University of Virginia Library

II
Grammar: Alterations in Third Edition

Sherbo's essay, much less inclusive than Keast's, concentrates on Johnson's revisions for his fourth edition (1773) of Shakespeare's plays (of which George Steevens was a collaborator) and for the fourth folio edition (also 1773) of his Dictionary. But Sherbo mentions (p. 19) that in the third edition of the Dictionary Johnson, surely reacting to John Wilkes's witty remark on the mistake, modified his original comment, in the Grammar, about the letter H to read: "It seldom, perhaps never, except in compounded words, begins any but the first syllable" (our italics). His principal subject being Johnson's revisions in 1773, Sherbo says nothing else about authorial changes in the 1765 version of the Grammar.

However, our collation of the entire text reveals that Johnson did not revise the Grammar in the second edition but that he altered it in the third edition more often than anyone has ever pointed out. For example, of the thirty-three substantive variants originating in this edition, twenty-three can be confidently labeled "authorial." Fifteen of the twenty-three are additions which divide the Grammar into parts and sections—namely, "PART I. Of ORTHOGRAPHY." (with three roman section numbers), "PART II." (with six roman section numbers), "PART III.", and "PART IV." (with two roman section numbers) (sigs. L1r, L1v, L2v, M1r, M1v, N1r, N2r, and N2v). Four more are also additions—three the identifications of authors of quoted passages on sig. O1r ("Pope.", Elijah "Fenton.", and David "Lewis."), the fourth an entire sentence following that on sig. L2r ending in "as frosty winter"

("Yet I am of opinion that both w and y are always vowels, because they cannot after a vowel be used with the sound which is supposed to make them consonants"). The four remaining authorial revisions consist of clarifications or corrections in as many statements: (1) "None of the small consonants have a double form, except s, s" is altered to "None of the consonants have a double form, except the small s, s" (sig. L1r); (2) as pointed out above, the assertion that the letter H "seldom, perhaps never, begins any but the first syllable . . ." becomes "seldom, perhaps never, except in compounded words, begins . . ." (sig. L2r); (3) "But it may be observed of y as of w, that it follows a vowel without any hiatus, as rosy youth" is changed to "It may . . . youth, but yet that it cannot be sounded after a vowel" (sig. L2v); and (4) in the sentence "The verse of twelve lines, called an Alexandrine, is now only used to diversify heroick lines," "twelve lines" is corrected to "twelve syllables" (sig. O1r). We have adopted all the revisions in this group except two, the identification of the author David "Lewis," which was expanded to "Lewis to Pope" in the fourth edition (see below) and (3) above, which was superseded by Johnson's revision in the fourth edition (see below).

The rest of the substantive alterations can be divided into the four which we think Johnson probably made and the six which he possibly made. The first of the former group changes "gradation" to "gradations" in the sentence reading, in part, "In treating on the letters, I shall not . . . enquire into the original of their form . . . ; nor into the properties and gradation of sounds . . ." (sig. L1r). "Gradations," denoting plurality and diversity, clearly describes human sounds more accurately than does the singular "gradation," as the accompanying term "properties" evinces. And Johnson seems the plausible cause of the shift, although a careful compositor or proof corrector cannot be entirely ruled out. Likewise, the "as" inserted after "consonant" in the phrase "is a consonant, as ye, young" (sigs. L2r-L2v) and that inserted after "ain" in the phrase "except words in ain, as cértain" (sig. N2v) are likely authorial additions, although, again, another person might have been responsible for them. The same comment applies to the correct replacement of "hung" by "stunk" in the original sequence of "drunk, sunk, shrunk, hung, come" (sig. M2v). We have adopted all these variants.

The last group of changes could have been made, we conclude, either by Johnson or by another person. The first is the deletion of the superfluous "it" in the sentence on sig. L1v originally beginning "F, . . ., it is numbered. . . ." We have adopted this correction. The next three consist of alterations from the plural to the singular form of verbs—" Wr imply" to "implies," "Sw imply" to "implies," and "C1 denote" to "denotes" (sig. N1v). An examination of the context immediately reveals the reason for the change: to achieve conformity between the number of the verbs and that of neighboring comparable verbs. However, since, as we note below, the fourth edition of the Grammar contains the plural form of all the verbs just described, we have chosen the same form for our text. The fifth variant, occurring too in the fourth edition, corrects the letters "ly" to "ty" in the sentence "Words ending in ly have their accent on the antepenult, as pusillanímity, actívity" (sig. N2v). We have

adopted this correction. Added to the beginning of line 4 ("Shall that holy fire") of Michael Drayton's "An Ode Written in the Peake" (sig. N2v), the word "Or" is the final member of this group. Johnson or someone else, we surmise, noticing that the line, unlike the others in the poem, lacks six syllables, and not consulting an independent text, proceeded to regularize the line by prefixing the "Or" (cf. the two appearances of "Or" in the last stanza). But we have retained the five-syllable line (also in the fourth edition) because it occurs in all early editions of Drayton's poem, including that (1748) from which Johnson drew illustrative passages in the Dictionary.[2]

Besides the thirty-three substantive differences, the third folio edition of the Grammar contains thirty-seven accidental variations from both its predecessors and its successor, two only from its predecessors, and one only from its successor. Twenty-three of these are changes in punctuation, eight in spelling, four in accent marks, three in italics, one in the location of a sentence, and one in the position of a word. We have accepted two of the alterations in punctuation: (1) On sig. L1v, the single sentence reading in part: "C, . . . , never ends a word; therefore we write stick, block, . . ., in such words C is now mute" becomes two sentences by the replacement of a period for the comma immediately preceding the phrase "in such words"; Johnson seems to us the probable source of the change. (2) On sig. L2v, the period following the sentence beginning "Had he written" and concluding "appeared thus" (succeeded by four lines of poetry) is changed to a colon (i. e., "thus:"); since the same change occurs in the fourth edition, Johnson might have been the cause of it. The remaining twenty-one variants in punctuation are either manifestly improper or less suitable to their contexts than their alternatives in the first, second, and fourth editions.

We have adopted two of the changes in spelling. (1) On sig. M1r, below the line of poetry ending "noble savage ran" (Dryden's Conquest of Granada, Part I, l. 1.209), "Dryd." is expanded to "Dryden," which, possibly Johnson's revision, harmonizes with "Milton" located directly above (and below a passage from Paradise Lost, I, ll. 1-3). (2) On sig. O1r, in the line of poetry beginning "Fairest piece" (Edmund Waller's "To Zelinda," l. 1), "welform'd" becomes "well-form'd," the same spelling as that in the wordlist of Johnson's Dictionary, where Waller's line is cited under well (adverb, sense 13). The other six variants are either errors or less preferable than their alternatives.

On sig. N2v, in the sentence beginning "1. Of dissyllables," an accent mark is correctly placed above "fáirer," thereby correcting a mistake in the first, second, and fourth editions. Similarly, in the sentence beginning "4. All dissyllables" accent marks are correctly placed above "cránny," "lábour," and "fávour," thus remedying an omission in the first, second, and fourth editions. Johnson might have been responsible for all these changes, which we have adopted.

On the other hand, we have accepted none of the remaining groups of

accidental variants in the third edition. The first three changes diverge from the pattern of italicizing evident in the first, second, and fourth editions. Forming the fourth difference, a short sentence is transferred—inadvertently, we assume—from its proper location, as evidenced by the context, to the end of the next paragraph. And the final difference—the location of the name of an author—is superseded by the location of the same word in the fourth edition.