4 Overall survey of the data

4.1 Results

The texts yielded 498 examples, with a be/have ratio of circa 4:1 (397:101).

When comparing the various text types, the striking fact is that the letters have a considerably higher have ratio than the other types. The Lyme letter collection is only slightly more in favour of have than average, while the two other letter writers show considerably higher figures than both the synchronic and the diachronic average (i.e. the average of the decade(s) they wrote in vs. that of the whole period. See Table 3). Common to the letters is that they were written by well-educated people with a high status. All letters were also meant to be read by a second party. No significant statistical difference in the usage of have can be found between the male and female letter writers.

The comedy authors show some individual differences worth noting. Sir Edmund Gosse writes as follows about the various dramatists in his introduction to Restoration Comedies (Everyman's Library 1932):

Sir George Etheredge [...] tried with insufficient energy to raise his audiences to the Parisian level
...
There followed Wycherley [...] a dramatic artist of more skill and weight than Etheredge [...] his picture of society under Charles II. was too heartless and too violent to be anything but a caricature. Nevertheless, Wycherley has vitality [...] His chief faults arose from his violence, his want of decorum.
...
But Congreve [...] was the one man who brought intellect to bear on the comic stage of the Restoration. His first merit resides in his style, in the extreme beauty of the language that he uses.
...
The great successes [...] of Sir John Vanbrugh [...] where the language is negligent, but where passages of everyday life is presented with great fidelity
...
George Farquhar [...] lies even further from literature than Vanbrugh [...] In him, the step between the play and the novel was finally taken

 

An examination of Table 2 in the light of Gosse's evaluation shows that Congreve, the stylistically most sophisticated author, has a relatively high have-percentage of 28%, compared to 22-24% which was the average of the periods in which he wrote; while William Wycherley, who was described as "coarse" and "scandalously farcical", has the lowest have-ratio of the individual writers except for Josselin - the simple clergyman - who had no occurrences of have + PP.

The conclusion drawn from this in combination with what was said above about the letter writers is that the usage of have rather than be grew more common first among educated people and especially in written language, while the spoken language of the lower classes was somewhat more conservative. Cf. however section 3.1 on the reliability of comedies as a source of spoken language. It can be pointed out that Rydén & Brorström (1987:205 f.) remark that "recent research has shown that the class most particular about 'correct' linguistic usage is the (lower) middle class".

In spite of the short time span within which the texts were written, a certain diachronic development can be evidenced in the statistics. In the period 1660-1669, the be/have ratio was 87:17 (16% have) while the period 1700-1709 has a ratio of 34:11 (24% have). The overall figure for the whole period is 80% be and 20% have, which is very close to earlier estimations. It should however be remembered that the letters comprise a smaller part of the corpus than the other text types, and it is probable that if the letters had been equally represented, the overall ratio would have been more in favour of have than it turned out to be.

An obvious source for comparisons is Johannes Söderlind's work (1951) on John Dryden's verb usage; many of the statements he makes on the be/have usage in Dryden hold for this material as well. Only a few rather minor discrepancies were found, most notably with the verb land which Söderlind found with have only, but which in the present corpus was evidenced only with be.10 This can, however, easily be explained with the fact that Söderlind's only instance of land is found in a have-promoting context, namely that of past perfect in a modal clause. The verbs flee, fly, retire, steal and set were in Söderlind's corpus found with be only, but have been found with have in the present material; fly and steal only with have.

 

4.1.1 Resolve, engage and agree: the problem of classifying ambiguity

In section 2.1, the difficulty of ascertaining to which of the categories in section 1.1 some semi-adjectival verbs belong was discussed. Especially difficult are the verbs resolve, engage and agree.

In the present corpus, resolve occurs in clauses of similar structure with both variant forms of the be/have paradigm:

 

[13] They have resolved to make you some amends (Vanbrugh 1696 p. 312)
 
[14] ... my lady and the whole knot of the virtuous gang, as they call themselves, are resolved upon a frolic of coming to you tonight in a masquerade (Wycherley 1675 p. 289)

 

This makes the verb very ambiguous as to its transitive/intransitive status.

Even more troublesome to classify is the example

 

[15] Therefore I am resolved to confess the truth to the whole world, though I die a beggar for it (Vanbrugh 1696 p. 384)

 

where the variant of be cannot be definitely put into any of the categories A-C in 1.1. In this example, I am resolved can be substituted with the clearly transitive VP I have decided, which gives the phrase a "feeling of transitivity". However, if the example were transitive, the use of be would mean that it had to be passive which is clearly not the case.

Neither can the participle be intransitive nor adjectival, since it is followed by a to-infinitive as object.

There are several instances using the same type of construction as the example just mentioned:

 

[16] Well! we are in too desperate a condition to stand upon articles, and are resolved to yield on any terms (Etheredge 1668 p. 19)
 
[17] If Mr. Courtall had killed thee I was resolved not to survive thee (Etheredge 1668 p. 59)
 
[18] I was fain to come up and fetch him, or go without him, which I was resolved not to do (Wycherley 1675 p. 275)
 
and also example 13 above.

 

Ando (1976:165) assumes that be resolved is a passive construction of the type "statal passive", and makes the following interesting remark:

We cannot satisfactorily explain the provenance of passives of this type unless we posit that they have developed from the corresponding reflexive forms, e.g., resolve oneself (=F. se résoudre) -> be resolved. In other words, I assume that reflexive forms must have existed initially before the corresponding passives came into being. Just as reflexive forms are semantically intransitive, so the whole sequence be+en may be interpreted as constituting a virtually intransitive verb.

 

Ando's discussion about the possible origin of the forms left aside, the interpretation of be/have+resolved as a "virtually intransitive" verb is worth considering, since it gives a possible solution to the problem stated above. Neither of the examples he quotes seem, however, to be of the type be/have resolved to + VP.

The ambiguity is too great to make an inclusion of resolve in the statistics advisable, and there are too few examples in the corpus to make separate statistics on resolve alone.

Engage and agree are found in contexts where it is impossible to decide whether they are verbal or adjectival, and, if verbal, whether they are intransitive or transitive (when used with have as in example 20 below). As stated in 2.2, all instances of these verbs found have been excluded.

 

[19] You need not trouble yourself about that, madam, he has engaged to dine with Mr. Courtall at the French House (Etheredge 1668 p. 11)
 
[20] We are engaged about a business the knowledge of which may make you laugh hereafter (Etheredge 1676 p. 196)
 
[21] You are agreed, then, at last? (Vanbrugh 1697 p. 401)
 
[22] So pray let's all be friends, she and I are agreed upon the matter before a witness (Congreve 1700 p. 400)

 

4.2 The factors examined11

The factors examined and recorded were listed in section 2.1. A more thorough discussion of the various factors and how they have influenced the choice of variable in the aux+PP paradigm will follow here.

The parameters have been listed in this order: the first four factors (of which the second and third are found under the same heading) are textual and grammatical factors; factors iv and v are textual and lexical; and the last factor is contextual. See section 2.1.

The statistics for the various parameters are given in Table 4.

4.2.1 The -ing form (ING)

 

[23] ... and being come now to an agreement with my barber to keep my periwig in good order at 20s. a year, I am like to go very spruce, more than I used to do (Pepys p.237)
 

Instances where the aux is in the progressive form comprise about 9% of the total examples. The factor seemed slightly be-promoting when isolated, where the be ratio was somewhat higher than the average. See also under the various verbs below. No statistic evaluation could be made on +ING with other parameters, owing to the small number of such instances.

Only one instance of +ING was found in the comedies, and that was not part of the dialogue of the drama:

 

[24] Here Archer, having come round and seized the pistol, takes Gibbet by the collar, trips up his heels, and claps the pistol to his breast (Farquhar 1707 p.661)
 

This suggests that the -ing form of auxiliary verbs, which is a useful form for contracting clauses in written language12 but sounds rather clumsy in speech was not used in spoken language at the time.

Visser (1973:2228) makes the following remarks on the history of the progressive forms of the auxiliaries: "It is a fact that by Shakespeare's time the having + past participle colligation had become a well-established and fairly frequent idiom [...] In the centuries subsequent to the Elizabethan period the idiom went on gaining ground". On page 2229 he argues that being is a collocation of having been. This last claim may be contested on the basis of the following two instances, which do not seem to be quite as resultative as is expressed with the form having been:

 

[25] ... the Lords resumed their House, the King being gone (Pepys p. 225)
[26] My Lord Clifford being about this time return'd from Tunbridge (Evelyn p. 324)

 

4.2.2 The perfect infinitive (PI) and modal auxiliaries (MOD)

 

[27] 'Tis the best plot in the world: your mother, you know, will be gone to church, my spouse will be got to the ale-house with his scoundrels, and the house will be our own (Farquhar 1707 p. 622)
 
[28] Our townsmen are mighty ashamed of their bringeing the Earl of Anglesey hither [...] when ye alderman and cheife of ye town would have gon out to meet him, he utterly refused to be received in such manner (Prideaux p. 102)
 

All the 21 instances of +MOD occurred in combination with the parameter +PI. 13 out of the 21, or 62%, had have+PP in the VP, which confirmed the construction to be strongly have-promoting.13 It is also worth noticing that in the category +PI & -MOD, 67% of the examples take be:

 

[29] Be you sure only to keep awhile our great secret, till he be gone (Wycherley 1676 p.474)
 

Thus it can be inferred that the perfect infinitive by itself does not promote have to any special degree. Note, however, that the VP be gone is often adjectival.

 

[30] Why will you be gone so soon? (Etheredge 1676 p. 217)
 
[31] Now slip you down stairs, and wait at the door till my husband be gone (Farquhar 1699 p. 179)

 

4.2.3 The past perfect (PAST)

[32] But he and all this day's company and Hales were got to the Crown tavern (Pepys p. 216)
 
[33] When your husband ran away from you, if you had fallen into some of my acquaintance, 'twould have saved you many a tear (Vanbrugh 1696 p.341)
 

+PAST in combination with +COND occurred only with have (cf. section 4.2.6 below). The total ratio for the past perfect is 58:28, or 33% with have. Notable is the context +PAST & -COND, the have-percentage for which was 17%. Apparently the factor +PAST by itself does not promote have to any degree.

The present perfect had a ratio of 278:54, or 16% have.

 

4.2.4 Adverbs stressing recent action or recent result of action (REC)

[34] I was newly come into the Exchange and intended to call at your shop before I went home (Etheredge 1668 p. 29)
 

Adverbs like newly, recently and lately proved a near-knockout factor in favour of be. The ratio was 65:2. The probable cause of this is the clearly resultative stress these adverbs put on the sentence, the consequence of which is that the new state is highlighted rather than the action of bringing it about.

 

4.2.5 The negative (NEG)

[35] I wonder sir Oliver is not come yet (Etheredge 1668 p.47)
 
[36] Your lady is the most virtuous woman in the world, and nothing has passed but an innocent frolic (Vanbrugh 1697 p. 362)
 

The negative in its various realisations has in earlier research been found to favour have (see Rydén & Brorström 1987:188, Fridén 1948:46). The present material shows that after the negator not, the be/have percentage is 64:36; but if the instances (six examples) with not yet, all of which occurred with be, were removed the remaining examples with not (19 examples) had a be/have percentage of 53:47, which is remarkably high.

Other negators found were none (one example with be, one example with have); nothing (two examples with have); no (two examples with have) and neither (one example with be).

 

4.2.6 Conditional contexts (COND)

[37] Sir Phil gave baile into ye Court being a Member of Pt or he had gone to ye condemn (Lyme p. 37)
 
[38] When your husband ran away from you, if you had fallen into some of my acquaintance, 'twould have saved you many a tear (Vanbrugh 1696 p.341)
 

The context of rejected hypothetical condition (20 examples) proved a knockout factor in favour of have. This context coincided in some cases with the past perfect, in which cases the influence was uncertain since the context +PAST & +COND invariably promoted have. See Rydén & Brorström (1987:186).

The overall ratio for conditional contexts was 7:22. Several of the instances were both +COND and +MOD.

The division of the parameters into textual versus contextual and grammatical versus lexical categories did not result in any statistically significant difference between the various categories. Nevertheless, they were kept to serve as a descriptive model for the parameters.

 

4.3 Verbs with more than 10 occurrences

65 verbs were found, of which 38 occurred only once. Separate statistics were made on those verbs which occurred more than 10 times.

 

4.3.1 Become

Be/have ratio: 21:0 (100%:0%)

Become did not appear in any have-promoting context such as negative or conditional contexts.

A caveat must be inserted here: in Etheredge 1676 (p. 189) an instance of become with have is found:

 

[39] Bellinda, what has become of you? We have not seen you here of late
 

This rather peculiar occurrence, lacking all have-promoting features, caused a search in other editions of the comedy. Two other editions of the text14 had the variant form has been become, which seems much more likely than has become in this resultative context.15 There is an example with be+become only a few pages earlier in the text within a similar (although not identical) context:

 

[40] Madam, what will your mother think is become of you? (p. 185)
 

As shown in Appendix A, the editor of this edition claims to have used the first edition of the comedies, and thus it cannot be certain which printing is the correct one. The assumption made is that this case was a misprinting.

Fridén (1948:106), Ando (1976:121) and Söderlind (1951:49) all report a marked preponderance of be with become. Visser's earliest instance with have+become is from as late as 1623 and he cites several occurrences of be+become from the 20th century (1973:2047). When an archaic form like be+become is used in a 20th century text, it can, however, be supposed that it is a conscious stylistic choice made by the author (cf. Rydén & Brorström 1987:212).16

 

4.3.2 Come

Be/have ratio: 109:21 (84%:16%)

Seven of the examples with have were found in conditional contexts, four of which were in the form of a hypothetical clause of rejected condition.

The -ing form with come had a ratio of 9:1; probably owing to the fact that the combination "being/having come" is an expression of state.

The abstract, non-physical meaning of come, in expressions like "come to one's senses" and "come to do something" had a ratio of 20:4.

An meaning of come which does not seem to fit into any of the meanings of come listed in the OED is found in the instance

 

[41] My Ld Lovelace hath been very busy makeing mayors at Woodstock and Wallingford, but hath come of very dully in both places (Prideaux p.105)

 

This example has been included in the statistics and counted as an abstract meaning of come.

 

4.3.3 Fall

Be/have ratio: 12:8 (60%:40%)

Visser (1973:2058) states that have+fallen was "already extremely frequent throughout the Middle English period". Fridén (1948:80) found 29% have in Shakespeare.

The expression fallen out is found in two instances with have and one instance with be: The expression can focus either on the action itself or on the result of the action, as shown in these instances:

 
[42] ... since things have fallen out so luckily, you must needs receive your woman into favour again (Etheredge 1668 p. 67)
 
[43] There is a business, Mr. Alderman, fallen out (Farquhar 1699 p. 165)
 

The second instance with have was found in a hypothetical clause of rejected condition:

 

[44] If this lucky business had not fallen out, I intended with your good leave to have outbid you for her ladyship's favour (Etheredge 1668 p.51)

 

4.3.4 Get

Be/have ratio: 19:4 (83%:17%)

Intransitive get can either mean "become" or "reach".17 There is, however, only one instance with get = "become" in the present material:

 

[45] Sir Oliver, madam, before I did perceive it, was got near that pitch of drunkenness which makes him come reeling home (Etheredge 1668 p. 20)

 

The earliest instance of have+get in Visser (1973:2060) is from 1702,18 but the present material yielded two instances from 1675:

 

[46] When all's done, a rival is the best cloak to steal to a mistress under, without suspicion; and when we have once got to her as we desire, we throw him off like other cloaks (Wycherley 1675 p. 254)
 
[47] So, she has got into my chamber, and locked me out (Wycherley 1675 p. 275)

 

Fridén (1948:91) states that "get is found with the have-construction in Spenser and in Shakespeare", and the two instances he quotes are of the type get = "reach".

 

 

4.3.5 Go

Be/have ratio: 84:23 (79%:21%)

Eight instances of go+have were found in a context of hypothetical unrealised condition. Three of the other cases were in the context +PI.

The +ING ratio is highly be-favouring for this verb, with 11 cases of be and one of have. This last instance is clearly stressing the aspect of action (durative context, see section 1.1):

 
[48] I fear the Drs designes are above his strength, by reason of his age, which is great, he haveing gon chaplain to the ambassadeur at Constantinople before our King was born (Prideaux p. 43)
 

Instances with the imperative be gone, written as one or two words, were excluded since there are no cases of have in this function.19

 

[49] If my words frighten thee, be gone the sooner (Wycherley 1676 p. 390)

 

A rather peculiar usage of be+gone is found in Russell p. 84:

 

[50] ... for the day I sent it to the coach, that was newly gone; and the next you was gone, I believe

 

The abstract or figurative sense of go, found in expressions like "go against something", had a be/have ratio of 1:8. The instance with be was found in the context +ING.

 
[51] I hear that the Queen hath miscarried of a perfect child, being gone about ten weeks (Pepys 224)

 

The ratio in Dryden (Söderlind 1951:52) was 55:17, or 24% have. Ando (1976:125) reports a ratio in Marlowe of 29:3.

 

4.3.6 Grow

The verb grow, which had 36 occurrences, was not found with have.

Söderlind (1951:53) has one instance with have from 1671. Fridén (1948:109 ff) gives a very low percentage of have. Visser (1973:2048) states an uncertain occurrence of hadde growid (possibly was growid) from circa 1380, one from 1573 and one from Hamlet (which is probably the same case that according to Ando (1976:126) is the only occurrence of have+grown in Shakespeare). On the other hand, Visser (idem) cites several examples of be+grown from the 20th century.

Although the verb from the evidence quoted above could have been assumed to have a preference for be, the total reign of be can only be explained with the fact that the examples with grow had no have-promoting parameters like modal auxiliaries or hypothetical rejected context. Indeed, the only of the parameters listed above found with grow was the negative, which was only found in one instance of the be-promoting not yet.

 
[52] I am sorry he is not yet grown up to be old enough to succeed the Dr (Prideaux p.44)

 

4.3.7 Pass20

Be/have ratio: 10:11 (48%:52%)

Pass was the only verb found more than 10 times to have a have-percentage of more than 50%. The verb is fraught with difficulties, since it is sometimes clearly adjectival. However, all instances of passed and past found were included.

Fridén (1948:93) remarks that "in Shakespeare have seems to be the ordinary auxiliary in combination with passed both in transitive and intransitive functions, while be is used with past". Rydén & Brorström (1987:134) get similar results in their examination of 18th century texts. In the present material, 7 examples of be+past and one example of have+past were found while the figures for passed were 3 be and 9 have. Pass'd was found in one instance with have.

 

[53] My Wife is now, God be thanked, past the great danger (Lyme p. 167)
 
[54] This winter shall be the fiery trial of my virtue, Which when it once has past, You'll be convinced 'twas of no false allay (Vanbrugh 1696 p.318)
 
[55] How can you, after what is passed, have the confidence to ask me? (Farquhar 1707 p. 664)
 
[56] The nigh relation I have to you and the respect I know sir Oliver has for you, makes me forget all that has passed, sir (Etheredge 1668 p.67)
 
[57] Sure I have pass'd the gulf of silent death, / And now I land on the Elysian shore! (Farquhar 1707 p. 642)

 

Söderlind (1951:57) found that pass was "found slightly oftener with have than with be" in Dryden.

The high ratios for have are, supposedly, owing to the actional aspect of the verb.

 

4.3.8 Return

Be/have ratio: 15:1 (94%:6%)

The instance with have was found in a hypothetical clause of rejected condition:

 

[58] My Suit, if you had not return'd (Wycherley 1676 p. 392)
 

The contexts in which the instances were found were for the most part focused on the resultative aspect of having returned and not on the process of returning in itself.

 

[59] He went to France a plain, bashful English blockhead, and is returned a fine undertaking French fop (Etheredge 1676 p. 213)
 
[60] The King being return'd from Winchester, there was a numerous Court (Evelyn p. 459)
 
[61] Our Regius Professor is returned from his northern progresse with his two baronets with him (Prideaux p.94)

 

Söderlind (1951:56) states a high ratio for be+return in Dryden. Rydén & Brorström (1987:151) report that "in the 18th century be has still a distinct ascendancy".

 

To section 5

To the table of contents