How Joseph Smith Translated the Book of Mormon:
Evidence from the Original Manuscript

Introduction The Prophet Joseph Smith said very little about the actual process of translating the Book of Mormon. However, a thorough study of the original text of the Book of Mormon (including a detailed examination of both the original and printer’s manuscripts) and a careful review of statements made by those who witnessed Joseph Smith translating combine to provide valuable information about the translation process. Even details such as spelling corrections and textual insertions provide definite clues about how Joseph translated.

This evidence does not support theories that Joseph Smith composed the text himself or that he took the text from some other source. Instead, it indicates that the Lord exercised what I refer to as “tight control” over the word-by-word translation of the Book of Mormon. In particular, the evidence suggests that Joseph Smith saw specific words written out in English and read them off to the scribe, and that the accuracy of the resulting text depended on the carefulness of Joseph and his scribe. Indeed, this evidence is most compatible with the account that Joseph himself gave, that he translated the Book of Mormon “by the gift and power of God.”

Statements from Witnesses of the Translation Witnesses of the translation process make two kinds of claims. First of all, the witnesses provide valuable evidence of what they actually saw taking place. Generally speaking, their observations are consistent with the physical evidence in the original manuscript. On the other hand, these witnesses frequently made claims about matters that they themselves could not observe. For instance, some described what they believed Joseph Smith saw in the interpreters; and many claimed that Joseph could not go on until the scribe had written down letter-for-letter what Joseph saw. It turns out that these kinds of claims are not supported by the evidence in the original manuscript. Of course, the witnesses themselves did not see what Joseph saw. For these kinds of claims, the witnesses were either offering their own conjectures or perhaps recalling what Joseph might have told them. Nonetheless, all seemed to believe that Joseph Smith actually saw words in English, and there is evidence in the original manuscript to support this idea.

During the translation process, the witnesses were able to observe, in an open setting, the following:

•Joseph Smith placing the interpreters (either the Urim and Thummim or the seer stone) in a hat and placing his face into the hat;

•Joseph dictating for long periods of time without reference to any books, papers, manuscripts, or even the plates themselves;

•Joseph spelling out unfamiliar Book of Mormon names;

•after each dictated sequence, the scribe reading back to Joseph what was written so that Joseph could check the correctness of the manuscript;

•Joseph starting a dictation session without prompting from the scribe about where the previous session had ended.

The translation process that these witnesses observed was an open one—that is, others in the room could observe the dictation from Joseph Smith to the scribe. But early on in the translation, from late 1827 to early 1828, it appears that Joseph used a different process while translating. During this early period, Joseph would first copy some of the characters directly from the plates onto sheets of paper, from which sheets he would then translate his transcribed characters into English by means of the Urim and Thummim. During such a process, the plates were uncovered while Joseph translated (or at least while he copied the characters from the plates to paper); and since no one was permitted to see the plates until later, Joseph took precautions to prevent anyone from seeing him working directly with the plates. Martin Harris, in a couple of early statements, said that a blanket or curtain separated Joseph from him at the time he (Martin) obtained a sample transcript and translation to take to Professor Anthon in New York City.

In place of this early procedure, Joseph Smith soon turned to a method of translation that depended directly on the interpreters alone, so that the plates did not have to be viewed, and thus the translation could be done openly. All witnesses that refer either to the translation of the lost 116 pages or to our own current Book of Mormon (Emma Smith, Martin Harris, and members of the Whitmer family) openly observed this translation process—one without a curtain or blanket separating Joseph from his scribe. In fact, according to Emma, the plates were wrapped up and not directly used.

On the basis of the witnesse’s statements, we can identify the following stages in the translation process:

  1. Joseph Smith sees (in some way) the English text;
  2. Joseph reads off the text to the scribe;
  3. the scribe hears the text;
  4. the scribe writes the text.

Evidence from the original and printer’s manuscripts suggests that the only revealed stage in the translation process was what Joseph Smith himself saw by means of the interpreters. Witnesses seemed to have believed that Joseph actually saw an English text in the interpreters, but it is possible that Joseph saw the text, so to speak, in his “mind’s eye”. In any event, all other stages—from Joseph Smith reading off that text to the scribe writing it down—potentially introduced human error and had to be carefully monitored.

The idea of a revealed text raises an important question: To what degree did the Lord control the dictation of the Book of Mormon? There appear to be three possible kinds of control over the dictation of the text:

  1. Loose control: Ideas were revealed to Joseph Smith, and he put those ideas into his own language (a theory advocated by many Book of Mormon scholars over the years);
  2. Tight control: Joseph saw specific words written out in English and read them off to the scribe—the accuracy of the resulting text depending on the carefulness of Joseph and his scribe;
  3. Iron-clad control: Joseph (or the interpreters themselves) would not allow any scribal error to remain (including the spelling of common words).

One can also conceive of mixtures of these different kinds of control. For instance, one might argue for tight control over the spelling of specific names, but loose control over the English phraseology itself.

A number of statements from the witnesses definitely show that virtually all of them believed in the iron-clad theory:

Joseph Knight (autograph [between 1833 and 1847]):

But if it was not Spelt rite it would not go away till it was rite, so we see it was marvelous.

Emma Smith (Edmund Briggs interview, 1856):

When my husband was translating the Book of Mormon, I wrote a part of it, as he dictated each sentence, word for word, and when he came to proper names he could not pronounce, or long words, he spelled them out, and while I was writing them, if I made a mistake in spelling, he would stop me and correct my spelling, although it was impossible for him to see how I was writing them down at the time.

Martin Harris (Edward Stevenson’s 1881 account):

By aid of the seer stone, sentences would appear and were read by the Prophet and written by Martin, and when finished he would say, “Written,” and if correctly written, that sentence would disappear and another appear in its place, but if not written correctly it remained until corrected, so that the translation was just as it was engraven on the plates, precisely in the language then used.

David Whitmer (Eri Mullin interview, 1874):

. . . the words would appear, and if he failed to spell the word right, it would stay till it was spelled right, then pass away; another come, and so on.

Evidence in the Manuscripts We now turn to the original manuscript and what it can specifically tell us about the translation process. In a number of instances it provides valuable support (or at least consistent evidence) for what the witnesses saw. This manuscript also provides valuable evidence for procedures that none of the witnesses described in any of their statements. The five conclusions that follow are based on evidence in the original manuscript and, to a lesser extent, on corroborating evidence in the printer’s manuscript.

The original manuscript was written from dictation

All witnesses of the translation stated that Joseph Smith dictated the text of the Book of Mormon. This claim is supported by certain errors in the original manuscript which clearly resulted from the scribe mishearing what Joseph had dictated. These errors were not the result of the scribe misreading while visually copying from some other manuscript or even from a copy of the King James Bible. As an example of this kind of error, consider the difficulty the scribe had in hearing the difference between and and an. In 1 Nephi 13:29 of the original manuscript the scribe (not yet identified, but designated as scribe 2) wrote down the following:

& because of these things which are taken away out of the gosple of the Lamb & exceeding great many do stumble

Obviously, scribe 2 misheard “an exceeding great many” as “and exceeding great many”. The scribe’s use of the ampersand (&) shows that the error was not based on visual similarity. Hearing an, the scribe interpreted it as the casual speech form an’ for and.

In contrast to this error from the original manuscript, the errors that are found in the printer’s manuscript show that this second manuscript was visually copied. As Oliver Cowdery was copying from the original manuscript into the printer’s manuscript, he sometimes incorrectly read the original manuscript. In many cases, the error leads to a more difficult reading, as in the following example: •Alma 30:52

original manuscript:

yea & I always knew that there was a God

printer’s manuscript:

yea & I also knew that there was a God

This error was due to visual similarity between the words always and also. This kind of error does not appear in the original manuscript, because the scribes were not copying from another written source but were hearing the words dictated by Joseph Smith.

Joseph Smith was working with at least twenty words at a time

There is some evidence in the original manuscript to suggest the minimal amount of text Joseph Smith viewed as he was dictating. Consider, for instance, the evidence from scribal anticipations. Frequently the scribe, in attempting to keep up with Joseph’s dictation, jumped ahead of the actual text. In the following example Oliver Cowdery anticipated the text in Alma 56:41 of the original manuscript:

& it came to pass that again when the light of the morning came we saw the Lamanites upon us

(Here and elsewhere in this article, the angled brackets refer to a crossout.) This example suggests that Joseph and Oliver started out together, but by the time Oliver finished writing “& it came to pass that again,” Joseph had moved along far enough that he was then dictating “we saw the Lamanites upon us” and Oliver started to write that down when he realized he had skipped the intervening text (“when the light of the morning came”), so he immediately crossed out “we saw the Lamanites” and wrote the correct sequence, possibly with Joseph repeating the correct text for him. If this explanation is correct, then it indicates that Joseph had at least twenty words in view as he was dictating.

Joseph Smith could see the spelling of names

Several witnesses to the translation process claimed that Joseph Smith sometimes spelled out names to the scribe. And we find evidence in the original manuscript in support of this process. Frequently the first occurrence of a Book of Mormon name is first spelled phonetically, then that spelling is corrected; in some instances, the incorrect spelling is crossed out and followed on the same line by the correct spelling, thus indicating that the correction was an immediate one. For example, in Alma 33:15 the text of the original manuscript reads as follows:

for it is not written that Zenos alone spake of these things but Zenoch also spake of these things

Oliver Cowdery first wrote Zenock using the expected ck English spelling for the k sound when preceded by a short vowel. But then Oliver crossed out the whole word and immediately afterwards, on the same line, wrote Zenoch, thus indicating that the spelling agrees with the biblical name Enoch. This example also suggests that Joseph Smith spelled out the ch sequence for Oliver, although it is possible that Joseph could have repronounced the ch sequence with the incorrect ch sound rather than with the correct k sound in order to help Oliver get it down right.

But there are also examples for which it is impossible to find a repronunciation that will guarantee the correct spelling. For instance, in Helaman 1:15 Oliver Cowdery first wrote the name Coriantumr phonetically, as Coriantummer, then he crossed it all out and wrote out the correct spelling, Coriantumr:

& they were lead by a man whose name was Coriantumr

In this case, no matter how slowly or carefully Joseph Smith might have repronounced Coriantumr, it would have been impossible for him to have indicated that there was no vowel between the m and r at the end of the name except by actually spelling out the separate letters m and r. Nor could Oliver have guessed this spelling since no word (or name) in English ends in mr. In fact, Oliver ends the correct spelling Coriantumr with a large flourish on the final r, which Oliver produces nowhere else in either the original or the printer’s manuscript. This addition may reveal Oliver’s frustration at having to guess at such a weird spelling.

The scribe repeated the text to Joseph Smith According to David Whitmer, a dictation of words was followed by a checking sequence in which the scribe would read back the text to Joseph Smith. If an error was discovered, Joseph would presumably then read off the correct text once more until he was satisfied that the scribe had written it down correctly.

The specific evidence from the original manuscript is consistent with the claim that the scribe read back what had been written. In such a process, Joseph Smith would be checking what he was hearing from the scribe against what he was viewing in the interpreters. But such agreement would not guarantee the accuracy of the manuscript. For instance, the name Amalickiah was frequently spelled as Ameleckiah in the original manuscript. This misspelling shows that Joseph placed the stress on the first syllable, not the second. If the stress had been on the second syllable, the scribe would have consistently spelled the second vowel with the letter a. Therefore, given Joseph’s pronunciation of Amalickiah with stress on the first syllable, there would be no way for him to detect the incorrect spelling Ameleckiah when the scribe, in reading back the text, pronounced the name as Joseph did (with stress on the first syllable). This same difficulty in hearing word differences applies to phonetically similar words (such as and versus an). Many of the undetected errors that remain in the original manuscript could not have been caught when read back because there was little if any difference in pronunciation.

Corrections in the original manuscript are also consistent with a repetition sequence. The clear majority of changes in the original manuscript were made immediately; that is, the scribe caught the error during Joseph Smith’s initial dictation. Evidence for these immediate corrections include: corrections following on the same line, erasures showing ink smearing (since the ink had not yet dried), or corrections above the line or insertions in the line with no change in the level of ink flow or difference in the quill. These immediate corrections also include numerous cases where the crossed-out word is only part of the intended word or is obviously miswritten.

On the other hand, there are also numerous changes that are consistent with a process of correcting errors found while reading back the text. In these instances, the original form is complete and the error is usually not obvious (that is, the reading is not a difficult reading); the correction is supralinear or inserted in line, but there is no erasure, only a crossout, and the level of ink flow is usually different.

In addition, some corrections were made well after the repetition sequence. In fact, a few of the later corrections in the original manuscript were apparently made when the printer’s manuscript was being copied from the original manuscript, or even later when sheets of the 1830 edition were being proofed. Sometimes the change was by a different scribe or in a different medium (pencil rather than ink). In virtually every case these few corrections eliminated difficult readings in the original manuscript.

The word chapter and the corresponding chapter numbers were not part of the revealed text

Evidence from both the original and printer’s manuscripts shows that Joseph Smith apparently saw some visual indication at the end of a section that the section was ending. Although this may have been a symbol of some kind, a more likely possibility is that the last words of the section were followed by blankness. Recognizing that the section was ending, Joseph then told the scribe to write the word chapter, with the understanding that the appropriate number would be added later.

There is considerable evidence in both manuscripts to support this interpretation. First, the word chapter is never used by any writer in the text itself, unlike the term book, which is used to refer to an individual book in the Book of Mormon (such as the book of Helaman) as well as a whole set of plates (such as the book of Nephi, meaning the large plates of Nephi).

Second, chapters are assigned before the beginning of a book. For instance, in the original manuscript, we have the following at the beginning of 2 Nephi:


〈Chapter VIII〉

second            Chapter I

The ^ Book of Nephi ^ An account of the death of Lehi . . .


Oliver Cowdery first wrote Chapter at the conclusion of the last section in 1 Nephi—that is, at the conclusion of Chapter VII in the original chapter system; our current chapter system dates from Orson Pratt’s 1879 edition of the Book of Mormon (which has 22 chapters in 1 Nephi). At this point, Joseph Smith had no indication that a new book was beginning. All he could see was the end of Chapter VII (namely, the words “and thus it is Amen” followed probably by blankness or maybe a special symbol). Later, when Oliver was adding the chapter numbers, he first assigned the Roman numeral VIII to this first chapter of 2 Nephi. But when he realized that this was actually the beginning of a new book, he crossed out the whole chapter designation and inserted (with slightly weaker ink flow) “Chapter I” after the title of the book, which originally was simply designated as “The Book of Nephi”. Later he realized that there was more than one book of Nephi, which led him to also insert the word second (with considerably heavier ink flow).

This system of assigning chapters also explains why the two manuscripts have chapter numbers assigned to the short books found at the end of the small plates (Enos, Jarom, Omni, and the Words of Mormon) as well as 4 Nephi. These books contain only one section, but at the beginning of each of these short books, Joseph Smith apparently had no knowledge that this was the case. This fact further shows that Joseph himself did not know in advance the contents or structure of the text.

Probably the strongest evidence that the word chapter is not original to the revealed text is that the chapter numbers are assigned later in both manuscripts. The numbers are almost always written with heavier ink flow and more carefully. In many cases, Oliver Cowdery took time to add serifs to his Roman numerals. On the other hand, his Chapter is always written rapidly and with the same general ink flow as the surrounding text. In the printer’s manuscript, at the beginning of Chapter XVII in Alma (now the beginning of Alma 36), the Roman numeral XVII was written in blue ink, not the normal black ink. This example clearly suggests that this part of the original manuscript itself did not yet have chapter numbers assigned to it when Oliver started to copy it, perhaps six months after it had been dictated.

Evidence in the Original Text The evidence for loose control seems to rely heavily upon the notion that the nonstandard use of English in the original text could not have come from the Lord (since some suppose he only speaks “correct” English). The use of dialectal English, in this view, is said to be Joseph Smith’s contribution, thus by inference the Lord only gave Joseph Smith ideas, not specific words. However, Joseph’s practice of spelling out names definitely suggests that a theory of loose control must be revised in some way: Joseph had some view of the specific spelling for names, in particular, names with impossible spellings for English literates.

Non-English Hebraisms in the Original Text One of the interesting complexities of the original English-language text of the Book of Mormon is that it contains expressions that appear to be uncharacteristic of English in all of its dialects and historical stages. These structures also support the notion that Joseph Smith’s translation is a literal one and not simply a reflection of either his own dialect or the style of early modern English found in the King James Version of the Bible.

For instance, in the original text of the Book of Mormon we find a number of occurrences of a Hebrew-like conditional clause. In English, we have conditional clauses like “if you come, then I will come”, with then being optional. In Hebrew this same clause is expressed as “if you come and I will come”. In the original text of the Book of Mormon, there were at least fourteen occurrences of this non-English expression. One occurrence was removed in 1 Nephi 17:50 as Oliver Cowdery was producing the printer’s manuscript by copying from the original manuscript:

1 Nephi 17:50

original manuscript:

if he should command me that I should say unto this water be thou earth and it shall be earth

printer’s manuscript:

if he should command me that I should say unto this water be thou earth

it should be earth

The remaining thirteen occurrences were all removed by Joseph Smith in his editing for the second edition of the Book of Mormon, published in 1837 in Kirtland, Ohio. One example comes from the famous passage in Moroni 10:4 where Joseph removed the extraneous and in the second edition of the Book of Mormon:

and if ye shall ask with a sincere heart with real intent having faith in Christ and he will manifest the truth of it unto you by the power of the Holy Ghost

This original use of and is not due to scribal error, especially since this same if-and expression occurs seven times in one brief passage:

Helaman 12:13—21

  1. yea and if he sayeth unto the earth move and it is moved
  2. yea if he say unto the earth thou shalt go back that it lengthen out the day for many hours and it is done . . .
  3. and behold also if he sayeth unto the waters of the great deep be thou dried up and it is done
  4. behold if he sayeth unto this mountain be thou raised up and come over and fall upon that city that it be buried up and behold it is done . . .
  5. and if the Lord shall say be thou accursed that no man shall find thee from this time henceforth and forever and behold no man getteth it henceforth and forever
  6. and behold if the Lord shall say unto a man because of thine iniquities thou shalt be accursed forever and it shall be done
  7. and if the Lord shall say because of thine iniquities thou shalt be cut off from my presence and he will cause that it shall be so

These examples of the if-and construction in the original text suggest that Joseph Smith did not simply get the idea of a conditional construction in his mind and then put it into his own words. If that had been the case, he should have translated that idea using the English if-then construction, possibly without the then, but certainly without the connective and. The multiple occurrence of the non-English if-and construction suggests that even the word and was controlled for.

Consistency of Phraseology in the Original Text There is substantial evidence within the text itself for tight control over specific words, phrases, and sentences of English. For instance, John Welch and Tim Rathbone have pointed out an interesting case where the Book of Mormon makes the same identical (nonbiblical) quote in widely separated parts of the text. The example they give is based on Lehi’s vision of the kingdom of God as found in 1 Nephi 1:8 and Alma 36:22:

•1 Nephi 1:8

and he thought he saw God sitting upon his throne

surrounded with numberless concourses of angels in the attitude of singing and praising their God

•Alma 36:22

and methought I saw

even as our father Lehi saw God sitting upon his throne

surrounded with numberless concourses of angels in the attitude of singing and praising their God

This identity of quotation provides striking support for a theory of tight control over the translation.

In numerous cases we find that the original Book of Mormon text is consistent in its phraseology, but subsequent copying errors or changes due to editing have created exceptional phrases uncharacteristic of the text. For example, the Book of Mormon originally referred to people pressing their way forward, but never feeling their way, as in Lehi’s dream of the tree of life:

•1 Nephi 8:21

and I saw numberless concourses of people many of whom were pressing forward that they might obtain the path

•1 Nephi 8:24 (2 times)

and it came to pass that I beheld others pressing forward . . . and they did press forward through the mists of darkness

•1 Nephi 8:30 (2 times)

behold he saw other multitudes pressing forward . . . and they did press their way forward

Yet in one case Oliver Cowdery accidentally miscopied the correct pressing as feeling:

•1 Nephi 8:31

original manuscript:

and he also saw other multitudes pressing their way towards that great and spacious building

printer’s manuscript:

and he also saw other multitudes feeling their way towards that great and spacious building

Other examples of “pressing forward” are found in 2 Nephi 31:20 (2 times) and Ether 14:12. But there are no scriptural uses of the modern expression “feeling one’s way”, in either the Book of Mormon or anywhere else (including the King James Bible)—except by accident here in 1 Nephi 8:31.

A second example is in 1 Nephi 12:18, where the current text refers to “the word of the justice of the eternal God”. Yet in the original manuscript, this passage refers to “the sword of the justice of the eternal God”:

•1 Nephi 12:18

original manuscript:

and a great and a terrible gulf divideth them yea even the sword of the justice of the eternal God

printer’s manuscript:

and a great and a terrible gulf divideth them yea even the word of the justice of the eternal God

Oliver Cowdery misread sword as word when he was copying the original manuscript into the printer’s manuscript, yet neither he nor subsequent editors noticed that the Book of Mormon nowhere else refers to the word of the justice of God, only to the sword of the justice of God:

“the sword of his justice” (Alma 26:19)

“the sword of justice” (Alma 60:29)

“the sword of justice” (Helaman 13:5, 2 times)

“the sword of my justice” (3 Nephi 20:20)

“the sword of his justice” (3 Nephi 29:4) “the sword of the justice of the eternal God” (Ether 8:23)

In fact, the phraseology in the last example is precisely the same as the original reading in 1 Nephi 12:18.

Conscious editing has also created exceptional readings in the text. For instance, in his editing for the second (1837, Kirtland) edition of the Book of Mormon, Joseph Smith changed one case of had hope to had hoped:

•Jacob 5:46

printer’s manuscript and 1830 edition:

and these I had hope to preserve to have laid up fruit thereof against the season unto mine own self

1837 edition:

and these I had hoped to preserve to have laid up fruit thereof against the season unto mine own self

By adding the d to hope, Joseph reinterpreted the original main verb had and the direct object noun hope as a past perfect verb phrase (had hoped). Nonetheless, this change created an exception to the rest of the Book of Mormon text. Elsewhere, there are 17 examples of the main verb have taking the noun hope as a direct object, including two examples of had hope, one of which is close by:

•Jacob 7:5

and he had hope to shake me from the faith

•Mormon 6:4

and here we had hope to gain advantage over the Lamanites

Except for the change in Jacob 5:46, there are no other examples of the verb hope occurring with the perfect auxiliary have—that is, there are no other examples of have hoped, has hoped, hath hoped, or had hoped in the entire Book of Mormon.

Conclusion Evidence from the original manuscript supports the traditional belief that Joseph Smith received a revealed text by means of the interpreters. This idea of a controlled text originates with statements made by witnesses of the translation. The evidence from the original manuscript, when joined with internal evidence from the text itself, suggests that this control was tight, but not iron-clad. The text could be “ungrammatical” from a prescriptive point of view, but the use of nonstandard English is not evidence that the text was not being tightly controlled, or that it did not come from the Lord, who apparently does not share our insistence on “proper English” (see D&C 1:24). In fact, the occurrence of non-English Hebraisms such as the if-and construction strongly suggests that the text was tightly controlled down to the level of the word at least. This tight control is also supported by the consistent phraseology in the original text. And the spelling of names such as Coriantumr suggests that control could be imposed down to the very letter.

All of this evidence (from the witnesse’s statements, the original manuscript, the printer’s manuscript, and from the text itself) is thus consistent with the hypothesis that Joseph Smith could actually see (whether in the interpreters themselves or in his mind’s eye) the translated English text—word for word and letter for letter—and that he read off this revealed text to his scribe. Despite Joseph’s reading off of the text, one should not assume that this process was automatic or easily done. Joseph had to prepare himself spiritually for this work. Yet the evidence suggests that Joseph was not the author of the Book of Mormon, not even its English language translation, although it was revealed spiritually through him and in his own language.

Note: This paper is a revision of my article “Translating the Book of Mormon: Evidence from the Original Manuscript” in Book of Mormon Authorship Revisited: The Evidence for Ancient Origins, ed. Noel B. Reynolds (Provo, Utah: FARMS, 1997), pp. 61-93. See that article for additional examples and complete references.