark(1991) also have the same description.

3 Aim and Method

________________

¹ For instance, the expression “I guess” for “I think” was used in Britain in the 14 c. However, it had gone out of fashion until the 17c. In North America, on the other hand, this expression is still very common.

It is worth while to examine, using Brown, Frown, LOB and FLOB, whether the tendency mentioned by Quirk et al. really exists in BrE and AmE, and if it does, whether there is some change in its use from the 1960s to 1990s.

First, we selected the following 33 suasive verbs and 9 adjectives which have possibility to lead a that-clause with mandative subjunctive.

(Verbs)

advise, agree, appeal, arrange, ask, beg, care, command, decide, decree, demand, desire, determine, dictate, direct, expect, fear, hint, insist, move, offer, order, permit, plan, pray, propose, provide, recommend, request, require, suggest, take, urge

(Adjectives)

advisable, desirable, anxious, urgent, careful, essential, imperative, important, necessary

Next, sentences which have a that-clause within 5 words after these words² were chosen from the four corpora, and classified into three groups: ‘present subjunctive’, ‘should’ and ‘other including indicative’. The auxiliary verbs except should were classified in ‘other’ group. In this classification the sentences as follows were excluded.

1) Sentences in which the tense of the verbs in subordinate clauses is older

than that of the matrix verbs³

ex. Some teachers, parents, and students - both black and white –

insist that the signs of trouble were already there at the high

school. (Frown)

2) Sentences in which it is difficult to distinguish present subjunctive from

indicative mood because the tense of the main clause is present and the subject is plural or you⁴

ex. (…) suggests that many women read Behn’s work. (FLOB)

________________

² If it is a verb, its inflections are also counted.

³ Greenbaum et al. (1988:650) state that should is commonly used in that-clauses after expressions that express a future necessity, plan, or intention or that convey an emotional reaction to a future event. Generally speaking, if the tense of the verbs in the subordinate clause is older than that of the verbs in main clauses, it is apt to express not a speaker’s opinion, but a fact.

⁴ Certainly one can easily differentiate subjunctive from indicative if it is read carefully in the context. However, it is not practical when you deal with a large quantity of data.

3) Sentences which have a that-clause led by nouns deprived from the

assertive verbs⁵

ex. (…) the requirement that they be transcended by faith. (Frown)

4 ‘Present Subjunctive’, ‘Should’ and ‘Other including Indicative’ in the Four Corpora

The frequencies of these three categories of 33 verbs and 9 adjectives are listed in appendix. Overall frequencies are shown in Table 1. The total numbers don’t show many differences but we can find some interesting facts in each list.

Table 1.

The frequencies of subjunctive, should and other verbal forms after certain verbs and adjectives expressing the idea that an action is necessary

(tripled numbers given in brackets to compare with Informative Prose)

	Present Subjunctive	Should	Other including Indicative	Total
Brown Informative Prose Imaginative Prose	101 90 11 (33)	21 17 4 (12)	148 133 15 (45)	270
Frown Informative Prose Imaginative Prose	78 59 19 (57)	18 16 2 (6)	190 170 20 (60)	286
LOB Informative Prose Imaginative Prose	15 12 3 (9)	105 98 7 (21)	216 183 33 (99)	336
FLOB Informative Prose Imaginative Prose	36 32 4 (12)	55 48 7 (21)	219 198 21 (63)	310

________________

⁵ It is desirable to include this type of sentences in order to get more data. In this paper, however, these are omitted in order to reduce the number of words searched by one half.

4.1. Differences of the Use of Subjunctive in BrE and AmE

Figure 1.

Frequencies of Subjunctive, Should and Other in That-clauses in Brown and LOB

Figure 1 indicates that, in the 1960s, the use of the present subjunctive was far more remarkable in AmE than in BrE, while should was more used in BrE than in AmE. Subjunctive is used in AmE as often as should is used in BrE. ‘In Am.E. (…) the subjunctive’, Declark(1991:354) says, ‘is quite common in written language, especially in journalese.’ Table 1 agrees with this statement. The amount of original data of informative prose in each corpus is roughly three times larger than the amount of imaginative prose. Even while taking this into consideration, the frequency of present subjunctive in informative prose is still higher than that in imaginative prose in the Brown corpus.

In the Frown corpus, subjunctive appears 59 times in informative prose and 19 times in imaginative prose. If you adjust the number in imaginative prose by tripling 19, you get 57, which is close to 59. We may say that, in the use of subjunctive in AmE, the difference between resisters is getting minor.

Figure 2.

Frequencies of Subjunctive, Should and Other in That-clauses in

Frown and FLOB

How do these features seen in the 1960s change in the 1990s? Figure 2 shows the distribution of present subjunctive, should and other verbal use in that-clauses. Subjunctive is still mainly used in AmE and should is in BrE. The frequencies of subjunctive in AmE and should in BrE have decreased by roughly 25% and 50% respectively. The frequency of should in AmE is still low, while the frequency of subjunctive in BrE has doubled. Quirk et al.(1985:1013) refer to the fact that the use of the present subjunctive seems to be increasing in BrE in that-clauses. ‘This subjunctive’, Chalker et al. (1994:381) also say, ‘has made a considerable comeback in British English recent years, probably under American influence.’ Figure 2 justifies this assertion.

4.2. Present Subjunctive in BrE

Quirk et al. (1985:157) remark that there is a tendency in BrE to choose the subjunctive particularly when the finite verb is be. Fowler(1996:747) also states that the subjunctive is increasingly found in BrE., showing its example as follows:

Fundamentalist Islam … decrees that men and women be strictly segregated. (Listener, 1988)

In order to verify this statement, we have chosen the sentences in which present subjunctive be is used in a that-clause from the data in Table 1. The result is shown in Table 2.

Table 2. Frequencies of the present subjunctive be in that-

clauses

(tripled numbers given in brackets to compare with Informative Prose)

Corpus		Present Subjunctive be
Brown	Informative Prose Imaginative Prose	49 4(12)	53
Frown	Informative Prose Imaginative Prose	32 1(3)	33
LOB	Informative Prose Imaginative Prose	8 1(3)	9
FLOB	Informative Prose Imaginative Prose	18 2(6)	20

It is clear that the present subjunctive be is still used a lot in informative prose in Brown and Frown (See Table 2). On the other hand, the frequency of present subjunctive be in BrE is not as high as in AmE. We should not, however, overlook the fact that be is found in only 9 cases in LOB, but 20 cases in FLOB. This also supports the revival of present subjunctive in BrE which we have verified in 4.1.

4.3. Collocation Between Predicators and Modality in That-clauses

Quirk (1995:136-139) suspects that the choice of VP is influenced by whether the subject of the dependent clause was willing or reluctant to undertake the proposed action. He conducted an experiment in which he showed these two sentences and asked the examinees to fill the blanks with go, should go, or went.

He wanted to see the play, so I suggested that he …….

He was very reluctant to leave, but I suggested that he …….

The result of this experiment was that when the subject of the dependent clause was willing, most chose should go, and when the subject was reluctant, most chose went. There was no meaningful difference in the number of people who chose go. When the verb suggested was changed into recommended, this tendency was strengthened.

Then, isn’t it possible that the matrix verb itself affects on the choice of the verb in the dependent clause? Table 3 shows that some main clause can influence on the frequency of present subjunctive or should in the dependent clause. The verb require collocates well with present subjunctive both in Brown and Frown, while recommend is often used with should both in LOB and FLOB. Since both of the verbs show strong assertion, speakers seem to choose present subjunctive or should rather than indicative mood when they want to assert their opinion strongly.

Table 3 Collocation between the verbs of the main clauses and

the verbs of the dependent clauses

	Predicator	Present Subjunctive	Should
Brown	demand recommend require	18 10 9	0 0 0
Frown	require insist demand	11 10 9	0 0 0
LOB	recommend	1	10
FLOB	recommend	2	16

5. Conclusion

In this research, suggest has the greatest frequency among 33 verbs which have a that-clause within 5 words after them. In Brown, suggest including its inflections appears 49 times, and 33 cases are indicative or other. When suggest is used in the meaning of ‘to propose’, mandative subjunctive or should might be used in a following that-clause. In most cases, however, suggest means ‘to indicate’, which leads to indicative mood in a that-clause. You can find whether the indicative mood is just the expression of a fact or the substitute of subjunctive only when you put the sentence back in the original context and read it through. However, the fuller study of these cases lies outside the scope of this paper.

At any rate, using the four corpora, we have shown that the use of present subjunctive in that-clauses was highly remarkable in written AmE in the 1960s, while the modal auxiliary should was conspicuous in BrE in the same period. Though this tendency can still be observed in the 1990s in both languages, it is not as remarkable as it used to be in the 1960s. Namely, the frequency of present subjunctive in AmE and the use of should in BrE have decreased. On the other hand, present subjunctive seems to have revived in BrE. The data of these corpora indicate that present subjunctive is not a hackneyed and old-fashioned expression but an active and useful expressive function in written English today. We have also observed that some matrix verbs, such as require and recommend, have strong influence on the choice of VP in a that-clause.

With the development of new information technology, especially the appearance and the tremendous progress of the Internet, this type of variations in some syntactical expressions and the mobility of languages will be more and more noticeable.

Appendix

Brown

Search Words	P. Subjunctive			Should			Other			Total
	Inf	Ima	Total	Inf	Ima	Total	Inf	Ima	Total	Inf	Ima	Total
advise	2		2	1		1	4		4	7		7
agree				3		3	20	2	22	23	2	25
appeal
arrange								1	1		1	1
ask	4	1	5				1		1	5	1	6
beg
care								1	1		1	1
command
decide				1		1	18	2	20	19	2	21
decree							2	1	3	2	1	3
demand	16	2	18							16	2	18
desire				1		1				1		1
determine	1		1	2		2	3	1	4	6	1	7
dictate
direct	2		2							2		2
expect	1		1	1		1	16	2	18	18	2	20
fear					1	1	8		8	8	1	9
hint							1		1	1		1
insist	6	2	8				11	4	15	17	6	23
move
offer
order		2	2	1		1				1	2	3
permit
plan
pray							2	1	3	2	1	3
propose	7		7	1		1				8		8
provide	3		3				6		6	9		9
recommend	10		10				1		1	11		11
request	6		6							6		6
require	9		9				2		2	11		11
suggest	9	3	12	3	1	4	33		33	45	4	49
take
urge	4		4				1		1	5		5
advisable
desirable
anxious	1		1							1		1
urgent
careful
essential	2		2		1	1				2	1	3
imperative
important	4	1	5	2	1	3	3		3	9	2	11
necessary	3		3	1		1	1		1	5		5
Total	90	11	101	17	4	21	133	15	148	240	30	270

LOB

Search Words	P. Subjunctive			Should			Other			Total
	Inf	Ima	Total	Inf	Ima	Total	Inf	Ima	Total	Inf	Ima	Total
advise	1		1	2		2		1	1	3	1	4
agree	1	1	2	8	2	10	25	8	33	34	11	45
appeal
arrange				1		1	2		2	3		3
ask	1		1	2		2				3		3
beg
care
command							1		1	1		1
decide				3		3	21	8	29	24	8	32
decree
demand	2		2	3		3	2		2	7		7
desire				1		1	1		1	2		2
determine				1		1	3		3	4		4
dictate				1		1				1		1
direct				1		1	2		2	3		3
expect							6		6	6		6
fear							11	2	13	11	2	13
hint							1	2	3	1	2	3
insist	1		1	6	2	8	5		5	12	2	14
move
offer
order		1	1	1		1	4		4	5	1	6
permit
plan				1		1				1		1
pray	1		1	1		1	2	3	5	4	3	7
propose				4		4	1		1	5		5
provide				3		3	12	2	14	15	2	17
recommend	1		1	10		10	1		1	12		12
request	1		1	2		2				3		3
require				2		2				2		2
suggest	3	1	4	21	1	22	73	6	84	97	8	105
take				2			3	1	4	5	1	6
urge				2			3		3	5		5
advisable
desirable				1						1		1
anxious				2						2		2
urgent
careful
essential				6	1		2		2	8	1	9
imperative
important				8			1		1	9		9
necessary				3	1		1		1	4	1	5
Total	12	3	15	98	7		183	33	216	293	43	336

Frown

Search Words	P. Subjunctive			Should			Other			Total
	Inf	Ima	Total	Inf	Ima	Total	Inf	Ima	Total	Inf	Ima	Total
advise
agree				4		4	24	2	26	28	2	30
appeal
arrange
ask		3	3				1	1	2	1	4	5
beg		1	1								1	1
care								2	2		2	2
command
decide				2		2	10	4	14	12	4	16
decree				1		1				1		1
demand	6	3	9				2		2	8	3	11
desire	1		1				1		1	2		2
determine
dictate
direct							1		1	1		1
expect
fear							19	4	23	19	4	23
hint
insist	3	7	10				16	2	18	19	9	28
move
offer
order	4		4		1	1				4	1	5
permit
plan
pray							1		1	1		1
propose	2		2	1		1	3		3	6		6
provide							12		12	12		12
recommend	7		7	1		2				9		9
request	6		6							6		6
require	11		11				2		2	13		13
suggest	9	4	13	5	1	6	74	3	77	88	8	96
take				1		1	2	1	3	3	1	4
urge	3		3							3		3
advisable
desirable
anxious
urgent								1	1		1	1
careful
essential
imperative	2	1	3							2	1	3
important	5		5				2		2	7		7
necessary
Total	59	19	78	16	2	18	170	20	190	245	41	286

FLOB

Search Words	P. Subjunctive			Should			Other			Total
	Inf	Ima	Total	Inf	Ima	Total	Inf	Ima	Total	Inf	Ima	Total
advise				2		2	2		2	4		4
agree				2		2	18	1	19	20	1	21
appeal
arrange								1	1		1	1
ask	1		1	2		2		1	1	3	1	4
beg
care							1		1	1		1
command
decide				5	1	6	13	8	21	18	9	27
decree					1	1		1	1		2	2
demand	4	1	5	5	1	6	3		3	12	2	14
desire								1	1		1	1
determine
dictate	1		1				2		2	3		3
direct
expect
fear							8	3	11	8	3	11
hint
insist	2	1	3	1		1	16		16	19	1	20
move
offer
order	4		4	1	2	3				5	2	7
permit
plan
pray							1	1	2	1	1	2
propose	1		1	2		2	4		4	7		7
provide	1		1				15		15	16		16
recommend	2		2	16		16	1		1	19		19
request	4	1	5		1	1				4	2	6
require	6		6	2		2	4		4	12		12
suggest	2	1	3	5		5	96	3	99	103	5	108
take							2		2	2		2
urge	1		1	1		1	1	1	2	3	1	4
advisable
desirable	1		1							1		1
anxious
urgent
careful
essential				1		1	3		3	4		4
imperative
important	1		1	3	1	4	8		8	12		12
necessary	1		1							1		1
Total	32	4	36	48	7	55	198	21	219	278	32	310

References

Berk, L. M. (1999), English Syntax:From Word To Discourse. New York: Oxford University Press.

Biber, D., S. Johansson, G. Leech, S. Conrad and E. Finegan. (1999), Longman Grammar of Spoken and Written English. London: Longman.

Chalker, S., and E. Weiner. (1994), The Oxford Dictionary of English Grammar. Oxford: Oxford University Press.

Declerck, R. (1991), A Comprehensive Descriptive Grammar of English. Tokyo: Kaitakusha.

Fowler, H.W. (1996), The New Fowler’s Modern English Usage, 3rd ed edited by R. W. Burchfield. Oxford: Oxford University Press.

Greenbaum, S., and J. Whitcut. (1988), Longman Guide To English Usage. London: Longman.

Hundt, M. (1997), ‘Has BrE Been Catching Up with AmE Over the Past Thirty Years?’ In Corpus-based Studies in English: Papers from the seventeenth International Conference on English Language Researsh on Computerized Corpora, Stockholm (1996), ed. by Magnus Ljung. 135-151. Amsterdam; Rodopi.

McCawley, J. D. (1998), The Syntactic Phenomena of English 2nd ed. Chicago: The University of Chicago Press.

Quirk, R. (1995), Grammatical & Lexical Variance in English. London: Longman.

Quirk, R., S. Greenbaum, G. Leech and J. Svartvik. (1985), A Comprehensive Grammar of the English Language. London: Longman.

Swan, M. (1995), Practical English Usage 2nd ed. Oxford: Oxford University Press.