[Page 53]Abstract: Joseph Smith made various refining changes to the Book of Mormon text, most of them minor grammatical in nature. However, one type of textual change has been virtually unstudied in Book of Mormon scholarship: extemporaneous change that was present the moment Smith dictated the original text to his scribes. This type of change appears to have been improvisational, a fix or repair made in the middle of a thought or expression. I study these improvisations in depth — when they appeared historically, their purpose, and their authorship. The evidence of Article One points to ancient authors and editor-engravers whose extemporaneous changes appeared during the early layers of the Book of Mormon’s construction. But how were these improvisations affected by later contributors? In this paper, Part 2, we study the improvisational work of Moroni as compiler, finishing-editor, and conservator, and of Joseph Smith as modern translator. The findings tell us much about the Book of Mormon as scripture, and about the construction and compilation of scripture by ancient editors and authors.

This article is the second in a series on extemporaneous change and improvisation in the Book of Mormon, studying change that was present the moment Smith dictated the original text to his scribes. This change appears to have been improvisational, a fix or repair made in the middle of a thought or expression. We explore extemporaneous change as it appears in successive layers or strata of the Book of Mormon’s [Page 54]compilation and construction (see Figure 1). Article One studied the first three levels of Figure 1. Beginning at the bottom of the figure, some of the oldest and earliest cases of improvisation in the Book of Mormon appear in the quoted ancient texts of the seminal prophets of ancient religion (Layer 1 of Figure 1), such as Moses and Isaiah.

We then studied the improvisations found in the writings of authors and embedded authors who were prophets, judges, kings, military leaders, and teachers — such as Alma, Benjamin, Abinadi, Limhi, and other authors (Level 2 of Figure 1).¹ And we studied the improvisations of the editor-engravers themselves, including Mormon, Nephi, and Moroni [Page 55](Level 3 of Figure 1), the managing editors of the entire Book of Mormon corpus serving complementary and contrasting roles in its construction and development.

Now in this article we study improvisation at the final two levels of the Book of Mormon’s construction and delivery — Moroni’s final compilation and conservancy of the Book of Mormon codex, and Joseph Smith’s translation resulting in the original manuscript of 1829 and subsequent revisions found in the 1830, 1837, and 1840 editions. These improvisations are important because they enable us to explore the composition, design, and construction of the Book of Mormon at a deep and elemental level using the corrective tools that all authors use. Were these improvisations the work of many authors, a few authors, or one single author, and was their origin ancient or modern?

I use the terms improvisation and extemporaneous change interchangeably and measure improvisation using corrective conjunction phrases (CCPs) that are marked by a phrase or sentence that the author or speaker corrects or modifies “on the fly,” as it were, using a conjunction or hybrid conjunction (e.g., “or,” “or rather,” “or in fine”) followed by a correction, amplification, or explanation. It is helpful to recall the three types of improvisation:

• Type 1: Correcting an apparent error or mistake.
• Type 2: Amplifying, clarifying, or augmenting the meaning of the text.
• Type 3: Explanatory, providing a helpful literal translation of an unknown word or concept.

Article One confirmed statistically that the improvisational patterns we see in the Book of Mormon are broadly driven by its editor-engravers and its embedded authors (Levels 2 and 3, Figure 1); extemporaneous change is the product of not one, but multiple actors. However, these early actors were not the last to handle the engravings. Therefore, we turn to the next higher level of construction and delivery, Level 4, to Moroni as the compiler, finishing-editor and conservator of the final Book of Mormon scriptural canon to be delivered and translated in the modern day.

Moroni’s Use of Extemporaneous Change

In Article One we found that Moroni’s work as an author never contains improvisation. However, we do find improvisation in Moroni’s work [Page 56]as an editor-engraver, appearing in the embedded texts of two earlier authors: Mormon’s two epistles to his son Moroni containing two CCPs, and the Jaredite book of Ether containing four CCPs. Moroni defined the canon of Book of Mormon scripture by ultimately finalizing the composition, organization, and presentation of its independent source texts as a complete codex, including, for example, the abridgement of Mormon (including the lost book of Lehi), the liturgical teachings of Christ at Bountiful (Moroni 1–6), the Jaredite plates of Ether, the records of Nephi, and the letters of Mormon. We therefore must ask whether or not Moroni may have been responsible for some, many, or most of the improvisations we see while reading the text because of his role as final finishing editor of the Book of Mormon corpus.

According to Grant Hardy, Moroni shows an inclination toward editorial license with the texts of Ether, which leads us to wonder whether he might show a similar propensity to fix, modify, or even author some of the 170 extemporaneous changes we see in the modern text. This description by Hardy illustrates Moroni’s editorial propensity: “The challenge for Moroni, then, was to Christianize Ether’s book, making it appear more theologically consistent with his father’s history than it actually was. He does this by working an additional eighteen references to Christ’s name into his comments on the [much older] Jaredite record.”² Regarding Moroni’s paraphrasing, Hardy notes, “It is telling, however, that Ether’s message is always conveyed indirectly, and we may wonder how close the paraphrase was, or if there was any embellishment in Moroni’s summary.”³ John Welch said, “The fact that Moroni felt free to insert his own material into his abridgment of the book of Ether indicates that, in general, he was not attempting to produce a technically rigorous version of Jaredite history.”⁴

Did Moroni apply some of these embellishment tendencies to fix or modify the six improvisations found in the embedded author texts of his editorial narratives — in Mosiah’s translation of the book of Ether, and in Mormon’s letters to his son Moroni? As shown in the three examples below, these improvisations in fact do not appear to exhibit embellishment; they appear to be minimal changes, missing the creative impulse that Hardy notes of Moroni:⁵

And they did also carry with them deseret,
which by interpretation is a honey bee.
Earliest Text, 675; Ether 2:3 (Type 3 Explanatory)

[Page 57]and that the Lord God would send or bring forth
another people to possess the land,
by his power,
Earliest Text, 703; Ether 11:21 (Type 2 Amplifying)

And it came to pass that he came to the waters of Ripliancum,
which by interpretation is large or to exceed all;
Earliest Text, 714; Ether 15:8 (Type 3 Explanatory, Type 2 Amplifying)

These improvisations from the Jaredite record of Ether are consistent with what we might expect from a translator’s hand (Mosiah’s) of a newly discovered record of an unknown people — half are Type 3 Explanatory, an otherwise infrequently used form of extemporaneous change. What about the two improvisations from Mormon’s epistles to Moroni, embedded in Moroni’s engravings? These also are simple and fit the same improvisational pattern of Mormon’s other extemporaneous changes (57 percent Type 1 Correcting, 41 percent Type 2 Amplifying, 3 percent Type 3 Explanatory), studied in Article One:

Wherefore he that is not condemned
—or he that is under no condemnation—cannot repent,
Earliest Text, 728; Moroni 8:22 (Type 1 Correcting)

Behold, the pride of this nation—or the people of the Nephites—
hath proved their destruction except they should repent.
Earliest Text, 729; Moroni 8:26 (Type 2 Amplifying)

This handful of examples, admittedly a small sample, give no evidence of embellishment or intervention, suggesting that Moroni as an editor-engraver likely seems to retain intact the improvisations encountered from the authors whose texts lay before him, from Mormon and the book of Ether. We can see why Moroni might have retained, rather than modified, the improvisations he found when considering the challenge of working with an ancient translated record like the Jaredite texts of Ether. Hardy said:

Moroni’s Christianizing of the Jaredite experience is subtle and consistent, but his sixteen-year writer’s block becomes more understandable when we imagine him reading Mosiah’s translation of the record of a non-Israelite, non-Christian society that had missed out on the covenant at Sinai, and then comparing that with the very last words written by his father … What could the two records possibly have to do with each other? Mormon had not been [Page 58]able to see his way forward to a solution, yet he was counting on Moroni to fulfill his editorial promise to incorporate the Jaredites into his comprehensive history of the Nephites.⁶

The Evident Identity and Early Origin
of Extemporaneous Changes

But what about Moroni’s involvement with the full sample of 170 extemporaneous changes? To answer this question, let us extend our exploration by drilling down into the evident authorship of the remaining extemporaneous changes and see if there is evidence that they appeared later in the Book of Mormon’s chain of authoring, editing and construction (Figure 1), or earlier in history. Could a late-stage finishing editor like Moroni have been responsible for the remaining 164 improvisations, those appearing in the larger corpus of Mormon’s and Nephi’s editorial engravings, and the personal engravings of Jacob’s priestly lineage appearing on the Small Plates of Nephi?

It is impossible to pinpoint when a given extemporaneous change might have been engraved, but it should be possible to ascertain approximately where it occurred in the chain of authoring, editing, and construction (see Figure 1). There is additional evidence in the textual context, and sometimes in the extemporaneous change itself, that enables us to infer whether the improvisation likely originated earlier in the chain of authoring, editing, and construction with ancient authors, or later with late-stage finishing contributors like Moroni or Joseph Smith. Therefore, let us look at the extemporaneous changes themselves, how they are constructed, and what they tell us about the evident identity of their possible originators.

First, the Book of Mormon appears unmistakably as a collection of personal writings; similar to many biblical books it is usually straightforward to identify authorship as given in the text. Narrators write either in the first person voice or narrate in the third person, but the narrator’s identity is rarely lost in the narrative. This applies as well to extemporaneous changes. For example, in this paraphrased improvisation Nephi clearly identifies Lehi as the original voice, but the speaker at this point in the chain of authoring, editing, and construction appears to be Nephi (I treat paraphrases such as this as originating with the speaker paraphrasing):⁷

And it came to pass that while my father tarried in the wilderness,
he spake unto us, saying:
[Page 59]Behold, I have dreamed a dream,
or in other words, I have seen a vision.
Earliest Text, 21, 1 Nephi 8:2

Another extemporaneous change found in the book of Jarom appears in the author’s first person voice:

Wherefore we withstood the Lamanites
and swept them away out of our lands
and began to fortify our cities or whatsoever place of our inheritance.
Earliest Text, 184; Jarom 1:7

In Article One we studied improvisational patterns statistically by examining editorial voice. Consider some additional findings regarding the five types of editorial voice evident in the improvisations found in the edited works of the three editor-engravers. In Table 1 we can see that 19 percent of the improvisations found in Nephi’s edited works are in the editor’s third person voice, which can then be identified by the context of the textual passage; the remaining four types, totaling 81 percent, are identified directly in the first person or paraphrased voice of the original author. For Mormon’s edited works, 47 percent of improvisations are in the editor’s third person voice, and the remaining types, totaling 53 percent, are in the first person or paraphrased voice of the original author. And for Moroni’s edited works, 50 percent of improvisations are in the editor’s third person voice, and the remaining two types, comprising 50 percent, are identified in the first person or paraphrased voice of the original author.

In other words, the vast majority of the extemporaneous changes in the Book of Mormon are traceable to the voice of the author identified in the text. Of course we are witnessing the words of these authors after many centuries of recording, redacting, editing, and translating, raising questions of accuracy and identity. Yet we encounter the same issues with biblical texts and accept their authenticity with faith and advances in scholarship. For example, Jeremiah speaks in the first person voice, but his words were preserved through the efforts of his disciple Baruch ben Neriah working as scribe and editor (Jeremiah 36). Ezekiel’s words appear in his first person narratives, yet the book’s authorship and authenticity were questioned in the early twentieth century, only to be critically affirmed in recent decades as having been transmitted orally by the prophet and then preserved as a written and interpreted text by his school of followers.⁸[Page 60]

Corrective Signals as Identifiers of Extemporaneous Change

Another anecdotal marker of authorship is the type of corrective signal used to indicate that a change is being made in the narrative. Different authors use different corrective conjunction configurations as signals — some are complicated and clearly appear to be unique to that author, and others are simply routine and used by all authors. I discovered twelve different corrective conjunction forms that are used as corrective signals in the Book of Mormon. I have arrayed these twelve into an ascending hierarchy based on individuality, meaning used by few authors, versus commonality, used by many or most authors (see Figure 2).

The following example from Mormon’s writings shows a complicated and awkward corrective signal (bolded), appearing in the first person voice:

And behold, the city had been rebuilt,
and Moroni had stationed an army by the borders of the city.
And they had cast up dirt round about
to shield them from the arrows and the stones of the Lamanites.
For behold, they fought with stones and with arrows.
Behold, I said that the city of Ammonihah had been rebuilt.
I say unto you, yea, that it was in part rebuilt.
And because the Lamanites had destroyed it once,
because of the iniquity of the people,
they supposed that it would again become an easy prey for them.
Earliest Text, 451; Alma 49:3–4[Page 61]

[Page 62]Mormon dramatically interrupts the narrative to signal a change; and he is the only author to use this inimitable corrective signal. Note the diversity — and individuality — of corrective change signals in Figure 2. Nephi is the only author to use the corrective signal “I say, trample under their feet but I would speak in other words.” Alma is the only author to use the corrective signal “or I would say in other words.” He is also the only author to use the signal “or rather in other words” and “or I would say.” Four authors use the corrective signal “which being interpreted is” or “which by interpretation is;” five use “or in other words;” but then only Alma uses “or I would say;” and only Alma and Nephi use “or, in fine.” The corrective signal “or” is very common to all authors except Giddianhi.

Figure 2 also shows that some authors utilize a colorful diversity of corrective signals, appearing adept at improvisation and making extemporaneous changes. If the 170 extemporaneous changes were the work of one late-stage editor-engraver like Moroni, or one translator like Smith, then we might expect this one person to repeatedly fall back on the same, or similar, recurring corrective signal. But that is not the case. Alma uses nine different corrective signals, ranging from “or I would say in other words” to simply “or,” with broad variation in between. As an editor-engraver Nephi uses five different corrective signals, and Mormon uses five.

Especially striking is how palpably personal some of these mistakes and repairs are, and how uniquely each portrays the apparent traits and personality of the author, almost like a portrait. This example from the words of Alma illustrates the colorful candidness of a personality evident in the extemporaneous change — note how the word “murdered” triggered attention in mid discourse, prompting an immediate CCP:

Yea, I saw that I had rebelled against my God
and that I had not kept his holy commandments.
Yea, and I had murdered many of his children
—or rather led them away unto destruction—
yea, and in fine, so great had been my iniquities
Earliest Text, 407; Alma 36:13-14, emphasis mine

[Page 63]An example from the words of Nephi shows a personal determination to make an argument forcefully and precisely, appearing fastidious about clarity and meaning:

yea, even the very God of Israel do men trample under their feet.
I say trample under their feet,
but I would speak in other words:
they do set him at naught and hearken not to the voice of his counsels.
Earliest Text, 60-61; 1 Nephi 19:7

These anecdotal examples show authorial personality, adding context and individuality to the evident authorship of the voices emerging from the text. The notion that a single late-stage editor like Moroni could have added these complexities seems daunting for the entire Book of Mormon corpus.

The Shape of Extemporaneous Change

There is a shape to how CCPs are constructed, which reveals innate differences in how the authors and engravers approach extemporaneous change. Usually CCPs have a three-part structure consisting of (1) a target phrase to change (before the corrective signal); (2) the corrective signal, usually a conjunction “or,” or hybrid conjunction such as “or rather;” and (3) the modification or fix (after the corrective signal). In the following example, found in the words of Benjamin, the target phrase before the change consists of six words (underscored), the corrective signal of one word (bolded), and the modification or fix of twenty words (italicized):

and serve the devil, which is the master of sin,
or which is the evil spirit, which hath been spoken of by our fathers,
he being an enemy to all righteousness.
Earliest Text, 205, Mosiah 4:14

By contrast, another example shows a much simpler fix, found in the writings of Mormon; the target phrase consists of two words (underscored), the corrective signal two words (bolded), and the modification or fix two words (italicized):

And thus he cleared the ground,
or rather the bank which was on the west of the river Sidon,
Earliest Text, 285, Alma 2:34

[Page 64]Table 2 shows the average words for each of these three structural dimensions, by author. There is considerable variation in how extemporaneous changes get constructed among the writings of the various Book of Mormon authors.

For example, on average the extemporaneous changes found in Limhi’s texts (fourth from the top) are quite wordy — 13.50 words per modification or fix (after the corrective signal), and 10.33 words per target phrase (before the corrective signal). By contrast, the extemporaneous changes found in Mosiah’s texts (in the book of Ether, second from the bottom) are quite simple — 2.75 words per modification or fix (after), and 1 word per target phrase (before).

An improvisation found in Giddianhi’s epistle to Lachoneous (top of Table 2) shows a very elaborate extemporaneous change — 39-word modification or fix (after, italicized), and 30-word target phrase (before, underscored):

Therefore I write unto you,
desiring that ye would yield up unto this my people
your cities, your lands, and your possessions,
rather than that they should visit you with the sword
and that destruction should come upon you.
Or in other words, yield yourselves up unto us and unite with us
and become acquainted with our secret works and become our brethren,
that ye may be like unto us, not our slaves,
but our brethren and partners of all our substance.
Earliest Text, 569; 3 Nephi 3:6-7

The Complexity of Extemporaneous Change

A more broadly encompassing measure of extemporaneous change is the complexity of the change; some improvisations are complicated and intricate, and others very simple. I assess the complexity of an extemporaneous change based on six indicators that generally pivot around the corrective signal:[Page 65]

1. Multiple Thought Phrases: Improvisations with just one thought phrase are simple, and those with multiple thought phrases are more complex.
2. Compound Sentence Structure: Some complex improvisations use compound sentences with multiple independent clauses — multiple subjects and verbs — joined by conjunctions.
3. [Page 66]Restate the Target Phrase: More complex improvisations often restate the entire target phrase as part of the modification or fix, while simple improvisations do not.
4. Complex Corrective Signals: Complex improvisations often contain an obvious or elaborate corrective signal that appears as an emphatic indicator that a deliberate change is coming.
5. Asymmetry: Asymmetric improvisations contain an unequal number of words in the modification or fix after the corrective signal compared to the target phrase before. Asymmetric improvisations are usually more complex.
6. Sentence, Phrase, or Word: The simplest improvisations change only a word, moderately complex improvisations change a phrase, and complex improvisations change entire sentences.

I subjectively sort the 170 extemporaneous changes into four categories of complexity: Category 1, High Complexity; Category 2, Moderate Complexity; Category 3, Moderate Simplicity; Category 4, High Simplicity. This is based roughly on the rule that High Complexity (Category 1) improvisations will qualify on approximately three or more of the above six indicators, Moderate Complexity (Category 2) on two or occasionally three of the six, Moderate Simplicity (Category 3) on one or occasionally two of the six, and High Simplicity (Category 4) on none of the six and they change only a word rather than a phrase or sentence.

This excerpt from the texts of author Limhi talking about slain prophet Abinadi shows a High Complexity (Category 1) extemporaneous change:

And because he saith unto them that Christ was the God the Father of all things
and saith that he should take upon him the image of man
and it should be the image after which man was created in the beginning
—or in other words, he said that man was created after the image of God
and that God should come down among the children of men
and take upon him flesh and blood
and go forth upon the face of the earth—
and now because he said this, they did put him to death.
Earliest Text, 214; Mosiah 7:27-28

[Page 67]Here, the improvisation is asymmetric, with multiple thought phrases or ideas, a compound sentence, and a moderately complex corrective signal, and the change encompasses multiple clauses — qualifying on five of six indicators of high complexity above; on only one indicator did it not qualify: The passage did not restate the complete target phrase.

The following improvisation from the texts of Mormon as author shows a High Simplicity (Category 4) extemporaneous change, fixing only one word of the target phrase:

And Teancum by the orders of Moroni caused
that they should commence in laboring,
in digging a ditch round about the land, or the city Bountiful.
Earliest Text, 469; Alma 53:3

Table 3 summarizes the proportional distribution of extemporaneous change for the texts of each author contributing to the Book of Mormon text, by complexity. The top six authors in the table, those authors we have focused on with five or more CCPs, are highlighted. Note that Alma’s texts exhibit a higher share of Category 1 High Complexity improvisations than Mormon’s, 19 percent versus 7 percent, and about the same as Nephi’s, 18 percent. Improvisations in the texts of Abinadi exhibit either High Complexity (Category 1), 60 percent, or Moderate Complexity (Category 2), 40 percent. Similarly, most of the improvisations found in Benjamin’s texts are either High Complexity, 29 percent, or Moderate Complexity, 57 percent. Looking further down Table 3, the texts from embedded authors Isaiah, Lehi, Jacob, Moses, Samuel, the angel of Nephi’s vision (1 Nephi 14), Giddianhi, Alma1 and Lamoni all exhibit mostly Moderate or High Complexity improvisations.

Summarizing the findings of the last several sections, the anecdotal evidence of narrative voice, the inimitable corrective signal used, the personality evident in some improvisations, and the differences in shape and in complexity, all suggest that authorship of these extemporaneous changes occurred at points earlier in the historical chain of authoring, editing, and construction of Figure 1, rather than one person such as Moroni making detailed extemporaneous changes at a later stage in the fourth or fifth century AD. If Moroni as editor retained intact the six improvisations he encountered from earlier authors in the Jaredite record and his father’s epistles, the additional evidence we have seen suggests that the remaining 164 improvisations may have been retained by Moroni as well when preparing and compiling the final codex for delivery at a future day.[Page 68]

Joseph Smith’s Use of Extemporaneous Change as a Translator

[Page 69]We next turn to the highest level of Figure 1, Joseph Smith’s use of extemporaneous change during the Book of Mormon translation. He and his assistants (scribes, and typesetter John Gilbert)⁹ are the final contributors in the chain of authoring, editing and construction, resulting in the original manuscript and printer’s manuscript of 1829, and then subsequent manuscripts — the 1830 Palmyra edition, the 1837 Kirtland edition, and the 1840 Nauvoo edition, represented in Level 5 of Figure 1. We should expect that extemporaneous change might be an indispensable tool for translating from an ancient text to a modern language. Yet Smith was never definitive about the translation process. At a conference in 1831 he “said that it was not intended to tell the world all the particulars of the coming forth of the book of Mormon, [and] also said that it was not expedient for him to relate these things.”¹⁰ In an 1842 publication, he described the translation process in ambiguous terms: “Immediately after my arrival there I commenced copying the characters of all the plates. I copyed a considerable number of them and by means of the Urim and Thummin I translated some of them which I did between the time I arrived at the house of my wife’s father in the month of December, and the February following.”¹¹ In other contexts, he described the process simply as by the “gift and power of God.”¹²

Nonetheless, as we noted at the beginning of Article One, after Smith had completed the Book of Mormon translation in 1829 he continued to refine the text by making many changes, most of them grammatical, in subsequent editions. Amongst all of this translating and editorial activity we must therefore look for evidence of Joseph Smith’s use of extemporaneous change during the translation period of 1829, and the ensuing eleven-year period involving subsquent editions.

Smith’s Documented Extemporaneous Change, Its Significance

First, there is only one documented instance of Smith’s actual use of a CCP; this occurred while he made changes to the Book of Mormon for the third edition (the Nauvoo Edition published in 1840). The 1829 Earliest Text edition quotes Old Testament prophet Isaiah (48:1) in First Nephi (Earliest Text, 63; 1 Nephi 20:1) without using a CCP. However, the 1840 edition adds a CCP that alters the meaning of the passage considerably, seen in the following:

Hearken and hear this, O house of Jacob,
who are called by the name of Israel,
[Page 70]and are come forth out of the waters of Judah,
(or out of the waters of baptism,)
who swear by the name of the Lord,
and make mention of the God of Israel
1840 Nauvoo Edition, 53, emphasis mine

The addition of the amplified CCP (signaled by “or”) is surprising in appearance, enclosed within parentheses by Smith himself — almost a tentative change, but surprising too because the added phrase changed the meaning of the text to say that, even in Old Testament texts such as Isaiah, ritual baptism had been part of the worship of ancient Israel.¹³ Significantly, on August 15, 1840, while speaking at the funeral of Seymour Brunson in Nauvoo, Smith introduced for the first time the doctrine of proxy baptisms for one’s deceased ancestors, so the idea of ancient baptism was very much on his mind.¹⁴

This insertion to the 1840 edition suggests that Smith was familiar with CCPs as a useful device for making extemporaneous textual changes. But did he know and use them during the translation period of 1828–29, or even later as he reviewed the text in preparation for the 1837 or 1840 editions? At the least, we should expect to see changes in subsequent editions to fix, improve, or remove CCPs involving apparent errors or mistakes.

Remarkably, I find virtually no changes in CCPs across editions in my sample. From the reconstructed 1829 Earliest Text, to the 1830 edition, to the 1837 edition, to the 1840 edition, the CCPs I studied appeared to remain surprisingly constant over time in 166 of the 170 improvisations. There are only three minor one-word changes, and one one-character change, in the 170 CCPs that appear across manuscripts and editions from 1829 to 1840, shown here in underscored text:

• 1 Nephi 10:5, “concerning this Messiah, of which he had spoken, or this Redeemer of the world” in the 1830 edition, changed to “of whom he had spoken” in the 1837 edition and retained in the 1840 edition.
• 2 Nephi 5:12, “and also the ball or the compass which was prepared for my father by the hand of the Lord” in the Earliest Text 1829 manuscript, changed to “and also the ball, or compass, which was prepared for my father, by the hand of the Lord” in the 1830 edition.
• [Page 71]Jarom 1:14, “according to the writings of the kings, or those which they caused to be written” in the 1830 edition, changed to “those which they cause to be written” in the 1840 edition.
• Alma 9:1, “preach again unto this people, or the people which was in the city of Ammonihah” in the 1830 edition, changed to “which were in the city of Ammonihah” in the 1837 edition and retained in the 1840 edition.

Clearly Smith noticed these improvisational passages because he made small changes in them, one in the 1830 edition, two in the 1837 edition, and one in the 1840 edition. Yet the changes he made were purely grammatical and minor within the improvisation itself — but the integrity of the original CCPs remained otherwise completely intact. All of the remaining 166 CCPs of my sample were untouched. This was the case even for Type 1 Correcting CCPs, those apparently designed to fix an apparent error or mistake. In other words, over time and across editions, it is clear that the prophet retained the original awkward translations of CCPs, with their seemingly obvious errors and mistakes. Let’s consider just three examples that could have been easily imroved.

Example 1: Kept intact across all manuscript editions was this improvisation found in the words of Jesus spoken at the temple in Bountiful:

Therefore if ye shall come unto me or shall desire to come unto me
Earliest Text, 599; 3 Nephi 12:23

This improvisation could have been easily changed to read simply as:  “Therefore, if ye shall desire to come unto me.”

Example 2: Kept intact was another improvisation found in a letter from Moroni1, captain of the Nephite armies:

Behold, Ammoron, I have wrote unto you somewhat
concerning this war which ye have waged against my people,
or rather which thy brother hath waged against them,
Earliest Text, 472; Alma 54:5

This improvisation could have been easily changed to read simply as: “Ammoron, I have written unto you somewhat concerning this war which thy brother hath waged against my people.”

Example 3: Kept intact was this improvisation found in Mormon’s account of the people of Anti-Nephi-Lehi:

[Page 72]And thus we see that they buried the weapons of peace —
or they buried the weapons of war for peace.
Earliest Text, 368; Alma 24:19

This improvisation could have been easily changed to read simply as: “And thus we see that they buried the weapons of war for peace.”

So what are we to make of these three examples? Consider a simple test: If you were a translator, like Smith, and had encountered these several awkward improvisations during later proofreading, would you not have made these simple changes (italicized above) to enhance the readability of the text for your audience? The answer is you probably would have (they seem intuitively obvious to a modern reader), if the improvisations had been yours to begin with — if they had originated with you. But if not, then you might have left them alone. This hypothesis seems most plausible to me: That Smith himself deemed that the improvisations found in these passages belonged to someone before him — to the editor who engraved them, or to the original author whose texts were embedded by the editor, or to the final finishing-editor and conservator (Moroni), or perhaps to unknown scribes or redactors working with the texts at some point during the chain of authoring, editing and construction (Figure 1).

Old Testament theologian Rudolph Smend noted how the presentation of the Ten Commandments of the Pentateuch has been left intact over time, despite two deficient “points of overlap” that should have been reordered or revised: “that between the prohibitions of adultery [seventh commandment] and of coveting one’s neighbour’s wife [tenth], and that between the prohibitions of theft [eighth] and the coveting of one’s neighbour’s possessions [tenth].” He continued: “A decent commission [of editors or redactors] would hardly have allowed the existing doublets to stand. It would have either established a consistent order and defined the forbidden acts more precisely or confined itself to a single prohibition in each case, thereby gaining space for other topics; for it was space, not topics, that was lacking.”¹⁵ Why was the Decalogue retained intact over time? One answer, he says: because of its “incomparable author [God],” “the direct expression of a single, sovereign will, without the intervention of anyone else.”¹⁶ The Decalogue’s words or grammar may change with translation, but its structure, and its presentation, remains unaffected through the ages out of deference to its ancient authorship.

In a similar way, did Joseph Smith not only defer to these inimitable improvisations because they belonged to others before him, but also revere or respect them as being personally expressive of the prophets [Page 73]and ancient figures whose names appeared with them in the scriptural narrative — as if he were witnessing the distinctive persona of prophet-historian Mormon in his writings, or the personality of King Benjamin in speech, or of Moroni1 at the command of his armies? Smith broadly reported spiritual experiences with various personalities in scripture (Christ, Moroni, John the Baptist, Peter, James, John, Elijah, Elias, Moses), and appeared to relate to them in personal terms. In August 1830 he dictated prophecy of his presence at a future sacramental meal, saying, “Wherefore marvel not for the hour cometh that I [Christ] will drink of the fruit of the Vine with you [Joseph] on the Earth,” and then expanded this text in the 1835 edition of the Doctrine and Covenants by adding:

and with Moroni, whom I have sent unto you to reveal the book of Mormon … and also with Elias, to whom I have committed the keys of bringing to pass the restoration of all things … and also John the son of Zacharias … which John I have sent unto you, my servants, Joseph Smith, Jr. and Oliver Cowdery, to ordain you unto this first priesthood which you have received … and also Elijah, unto whom I have committed the keys of the power of turning the hearts of the fathers to the children and the hearts of the children to the fathers … and also, with Joseph, and Jacob, and Isaac, and Abraham your fathers; by whom the promises remain; and also with Michael, or Adam, the father of all, the prince of all, the ancient of days: And also with Peter, and James, and John, whom I have sent unto you, by whom I have ordained you and confirmed you to be apostles and especial witnesses of my name.

“Moroni was more than a visionary encounter in the Prophet’s memory; he was a mentor who had a lasting influence,” I wrote in Schooling the Prophet: How the Book of Mormon Influenced Joseph Smith and the Early Restoration. “In his 1838 history Smith devotes nearly half the narrative of his life before 1827 to detailed descriptions of Moroni’s visitations that began September 21, 1823.”¹⁷ On June 26, 1844, the evening before the prophet’s martyrdom, “brothers Joseph and Hyrum Smith spent their final hours while imprisoned in Carthage Jail reading and listening to Book of Mormon historical narratives of persons experiencing divine rescue.” This included, according to eyewitness Dan Jones, “copious extracts from the Book of Mormon, the imprisonments and deliverance of the servants of God for the Gospels sake.”¹⁸ Smith and his associates appear to have felt a personal intimacy with the ancient personalities of scripture.

[Page 74]It is possible of course that Smith may have tacitly judged these improvisations as inconsequential and not worthy of correction, although some of the more noticeably awkward improvisations make this seem implausible. Consider for example these ungraceful improvisations from the Earliest Text edition that one would expect to have been changed over time and across editions, but were left untouched:

even so if ye shall keep the commandments of my son,
or the commandments of God which shall be delivered unto you by him,
ye shall prosper in the land,
Earliest Text, 198; Mosiah 2:31

they were again brought before the king
and their bands were loosed;
and they stood before the king
and was permitted — or rather commanded —
that they should answer the questions which he should ask them.
Earliest Text, 211; Mosiah 7:8

For behold, upon them the record of our wars are engraven
according to the writings of the kings,
or that which they caused to be written.
Earliest Text, 185; Jarom 1:14

The last example especially underscores the paradox of Smith’s treatment of extemporaneous change: the awkward improvisation itself was retained completely intact, except for one letter — “caused” (1829 Earliest Text) changed to “cause” (1830 edition).

In summary, explaining why Smith persistently looked past or tacitly refused to tamper with these CCPs is a paradox of Smith’s relationship with the Book of Mormon and should draw us into deeper exploration of how he interacted with it. He acts unusually detached from these awkward improvisations compared to the many refinements he made elsewhere in the text, signaling implicitly a deference to, or perhaps even reverence for, those voices who speak from the text, the ancient prophets, historians, priests, military leaders, editors, and writers — including the mistakes and imperfections endemic to their writings and sayings.

These extemporaneous changes capture not only the imperfections, but the idiosyncrasies of each writer, each orator, and each editor, containing as it were in the words before him and us, their indigenous imprints — traits, personalities, peculiarities, and dispositions at a [Page 75]granular level. Moreover, the paradox of untouched but defective CCPs versus the countless grammatical changes across successive editions is so striking, revealing a tacit dichotomy in Smith’s actions — never articulated but clearly observable in his quiet choices to act, or to not act. Some content clearly was deemed changeable and transitory, such as grammar, syntax, and style. But other content was deemed unchangeable and untouchable, including the vast majority of these extemporaneous changes that were retained word-for-word as embedded in the original translation narratives.

Joseph Smith and the Adaptive Learning Hypothesis

In an early version of this research I hypothesized that Joseph Smith might have personally authored the improvisations of the Book of Mormon during the translation process to ensure he achieved an acceptable translation. I tested this hypothesis by examining the sequential moments in time in which the improvisations appear during the translation period of 1829. I called it the adaptive learning hypothesis; it is useful to share it here because of the insights it leads to. Hardy describes the Book of Mormon’s modern translation as a product of one-time-through dictation: it “is the transcript of a single, extended oral performance.”¹⁹ If the final Book of Mormon manuscript were the product of one translator who applied traditional human-translation skills, then we should expect to see an adaptive learning effect as the translator learned how to translate better — with fewer mistakes, improvisational corrections, and fixes. That is, we should see more extemporaneous change earlier in the translating process, but then progressively less as the translator got better at translating over time.

I organized the entire Book of Mormon into 56 sequential text segments of equal size in the order in which Smith translated them, beginning in early April 1829 with the book of Mosiah, and concluding in late June 1829 with the small books of Jacob, Enos, Jarom, and Omni.²⁰ I then measured the number of CCPs for each sequential text segment. After first translating and then losing the book of Lehi in 1828, Smith began translating again about April 7, 1829, with Oliver Cowdery as scribe. Rather than return to the beginning of First Nephi, he instead continued translating at where the 116-page lost Lehi text had left off, starting “with the speech of Benjamin (Mosiah 1–6), [then] translated to the end of the book of Moroni in May, then translated the Title Page, and finally translated the small plates of Nephi (1 Nephi–Omni) and the [Page 76]Words of Mormon before the end of June.”²¹ The results of my sequential text analysis are shown in Figure 3a.

Figure 3a shows that extemporaneous change is frequently greater during the initial 22 sequential text segments, aligning with the earlier periods of the translation, and then diminishes until a spike appears again at text segment 44, aligning with First Nephi. I have superimposed a computer-drawn trend line reflecting the slope of the pattern. The trending pattern appears to lend support to the adaptive learning hypothesis that Smith as translator indeed got better at translating over time.

Although another explanation, of course, is that what we may see here is not the improvisational patterns of one late-stage translator (Smith or, as Roger Terry suggested, Moroni),²² but of earlier editor-engravers — Mormon, Nephi, and Moroni — who themselves got better at editing, engraving, and working with the ancient texts before them. In fact, when we look at sequential text segments separately for Mormon, and then for Nephi, we can see separate learning patterns with progressively declining CCPs for each, shown in Figure 3b, consistent with adaptive learning for these two ancient editor-engravers.²³

A Discovery — Contrasting Editorial Designs

However, a third explanation is actually more plausible and leads to a surprising discovery by looking not just at the incidence of CCPs across sequential text segments, but also the source of where the improvisations came from. In Article One we noted that many of the embedded author texts found in Mormon’s editorial narratives often exhibit higher levels of improvisation than Mormon’s personally authored texts (see Table 2 of Article One). So where are these high improvisation embedded texts placed in Mormon’s editorial narratives?

The answer: in Mormon’s editorial work the texts of high-improvisation authors are embedded earlier in his work, resulting in the pattern shown in Figure 3b (left side). For example, Benjamin (11.61 CCPs per 7,000 words), Limhi (16.90), Abinadi (12.47), Gideon (17.28), and Alma1 (6.12) all appear in the book of Mosiah, the earliest of Mormon’s editorial engravings (not including the lost Book of Lehi). Next appears a mix of embedded texts from high and low improvisation authors in the subsequent book of Alma, such as Amulek (4.40 CCPs per 7,000 words), Ammon (5.13), Anti-Nephi-Lehi (12.92), Alma (9.00), Moroni1 (6.83), and Pahoran (8.57).[Page 77]

[Page 78]

[Page 79]In the last three of Mormon’s edited books (Helaman, Third Nephi, and Fourth Nephi) appearing before the personally authored book of Mormon, we find low improvisation authors, such as Helaman (5.00 CCPs per 7,000 words), Samuel (4.55), and Jesus (2.06) — and Giddianhi (14.08, an outlier with one CCP in a small text).

The effect of these embedded author improvisations can be seen in Figure 3c, where I remove the improvisations of embedded authors, leaving only the improvisations of Mormon as author (left side). The resulting computer-generated trend line of the incidence of CCPs is much flatter. Gone are the elevated early improvisations we saw in Figure 3b in the books of Mosiah and early Alma. The trend line appears to decline slightly, but this is driven by higher improvisation located in the middle of Mormon’s work – in sequential text segments 10, 12, 19, and 20. Clearly, the editorial design evident in Mormon’s work reveals the placement of high-improvisation authors earlier in his work, and low-improvisation authors later.

For Nephi, the editorial design is nearly the opposite in subtle ways, as shown in Figure 3c (right), compared to Figure 3b (right). Removing the embedded author improvisations from Nephi’s chart panel has little impact on the trend line of the incidence of CCPs; that is, the improvisational trending for Nephi as author (Figure 3c) is about the same as for Nephi as editor-engraver (Figure 3b). Why does the trending look the same? Because the early elevated improvisations of Nephi’s texts belong to Nephi as author. Nephi as editor-engraver embeds the texts of low improvisation authors: for example, Lehi (2.99 CCPs per 7,000 words), Isaiah (3.52), the angel from one of Nephi’s visions (3.11), and his brother, chief priest Jacob (1.65) — and they get placed mostly later in his work. I summarize these contrasts in Figure 3d, which compares for both Nephi and Mormon their personal-author improvisations (solid bars) with their embedded-author improvisations (striped bars). Here we can see clearly the editorial designs of both authors, placing high improvisation texts early in their respective works, but exhibiting very different ways to go about it — Nephi relying on his own improvisational texts, and Mormon importing the improvisational texts of other embedded authors.[Page 80]

[Page 81]In summary, the adaptive learning hypothesis — that the Book of Mormon’s translator got better at translating over time using increasingly less extemporaneous change — is not supported, and is actually reversed when we dig deeper. But it leads to an important finding: The most plausible explanation for the progressively declining trend in extemporaneous change across the translation period of 1829 points not to the translator Joseph Smith (Level 5, Figure 1), nor to the compiler and finishing-editor Moroni (Level 4), but to the improvisational texts of the embedded authors (Level 2) of the Book of Mormon and their placement within the respective works of the editor-engravers (Level 3).

Conclusions

In Article One we saw statistical evidence confirming that the distinctive improvisational patterns found within the Book of Mormon corpus are the product of not one, or a few, but various different ancient authors, embedded authors, and editor-engravers (Levels 2 and 3, Figure 1) — each with a distinctive improvisation pattern signature.

Now studying adaptive learning leads to another discovery of contrasting editorial designs by the editor-engravers (Level 3, Figure 1), with the placement of high improvisation writings early in their respective editorial narratives — as if to ensure that the narratives delivered to future readers were not only accurate with respect to authorial intent, but also compelling so as to motivate continued reading and immersion in the narrative. The often vivid and inimitable improvisations — whether corrective, amplifying, or explanatory — are a tacit signal to the reader that something important is being said here because the author interrupts the narrative to make sure it is right.

These findings correlate with existing research on Israelite scribal practice. As we saw, in Nephi’s work the high improvisation texts that get placed early in his narrative are his own (including those found in his paraphrases of his father Lehi’s texts), which seems consistent with what we might expect from one “trained for a different profession,” as Gardner noted — a trained scribe in Israel.²⁴ Karel van der Toorn said, “The scholars of Israel … were scribes who had specialized in the classic texts [such as Moses and Isaiah], which in their case made them scholars of the Torah.”²⁵[Page 82]

[Page 83]These trained Israelite scribes elaborated or improvised on the scriptural texts before them, a pattern Robert Eisenman noted in his analysis of the Dead Sea Scrolls called pesher: “citing a biblical passage or quotation out of context or even sometimes slightly altered … then proceed[ing] to give an idiosyncratic interpretation having to do with the history or ideology of the group.”²⁶ Indeed, we see a vivid example of this in Nephi’s paraphrase of one of his father Lehi’s prophecies, marked by three rapid Type 2 Amplifying improvisations that together comprise one of the more complex extemporaneous change passages of our study:

Yea, even my father
spake much concerning the Gentiles
and also concerning the house of Israel …
And after that the house of Israel should be scattered,
they should be gathered together again,
or in fine, that after the Gentiles had received the fullness of the gospel,
the natural branches of the olive tree or the remnants of the house of Israel
should be grafted in or come to the knowledge of the true Messiah,
their Lord and their Redeemer.
And after this manner of language did my father prophesy
Earliest Text, 26; 1 Nephi 10:12, 14-15

The passage shows Nephi not quoting, but more freely paraphrasing his father’s prophecy, which then provides the flexibility to elaborate and expand, in pesher form, with repeated improvisations to make his own prophetic point for future readers.

As we come to a close, consider two broader conclusions to be drawn from our study of extemporaneous change and improvisation in the Book of Mormon, one regarding the nature of scripture, and the other regarding the nature of the Book of Mormon itself.

About Scripture

First, consider the significance of Smith choosing to retain CCPs intact across successive editions of the Book of Mormon — mistakes and all — for the world to see. He presented a work that was at once in the realm of the divine, with miraculous visions and revelations from God; yet the text itself clearly was imperfect, not only in grammar (which he tried to perfect over time), but even in the awkward improvisations of its writers (which he evidently did not) — a truly telling paradox. We see evidence [Page 84]of the hand of God, but also the struggles of inexpert writers, editors, and translators trying to effect or implement this work of God.

A common assumption, said James Kugel, is that “there is no mistake in the Bible, and anything that might look like a mistake … must therefore be an illusion to be clarified by proper interpretation.”²⁷ Like many devout Christians believing the Bible to be the infallible and inerrant word of God, many Mormon believers also assume that the Book of Mormon was given by God with inerrancy — divinely and perfectly, without human fault, especially because of its revelatory origins. However, the improvisations we have studied in depth here suggest that the Book of Mormon is a commissioned work of God brought about by ancient authors, orators, and editors with imperfect skills, qualifications, and predispositions.

What does this say about scripture, about what it is, and how it gets written? Roger Terry, who studied the inconsistent usage of archaic pronouns and verbs in the Book of Mormon, said:

What can we learn from the idea that God didn’t prepare a perfect translation himself and miraculously present it to Joseph? This fact seems to support the homely metaphor a friend of mine once coined: “God doesn’t send cookies baked in heaven.” … [We] must assume that [God] left the [construction and] translation largely in the hands of his still imperfect children, mortal or immortal. For a volume as important as the Book of Mormon to come forth with such labor pains and such imperfections suggests perhaps a more hands-off God than some of us prefer to imagine. Subtlety and restraint appear to be two of his most prized attributes.²⁸

Our research here suggests similarly that the mistakes and improvisational fixes we see in scripture are evidence of ordinary actors, acting in the long-ago production of a sacred text, writing inspired words of seminal moments in time, and of visions, revelations, and impressions from God — but all moderated by the limited, or extraordinary, literary resourcefulness of the prophets, priests, teachers, and commissioned servants who wrote them.

About What the Book of Mormon Is

Upon its arrival in nineteenth-century North America, both believers and nonbelievers instantly framed the Book of Mormon as another “Bible.” Richard Bushman wrote, “Martin Harris [early Book of Mormon scribe] referred to the [printer’s] manuscript as the ‘Mormon Bible’ when [Page 85]he was negotiating with the printer. Newspapers derisively called it the ‘Gold Bible.’ Eber D. Howe, the Painesville, Ohio, editor who took an interest in Mormonism, described the recovery of the Book of Mormon as a ‘pretended discovery of a new Bible, in the bowels of the earth.’”²⁹ This early biblical framing of the Book of Mormon set a tall standard for acceptance and marked the beginning of an epoch struggle over what the book was and what it should become in the religious milieu of modern Christianity.

On the surface, the findings presented here seem to undermine the very premise of this new scripture’s claim to divine authenticity. If the Book of Mormon was a new Bible, the very word of God, an oracle from the heavens, then the fixes and patches, the improvisations discovered here, would seemingly undermine its legitimacy. By contrast, the King James Bible is a much more polished product, a “Bible designed to be read aloud, a Bible which came from preachers,”³⁰ said Lord Melvyn Bragg. The Old Testament certainly contains its own improvisations from its early authors, as we saw with Abinadi’s recitation of the Ten Commandments in which God clarifies in detail the meaning of “graven image” (discussed in depth in Article One). Indeed, the Bible itself is a jagged collection of sacred texts constructed over time. William Dever said, “These [Bible] stories were set down over a period of a thousand years, the whole finally woven into a composite, highly complex literary fabric sometime in the Hellenistic era (ca. 2nd century bc) … [containing] diverse and indeed contradictory literary forms.”³¹

In a similar way, woven into the fabric of the Book of Mormon narrative is its own jagged claim to historical scripture, not just by the sacred stories it records, but by the means the early authors and editors use to construct them through layers of authoring, editing, and constructing over a thousand years of history — including, as we’ve seen in this paper, the use and placement of extemporaneous change and improvisation. Thus, in the introduction of First Nephi we see Nephi deliberately use an extemporaneous change to palpably underscore the personal authorship of his own historic writings:

This is according to the account of Nephi,
or in other words, I Nephi wrote this record.
Earliest Text, 5; 1 Nephi 1: Introduction

Terryl Givens said, “Why such emphatic insistence on the literal origins of the record, with Nephi’s own hand? Clearly, unlike the impersonal voice with which Genesis opens the biblical account of creation, focusing as it does on cosmic history, epic events, and God’s [Page 86]primal acts of creation, the Book of Mormon’s first named author urgently presses upon his audience the very human, very local, and very historical nature of his narrative. It is as far removed from mythic beginnings and anonymous narratives as he can possibly make it.”³² Nephi’s words of personal authorship are strong and compelling, but they are made even more so by the insertion at that early point in the text of a small, but significant improvisation: “or in other words, I Nephi wrote this record.”

The extemporaneous changes of the Book of Mormon deepen this historical sensibility by showing the vivid improvisations of its engravers and authors. We see empirically and anecdotally their improvisational signatures as they engrave, write and speak, each signature different from the next — Nephi versus Mormon, Alma versus Abinadi, or Benjamin versus Limhi. Moreover, we see these improvisational impulses not merely in a simple linear story narrative, but in historic layers as if we were digging through literary strata — later translators who deferentially maintain the improvisations of earlier editor-engravers, who strategically embed the improvisational documents and discourses of original prophets, priests, kings, generals, and teachers, who themselves quote the improvisational ancient texts of earlier seminal prophets, like Moses and Isaiah.

The presence of these extemporaneous changes, preserved intact through time and edition, suggests that we are not merely reading history, but witnessing the very authors who constructed it in an earlier time. The improvisations we see in the CCPs are sometimes awkward and clumsy; clearly Smith as translator in the modern day, or Moroni as compiler, finishing-editor and conservator in the fifth century AD, or even Mormon as primary editor-engraver whose work would bear his name, should have cleared these up over time — but they plainly did not. Like the Decalogue, subject to minor changes in translation, word, or grammar but otherwise fixed in time, these improvisations help us understand the Book of Mormon in a different way as authentic sacred scripture. In so doing they help us understand what scripture is, and how it gets constructed to achieve the inspired editorial designs of those who compiled it long ago working under their own respective commissions from God.[Page 87]