Worthy of Another Look: John Gilbert's 1892 Account of the 1830 Printing of the Book of Mormon*
Occasionally historians and other observers of the past attempt to discredit someone’s account of a past event by referring to the age of the person when the account was given. Age frequently becomes an argument against the account if the historian or observer does not agree with the implications of that account.
Yet the real issue is how an account matches up with other accounts or, even more significantly, how it matches up with the physical evidence that remains. Independent, physical evidence can often be used to test the reliability of accounts. A good example of this procedure in analyzing accounts can be found in the analysis by Don Enders of numerous statements made in E. D. Howe’s 1834 Mormonism Unvailed, in particular, claims by some of the residents around Palmyra that Joseph Smith’s family were poor and lazy. Enders compared these claims against the original land and tax records and other local government papers from the 1820s and 1830s and discovered that the assessment of Joseph Smith Sr.’s property, based on the 1830 tax records, shows that the valuation of the Smith farm per acre exceeded that of nine out of ten farms owned by families who criticized the Smiths in Mormonism Unvailed. This finding calls into question the overall validity of these accounts in Howe’s book denigrating the Smiths’ work ethic.1
In this article, I would like to consider a statement made by John Gilbert, the compositor (or typesetter) for the 1830 edition of the Book of Mormon. Gilbert made this statement on 9 September 1892, when he was 90 years old. In his statement, a typescript located in the King’s Daughters Library in Palmyra, Gilbert describes events that occurred 63 years earlier (see pages 70–72 at the end of the article). Now of course we could dismiss his account (if we didn’t like what he was saying about the early publishing history of the Book of Mormon) by simply referring to his age, or the lateness in making this statement, or even his anti-Mormon bias. But the better procedure is to test this statement against what we have been able to discover about the printing of the first edition of the Book of Mormon.
This process includes evidence from the two Book of Mormon manuscripts: the original manuscript and the printer’s manuscript. The original manuscript is the dictated manuscript the scribes wrote down as Joseph Smith translated the Book of Mormon. During 1829–30, a copy of the original manuscript was made. This copy is called the printer’s manuscript because, for the most part, this was the manuscript that was taken to the printer’s shop in Palmyra, New York, where the type was set for the first edition of the Book of Mormon, published in 1830. About 28% of the original manuscript is extant. Most of the extant portions of the original manuscript are owned by the LDS Church. The printer’s manuscript is owned by the RLDS Church, now renamed the Community of Christ, and is extant except for three lines. Fragments of the original manuscript show that the original (dictated) manuscript rather than the copied printer’s manuscript was used to set the 1830 edition from Helaman 13 through Mormon 9.2
In addition to the two manuscripts, this analysis of John Gilbert’s statement has involved the examination of about one hundred copies of the 1830 edition, an original proof sheet of the 1830 title page, and a complete set of unbound sheets of the 1830 edition (sometimes called the “uncut sheets”) that Gilbert had saved.
I list here a number of claims Gilbert made in that statement about the printing of the 1830 edition and compare those claims with the extant physical evidence dating from 1829 and 1830.
A few pages of the manuscript were submitted as a specimen of the whole, and it was said there would be about 500 pages.3
There were 466 pages in the printer’s manuscript and probably a few more in the original manuscript, perhaps as many as 480 pages. (Of course, this number excludes the first 116 pages from the book of Lehi and the beginning of the book of Mosiah, lost by Martin Harris in the early summer of 1828.) In either case, the estimate of 500 pages that Gilbert remembered is close to the actual number of pages.
In the forepart of June 1829, Mr. E. B. Grandin, the printer of the “Wayne Sentinel,” came to me and said he wanted I should assist him in estimating the cost of printing 5,000 copies of a book that Martin Harris wanted to get printed, which was called the “Mormon Bible.” . . .
The contract was to print and bind with leather 5,000 copies for $3,000.
The number 5,000 agrees with other accounts of the press run for the 1830 edition. For instance, these same figures are found in Joseph Smith’s 1839 History. The final version reads:
Having finished the translation we went to Palmira Wayne Co, N.Y. and secured the copy right and engaged Mr. Gilbert Grandon to print 5,000 copies for the sum of $3,000.4
The size of the page was agreed upon, and an estimate of the number of ems in a page, which would be 1,000.
An em is a measure of type width equal to the point size of the font being used. There are about 1,075 ems per page in the 1830 edition, with 25 ems per line and 43 lines per page (excluding the header on each page). Gilbert’s recollection of the estimated number of ems is close to the actual count for an 1830 page.5
A page of manuscript would make more than a page of printed matter, which proved to be correct.
As already noted, there are 466 pages of manuscript in the printer’s manuscript and perhaps as many as 480 pages were in the original manuscript. The 1830 edition itself has 590 pages, which means that one manuscript page provided about one and a fourth pages in the 1830 edition.
Mr. Grandin got a new font of small pica, on which the body of the work was printed.
The “small pica” of the 1830 edition is a 10-point type. The type used in the 1830 edition is called Scotch Roman, a very common type designed about 1810 by Richard Austin in Edinburgh, Scotland. This type face was widely used throughout the nineteenth century.6
The type used in the 1830 edition had only a few pieces of broken type. The type imprint in 1830 copies is sharp and clean and shows little wear.
When the printer was ready to commence work, [Martin] Harris was notified, and Hyrum Smith brought the first installment of manuscript, of 24 pages, closely written on common foolscap paper.
The entire printer’s manuscript is a collection of gatherings of sheets. To form a gathering, Oliver Cowdery (the principal scribe for the printer’s manuscript, as well as the original manuscript) would typically take 6 sheets of foolscap paper (a size of paper), line them, and fold them down the center to form a gathering of 24 pages or 12 leaves. Later, after writing the text, he would secure the gathering by producing at least 4 holes (or “stabs”) along the fold (or “gutter”) and weaving in yarn and then tying it to hold the gathering together. The very first gathering for the printer’s manuscript starts at the beginning of 1 Nephi and goes up to 1 Nephi 14:21. Like most of the other gatherings in the printer’s manuscript, this first one contains 24 pages (6 foolscap sheets folded widthwise to form 12 leaves or 24 pages).
Foolscap paper originally referred to a watermark showing a fool’s cap, but by the 1800s this term was universally used to refer to a paper size. The sheets for the printer’s manuscript show some variance, but range from 31.4 to 33.1 cm in width and from 38.3 to 41.5 cm in length. Published accounts (given in the Oxford English Dictionary under “foolscap”) indicate that foolscap paper varied from 12 to 13.5 inches in width and from 15 to 17 inches in length (that is, from 30 to 34 cm in width and 38 to 43 cm in length). All the sheets in the printer’s manuscript are within these bounds, as are the extant sheets of the original manuscript.
The title page was first set up, and after proof was read and corrected, several copies were printed for [Martin] Harris and his friends.
One of the individuals in the print shop that day was Stephen Selwyn Harding, who later served as territorial governor of Utah (1862–63). Harding received one of these copies of the proof sheet of the title page and in 1847 donated his copy to the LDS Church. This copy has been on display at the Church Museum in Salt Lake City. In comparing this proof sheet with the title page as actually published, we see that a number of misspellings were corrected; in addition, the spacing (or “leading”) between the various lines, especially in the title and subtitle, was increased.
On the second day—[Martin] Harris and [Hyrum] Smith being in the office—I called their attention to a grammatical error, and asked whether I should correct it? Harris consulted with Smith a short time, and turned to me and said: “The Old Testament is ungrammatical, set it as it is written.”
For the most part, Gilbert did not edit out the grammatical “errors.” The vast majority of them were copied over straight from the manuscripts into the 1830 edition. In some cases some accidental correction seems to have occurred. And in a handful of cases we have specific evidence that either John Gilbert or Oliver Cowdery consciously corrected what was perceived to be pronominal redundancies. For instance, in Ether 9:8, the printer’s manuscript originally read as follows:
& now the brother of him that suffered death & his name was Nimrah & he was angry with his father because of that which his father had done unto his brother
While punctuating the manuscript to set the type for this part of the text, Gilbert placed the intrusive “& his name was Nimrah” in parentheses and then crossed out the words “& he” that followed. This kind of conscious editing is infrequent in the text. The vast majority of “ungrammatical” expressions were left unchanged.
Martin Harris, Hyrum Smith, and Oliver Cowdery were very frequent visitors to the office during the printing of the Mormon Bible. The manuscript was supposed to be in the handwriting of Cowdery. . . .
Cowdery held and looked over the manuscript when most of the proofs were read. Martin Harris once or twice, and Hyrum Smith once, Grandin supposing these men could read their own writing as well, if not better, than anyone else; and if there are any discrepancies between the Palmyra edition and the manuscript these men should be held responsible.
The printer’s manuscript is mostly in Oliver Cowdery’s hand (84.6%). A not-yet-identified scribe (referred to as scribe 2) accounts for 14.9% of the printer’s manuscript. This scribe basically transcribed two large portions (from Mosiah 25 to Alma 13, and from 3 Nephi 19 to the end of Mormon), but in the first portion, Hyrum Smith briefly took over for scribe 2 on five different occasions (from Mosiah 28 to Alma 5). Hyrum’s minor contribution amounts to only 0.5% of the text.
But the printer never saw the second portion done by scribe 2. Instead, the original manuscript was taken in for this portion of the typesetting. All extant fragments of the original manuscript from this part of the text (from Helaman 13 to the end of Mormon) are in Oliver Cowdery’s hand, so if we presume that all this portion of the original manuscript was in Oliver’s hand, the 1830 printer saw Oliver Cowdery’s hand for slightly over 91% of the text. By this calculation, scribe 2 then accounts for 8.5% of the text and Hyrum the remaining 0.5%. So Gilbert’s comment that the manuscript was supposed to be in Oliver’s hand is probably accurate for about 91% of the text.
Gilbert’s comment that Oliver Cowdery did most of the proofing, but that Martin Harris did it twice and Hyrum Smith once is intriguing, especially since these rankings are consistent with the frequency with which the printer set type from the handwriting of Oliver Cowdery, scribe 2, and Hyrum Smith. The additional statement from Grandin about proofing “their own writing” suggests that Martin Harris might have been scribe 2, although of course Gilbert’s initial statement about the handwriting implies that Oliver Cowdery was the only scribe. Except for his signature, there are apparently no identified extant examples of Martin Harris’s handwriting.7
We also have definite evidence that Oliver Cowdery was learning from his proofing of the 1830 edition. For instance, by the time he got into 3 Nephi, Oliver had learned that exceeding(ly) is spelled with two e’s after the c, not as exceding(ly), which is how he consistently spelled the word in the original manuscript as well as in the printer’s manuscript before 3 Nephi 12:12. From then on in the printer’s manuscript, Oliver always spelled exceeding(ly) correctly.
In addition, Oliver Cowdery also learned to hyphenate at the end of lines. Earlier he had always hyphenated at the beginning of the line (in the original manuscript and the first part of the printer’s manuscript). For example, in the original manuscript, if only accord of according fit at the end of a line, Oliver would have written accord at the end of the line and -ing at the beginning of the next line. But when he finally learned that hyphenation occurs at the end of the line, Oliver would have written accord- at the end of the line, but still he would have kept the hyphen at the beginning of the next line (that is, -ing), thus ending up with two hyphens.
Oliver Cowdery started this practice of double hyphenation at the beginning of 2 Nephi (page 49 in the printer’s manuscript), but here he put hyphens at the end of a line only once or twice a page, so that in this part of the printer’s manuscript most hyphenated words had only a single hyphen, at the beginning of a line. But by the time Oliver got through 200 pages of the printer’s manuscript, he started to hyphenate more frequently at the end of lines, so that ultimately in the last half of the manuscript we often find double hyphenation more than ten times a page.
Every chapter, if I remember correctly, was one solid paragraph, without a punctuation mark, from beginning to end.
. . . I punctuated it to make it read as I supposed the author intended, and but very little punctuation was altered in proof-reading.
Originally, very little punctuation appeared on the printer’s manuscript and virtually none on the original manuscript, including that portion (from Helaman 13 to the end of Mormon) used to set the type for the 1830 edition. For the first part of the printer’s manuscript, Oliver Cowdery copied the original manuscript without adding punctuation. He finally realized that he himself could add the punctuation, so beginning with page 106 of the printer’s manuscript, Oliver started to add a little punctuation, but only sporadically and never systematically. Moreover, Gilbert basically ignored Oliver’s punctuation.
Beginning with page 129 of the printer’s manuscript, Oliver Cowdery added paragraph marks as he prepared this manuscript, but by page 145 he stopped this practice, probably because he had realized that the compositor was ignoring his suggested paragraph breaks. In any event, all the original chapters in the Book of Mormon manuscripts were written as a single paragraph. Gilbert is responsible for the actual paragraphing in the 1830 edition, although he does not mention it in this statement. While inserting punctuation, he would also use the letter P (not the reversed paragraph symbol ¶) whenever he wanted to show the beginning of a new paragraph.
Scribe 2, unlike Oliver Cowdery, fairly consistently punctuated the portions of the printer’s manuscript that he was responsible for, although scribe 2 had only a single punctuation mark that sometimes looks like a period and sometimes like a small comma. This same mark is used interchangeably for both full and half stops. Once more, for the first portion of scribe 2’s handwriting (from Mosiah 25 to Alma 13), Gilbert ignored this rather confusing punctuation mark from scribe 2.
As Gilbert indicated, he basically typeset the 1830 edition with the same punctuation marks that he had placed in the printer’s manuscript. I would estimate that over 90% of Gilbert’s punctuation marks in the printer’s and original manuscripts were carried over without change into the 1830 edition.
Names of persons and places were generally capitalized, but sentences had no end. The character & was used almost invariably where the word and occurred, except at the end of a chapter.
In those portions of the original manuscript in the hand of Oliver Cowdery, the first word in a chapter was systematically capitalized (as were names). If the first word was and, it was written as And. Gilbert’s“end of a chapter” refers, of course, to the beginning of a new chapter, since one implies the other. But other sentence-initial words in the original manuscript were generally not capitalized by Oliver. And he wrote virtually all other examples of and as an ampersand (&). Oliver nearly always followed this same practice in the printer’s manuscript. In a couple in-stances in the manuscripts, Oliver did write and as and, but in each case he had accidentally started to write some other word and he then overwrote the incorrect word by writing out the full and rather than using the shorter ampersand.
On the other hand, it should be noted that in the book of Helaman, Oliver Cowdery started to occasionally show the beginning of a new sentence in the middle of a chapter by capitalizing the sentence-initial word, as in Helaman 5:5–6:
for they remembered the words which their father Helaman spake unto them & these are the words which he spake Behold my Sons I desire that . . .
Although Oliver never consistently applied this practice in the rest of the printer’s manuscript, still he occasionally did capitalize a few sentence-initial words in the middle of a chapter. And eventually, there are examples of mid-chapter sentences beginning with And instead of &, as in 3 Nephi 13:34–14:1:
sufficient is the day unto the evil thereof And now it came to pass that . . .
Although this sentence begins chapter 14 in our current chapter system (dating from Orson Pratt’s editing for the 1879 edition), originally this sentence occurred about one-third the way through chapter VI of 3 Nephi. But since this part of the printer’s manuscript was never seen by John Gilbert, he never saw this example of a mid-chapter And. Only in a few cases in Ether and Moroni of the printer’s manuscript could Gilbert have seen in the middle of a chapter an occasional And instead of Oliver Cowdery’s much more frequent &. In nearly all instances, Gilbert would have seen & in the printer’s manuscript.
The two other scribes in the printer’s manuscript (scribe 2 and Hyrum Smith) used both & and and interchangeably, but this variation would have occurred for only 8.9% of the text (from Mosiah 25 to Alma 13). In any event, Gilbert’s recollection of the massive use of & is accurate for the vast majority of the Book of Mormon text.
After working a few days, I said to [Hyrum] Smith on his handing me the manuscript in the morning: “Mr. Smith, if you would leave this manuscript with me, I would take it home with me at night and read and punctuate it, and I could get along faster in the day time, for now I have frequently to stop and read half a page to find how to punctuate it.” His reply was, “We are commanded not to leave it.” A few mornings after this, when Smith handed me the manuscript, he said to me: “If you will give your word that this manuscript shall be returned to us when you get through with it, I will leave it with you.” I assured Smith that it should be returned all right when I got through with it. For two or three nights I took it home with me and read it, and punctuated it with a lead pencil. This will account for the punctuation marks in pencil.
John Gilbert had to wait more than “a few mornings” after “a few days” before getting permission to take the printer’s manuscript home to punctuate it. In the first part of the manuscript, before page 73, there are only a few minor places where Gilbert added punctuation to the manuscript. These few punctuation marks are all in pencil. When Gilbert refers to reading down half a page of manuscript to determine the punctuation, he was apparently trying to determine the reading of the text and then adding the punctuation to the typeset text only, not on the manuscript itself (except in those few cases).
The first place where Gilbert began to systematically punctuate the printer’s manuscript is on page 73 (beginning with 2 Nephi 17:4). Since this place is about one-sixth the way through the manuscript, Gilbert’s impression about when he started to take the manuscript home is a little too early. I would estimate that he probably took the manuscript home sometime in the last half of September 1829, after at least one month of printing.
We do have evidence that Gilbert took the manuscript home for two days. For these two sessions, Gilbert marked the punctuation in heavy black ink, not in pencil. The first session covers pages 73–75 of the printer’s manuscript. The second session covers pages 77–79 and the first third of page 80.
After these two sessions, all of Gilbert’s subsequent punctuation marks on the printer’s manuscript (and on the original manuscript for Helaman 13 through the end of Mormon) are in pencil rather than ink. Gilbert’s penciling seems to be restricted to work actually done in the printing shop, not at home, especially since his punctuation marks are interspersed with take marks (also in pencil) that were made during the actual setting of the type. (These take marks show where in the manuscript the compositor finished setting the type for a portion of the text.)
Since the clear majority of Gilbert’s punctuation is in pencil, it is understandable that he might not have remembered that he used ink for the two nights he took the manuscript home to prepare it for typesetting.
The [Mormon] Bible was printed 16 pages at a time, so that one sheet of paper made two copies of 16 pages each, requiring 2,500 sheets of paper for each form of 16 pages. There were 37 forms of 16 pages each, 570 pages in all.
The 1830 edition has 16 pages to a signature and has 37 signatures. Of course, Gilbert could determine this by referring to a copy of the edition (or perhaps to his set of 37 unbound folded sheets). There are, however, 592 pages in the 1830 edition (37 x 16 = 592), of which the last two are blank, thus giving 590 printed pages, not 570. Perhaps the 570 is a typo for 590.
The 2,500 sheets for each signature would thus account for 5,000 copies since they were printing all 16 pages of each signature on both sides of the sheet. This process, called half-sheet imposition (or in more colloquial terminology, “work and turn”), requires that each sheet be properly oriented and lined up (a process referred to as registering) before printing the opposite side. Finally, the 2,500 larger sheets were torn or cut in two—so that prior to binding, 5,000 copies of each signature were available.
Examination of the unbound sheets shows quite clearly the torn side at the top of each of the 37 signatures. Here each of the original larger sheets was folded and cut along the crease with a bone cutter (personal communication from Don Enders), which left a rough, tornlike edge. The bottom edge has always been cut mechanically, whereas the sides always show a deckle edge—that is, the original uneven edge that results from the paper-making process itself. In addition, the two pinholes resulting from pinning down the middle of the full sheet to the tympan (the frame to which the sheet is secured during the presswork) can be found about half the time near the torn upper edge of the unbound sheets. Thus the unbound sheets clearly show that Gilbert’s statement about printing 2,500 sheets to produce 5,000 copies was entirely accurate.8
From these many examples, we can see that in every instance John Gilbert’s recollections regarding the printing of the 1830 edition of the Book of Mormon are either precisely correct or, where wrong, the error is easily explained. In a number of cases where he thought something held in every instance, the actual facts show that his recollection is still correct for the clear majority of cases. All in all, these examples show that Gilbert’s memory is very accurate, even at 90 years of age and 63 years after the fact.9
*This paper, updated here, was originally published in The Disciple as Witness: Essays on Latter-day Saint History and Doctrine in Honor of Richard Lloyd Anderson, ed. Stephen D. Ricks, Donald W. Parry, and Andrew H. Hedges (Provo, UT: FARMS, 2000), 383–405.
[The PDF version of this article includes a reproduction of the memorandum of John Gilbert from which his quotations used in this article were taken. This document retains the typing and spacing errors, but does adjust the line breaks and hyphenation. Ed.]
Royal Skousen, professor of linguistics and English language at Brigham Young University, has been the editor of the Book of Mormon critical text project since 1988. In 2009, Skousen published with Yale University Press the culmination of his critical text work, The Book of Mormon: The Earliest Text. He is also known for his work on exemplar-based theories of language and quantum computing of analogical modeling.
1. Donald L. Enders, “The Joseph Smith, Sr., Family: Farmers of the Genesee,” in Joseph Smith: The Prophet, the Man, ed. Susan Easton Black and Charles D. Tate Jr. (Provo, Utah: BYU Religious Studies Center, 1993), 213–25; see page 220 for the nine-out-of-ten statistic.
2. For further information about the use of the original manuscript as the printer’s copy, see Royal Skousen, “Piecing Together the Original Manuscript,” BYU Today (May 1992): 18–24; or Royal Skousen, “Why was one sixth of the 1830 Book of Mormon set from the original manuscript?” (22–24 February 2012, accessible online at timesandseasons.org).
3. To enhance readability, a few changes (mostly in punctuation and grammar) have been made in the quotations from the statement.
4. Joseph Smith Papers: Histories (1832–1844), Vol. 1 (Salt Lake City, Utah: The Church Historian’s Press, 2012), 353, commas added in numerals.
5. I wish to thank Jonathan Saltzman for help in determining these figures.
6. I wish to thank Jonathan Saltzman for identifying the type face for the 1830 edition. For further information about Scotch Roman, see W. Pincus Jaspert, W. Turner Berry, and A. F. Johnson, The Encyclopaedia of Type Faces, 4th ed. (Poole, Dorset, England: Blandford, 1970), 203.
7. For an early discussion of this problem, see Dean C. Jessee, “New Documents and Mormon Beginnings,” BYU Studies 24/4 (1984): 397–428.
8. I wish to thank Louis E. Crandall of the Crandall Historical Printing Museum (Provo, Utah) for his valuable assistance in identifying these aspects of the unbound sheets. For further information on printing in the 1800s and earlier, see Ronald B. McKerrow, An Introduction to Bibliography for Literary Students (1928; repr., New Castle, Delaware: Oak Knoll, 1994), 22–23 (for registering), 19 and 45–46 (for the tympan), 66–70 (for half-sheet imposition), and 103 (for deckle edge).
9. I wish to thank Richard L. Anderson and Larry C. Porter for providing copies of Gilbert’s 1892 statement. Scott Faulring provided access to some related documents; Matthew Empey, my research assistant, helped collect some of the information for this article. Don Enders also provided a helpful critique of this paper as well as some additional information.