[Page 105]Abstract: The works of Tolkien and the Book of Mormon have been compared in a variety of ways by multiple authors and researchers, but none have looked specifically at the unusual names found within both. Wordprint studies are one tool used in author attribution research, but do authors use specific sounds more than others — consciously or subconsciously — when selecting or inventing names? Some research suggests they may and that their patterns could create a “sound print” or phonoprint. This constitutes a fresh and unusual path of research that deserves more attention. The purpose of this exploratory study was to see if phonoprints surfaced when examining Dwarf, Elf, Hobbit, Man, and other names created by Tolkien and Jaredite, Nephite, Mulekite, and Lamanite names found in the Book of Mormon. Results suggest that Tolkien had a phonoprint he was unable to entirely escape when creating character names, even when he claimed he based them on distinct languages. In contrast, in Book of Mormon names, a single author’s phonoprint did not emerge. Names varied by group in the way one would expect authentic names from different cultures to vary. Although much more research needs to be done to establish the validity and reliability of using phonoprints for author identification, this study opens a door for future research.

One of the unexplained mysteries of the world is the Voynich Manuscript,¹ a 240-page, richly illustrated book, carbon dated to the [Page 106]14th or 15th century, written in characters a virtual army of worldwide linguists and code breakers have been unable to translate. It is considered by scholars to be either a real language or a skillfully invented language that has characters grouped with the appearance of forming words.² Its origin is unknown. It belonged at one time to Emperor Rudolph II of Germany (1576–1612), who thought it to be the work of Roger Bacon, an English Franciscan friar and philosopher (1219–1292),³ though this has been overridden by a number of analyses.⁴ The book is now housed in the Beinecke Library at Yale University.

Linguistic Analyses

Although no one has been able to interpret its characters or discern whether there is any meaning to its text,⁵ extensive linguistic analyses of the Voynich Manuscript have been made. A statistical study of the word equivalents claimed that the text adheres to linguistic rules.⁶ Jorge Stolfi, who dealt with such rules, designated core, mantle, and crust characters.⁷ German and Latin influences have been found, and some scholars have suggested that the manuscript might represent collaboration between individuals from Italy and Germany.⁸ After years of analyzing patterns in the linguistics of the manuscript, Marcelo Montemurro, a theoretical physicist from the [Page 107]University of Manchester, UK, and a colleague used computerized statistical methods to find “semantic networks” which appeared to them to be clustered patterns of content-bearing words, with new word structures indicating a shift in topic.⁹ Linguist Prescott Currier has found two different writing styles and text properties, thus being able to conclude that whatever it might say in whatever language, it is co-authored.¹⁰

If linguists can make fascinating analyses of unknown languages and communication systems in the Voynich Manuscript, then the Book of Mormon seems to open itself to similar investigation. Readers who believe the Book of Mormon claim it was translated from an ancient record represented in characters that no one but a prophet with divine intervention could be able to read. Other manuscripts that invite linguistic exploration are the works of J. R. R. Tolkien. He created worlds, cultures, and characters based on languages he invented himself.

Tolkien was a brilliant linguistics scholar who had mastered thirteen languages (ancient and modern) and had a working knowledge¹¹ of nine more; he invented imaginary languages in his spare time, having started this practice as a child. He knew how to create linguistic systems. Fortunately, he didn’t want his works to be unreadable like the Voynich book; the stories are told in English.

At the time the Book of Mormon was published, Joseph Smith had about a third or fourth grade education and no knowledge of languages or linguistics. He never claimed to have created the Book of Mormon — its people, its cultures, or the languages from which it was derived. He explained that he had translated an ancient record into English by the gift and power of God.

Purpose of the Study

One element The Lord of the Rings and the Book of Mormon have in common are hundreds of unusual, striking names. Tolkien created his names while Joseph Smith maintained the names in the Book of Mormon are authentic. Peter A. Huff (not a Latter-day Saint), referred to the Book of Mormon as “an extraordinary piece of literature” comparable to [Page 108]Tolkien’s The Lord of the Rings.¹² Although he is not the first or only person to compare the two works, the purpose of this study was to explore Huff’s analogy by looking closely at the names contained in both volumes.

Authentic personal names are derived from many different sources; few parents choose their children’s names according to linguistic forms and codes. In contrast, fictional names in creative works come primarily from one source, the author, and they are usually carefully and purposefully chosen or invented. Although Joseph Smith maintained he translated the Book of Mormon from an ancient record, critics claim he wrote it as any author writes a fictional work. If these critics are correct, he would have presumably chosen or created the names as fiction writers do.

Background

To examine and compare names from these two sources, we need to consider differences between authentic and fictional names in general as well as some background concerning names of Tolkien’s characters and names in the Book of Mormon.

Authentic Names

Personal names have cultural significance. Some societies use personal names to teach culturally important lessons or remind their owners of significant events. In other societies names can reflect social or financial position or incidents surrounding a child’s birth.¹³ Additional reasons for choosing names, at least historically, include naming a child with hoped-for characteristics,¹⁴ focusing on the location of the birth,¹⁵ or alluding to the occupations of parents.¹⁶

Today, with the interaction of many different cultures and peoples, personal names may derive from multiple languages and origins.¹⁷ Many [Page 109]people have a surname from the culture in which they reside and a given name chosen from the native language of the parents.¹⁸ In Western societies given names can derive from circumstances such as invasions of or trading with other cultures,¹⁹ conversions to Christianity,²⁰ pagan mythology,²¹ or outright coinage.²²

Regardless of the origins of names, their phonological features evolve over time. Names are shortened, combined, or spelled with variant letters until the same names can differ significantly.²³ The phonology of names may have personal as well as social significance. Native English speakers prefer, for example, female names that follow specific features (like ending in /i/ as in Debbi or /n/ as in Sharon) and male names that follow others (some ending in /r/ as in Brenner or in obstruents as in Brad); names will change or fall out of favor if they do not follow these features (such as Ashley and Whitney changing from boys’ to girls’ names). Other cultures prefer to keep the exact or very similar phonology to the pronunciation of the name in the borrowed language, even if it violates native language phonotactics.²⁴

In previous research, the authors studied the degree to which personal names differ in their phonology by examining a corpus of the 100 most prevalent male personal names in English in the 19th century, [Page 110]using the phonotactic calculator²⁵ created by Vitevitch and Luce.²⁶ This calculator determines the relative probability of each phoneme in a word occurring in the location that it does and also the probability of its occurring adjacent to the preceding and following sounds in the word (bifones or bi-phonemes as defined by Vitevitch and Luce). In addition, a word’s probability is determined by examining its neighborhood density. Words with many phonological neighbors (cap, cat, can, cash) score higher on the probability calculator than words with few neighbors (oriole, flask, etc.). The calculator determines neighborhood density and sound location probability by comparing a selected word to calculations already performed on a corpus of English created by Kucera and Francis.²⁷ For example, in the word box (/baks/), the calculator would determine how probable it is that a word would start with the sound /b/ and be followed by the sound /a/ in the second position, how likely a word would have /a/ in the second position and have the sound /k/ following it, and so on. We found that phonotactic probabilities of the names in the census varied greatly.

In addition to varying at the sound level, personal names also differ when they are examined as whole units. When people buy plane tickets, their names are examined by name recognition software that searches for matches with names on no-fly lists but also identifies the background of the names with amazing accuracy by comparing them to databases containing millions of names.²⁸

Thus personal names come from a variety of sources, vary in their phonological features and properties, and vary when considered as whole units in comparisons such as databases. In considering the names created by Tolkien and the names in the Book of Mormon, we asked if a writer of fiction creating all the names in a novel could replicate this [Page 111]variety: if one author could imitate the diversity found in a corpus of personal names from actual languages — past or present.

Names in Fiction

Some authors choose to name their characters using authentic names they have encountered in experience; found on internet lists; or noticed consulting newspapers, telephone directories, road signs, or tombstones.²⁹ Others invent unique names — especially when they create fantasy or science fiction. Sometimes these supposedly come from languages unknown to us or from worlds beyond our own. Drawn from authentic sources or invented, names used by authors of fiction are chosen to match the personality of the character or to bring up stereotypes or archetypes.³⁰ When interviewed about their methods of choosing names for their fictional characters, successful authors spoke of choosing names that held personal significance for them, doing research to find names that were unusual or represented a particular time period or culture, and choosing names with phonology that appealed to them or would possibly appeal to potential readers.³¹

Scholars who have studied the Voynich Manuscript over the years cannot identify the language (or code) in which it is written and can hazard guesses at content only from the illustrations. But as noted previously, they have been able to identify aspects of the authors’ time period and nationality from the way the characters are grouped as words and what appear to be paragraphs — although they have no idea the sounds the characters and groupings might have made. They have even reconstructed linguistic rules and grammar from statistical analysis. One analyst claims to have figured out the meanings of 10 of the document’s words using linguistics, grammar, and repetition patterns.³²

Similarly, research has demonstrated that authors writing in English or other known languages have individual biases toward using specific words and syntactic structures when they write. Their choices constitute [Page 112]a “wordprint” (similar to but not as precise as a fingerprint) by which they can be identified.³³

Some criticize this technique because collecting sufficient data is usually difficult.³⁴ Nevertheless, this method of examining an author’s use of specific syntactic structures, type-token ratios, and other lexical features has been used regularly to identify or verify authorship of documents.³⁵ In fact, some scholars claim authors can be identified based on their use of function words alone (e.g., then, why, the, if, of, but, have, etc.).³⁶ Others have successfully verified or ruled out authorship on writing elements as simple as cross-textual comparisons of style markers such as variant forms of punctuation and spelling.³⁷ Scholars of the Voynich Manuscript were able to eliminate Roger Bacon as a possible author because no trace of an English wordprint emerged in their analysis.³⁸

An aspect of wordprints that has not been examined sufficiently in author attribution research is whether authors use some specific sounds (phonemes) more than others — consciously or subconsciously — when selecting or inventing names. Some research suggests they might do this and that their patterns may create a “sound print” or phonoprint.³⁹ This constitutes a fresh and unusual path of research that may merit more examination. Traditionally, words have been seen as the smallest building blocks over which authors have some freedom to choose. This new line [Page 113]of research expands the fundamental unit of text into phonemes and proposes the possibility that we could produce a phonoprint that would differ from author to author. Despite that authors have fewer sounds with which to create words than they have words with which to create prose and poetry, there is some evidence that authors favor certain sounds over others when choosing or inventing names.⁴⁰ We recognize that much more research will need to be completed to establish a baseline with which valid and reliable comparisons can be made. Many works of fiction by a variety of authors will need to be examined. Nevertheless, this exploratory study was completed to see if further research might be justified.

Tolkien’s Character Names

As Tolkien, a professor of Anglo-Saxon at Oxford, created languages based on some characteristics of natural languages, he deliberately used different sound systems for each of them, and he particularly enjoyed the names that evolved from these systems. He commented during a BBC radio interview that a “good name” gave him “great pleasure.” Although Tolkien claimed his names came from invented or ancient languages, sometimes he selected names already existing in Icelandic poems like Voluspa and Gylfaginning. For example, of the 13 dwarves’ names in Thorin’s company in The Hobbit, 12 come right out of Voluspa. Tolkien simply used the traditional Anglicization of the names. Other names are recognizable surnames in Great Britain such as Meriadoc and Faldor.

Tolkien once said that he sometimes started with a name: “Give me a name and it produces a story, not the other way about normally.”⁴¹ Sometimes he claimed his character names followed the patterns of invented languages, including phonotactics, and of the cultures he created for them. Other times he used names found in mythic poetry or his own experience.

Whatever the source of Tolkien’s names, this study analyzed character names from five different languages groups selected or created by Tolkien (elf, man, dwarf, hobbit, other). As few surnames or titles appear in the text, only first names were included. Names given to two or more characters were used only once. Of the 197 names found in [Page 114]Tolkien’s writings, 14 were discarded because they were names of beings for whom there were only a few representatives (dogs, ravens, goblins, etc.). Therefore, 183 names were used in the final analysis.⁴²

Book of Mormon Names

The Book of Mormon includes 337 proper names and 21 gentilics (analogous forms), with 188 of them found in no other source. Joseph Smith purported that the large number of Book of Mormon names represent a diversity of cultures in ancient America (e.g., Jaredites, Nephites, Mulekites, Lamanites) and thus the names were derived from the language backgrounds of these cultures, which included Egyptian, Hebrew, and other Semitic languages. (Joseph Smith had no knowledge of or experience with any of them at the time the Book of Mormon was produced, although he studied Hebrew later in his life.) Of the 337 proper names in the book of Mormon, 149 are found in the Bible as well (e.g., Samuel, Isaiah, Gideon, Benjamin, Aaron, Noah, Shem, Timothy, and Jacob) and thus were excluded from this study.⁴³ When place names were also excluded, 162 unique names remain as those of people. Of these, 32 could not be clearly classified by culture, so they were not used in this study. This elimination left 130 single names without ranks or titles: “Typical of the ancient Semitic languages from which the Nephite record is [said to have been] derived, the Book of Mormon does not use surnames or attach modern titles to its names such as … Professor, Reverend, Count or Earl.”⁴⁴

Of the unique names in the Book of Mormon, 41 are mentioned only once. In contrast, the most prominent names are given often, some of them assigned to multiple characters — particularly descendants of the first recorded character with that name (e.g., Nephi, Helaman, Laman, Moroni, and Zoram). Nephi is mentioned 77 times, the unique name [Page 115]represented most frequently. Ammon appears in 51 references, and Moroni is mentioned 44 times.⁴⁵ Name pronunciations and spellings in this study were consistent with Skousen’s work with the original manuscript.⁴⁶ Table 1 presents characteristics of names in the four Book of Mormon groups and the five Tolkien groups including the longest and shortest names in each group.

Table 1. Characteristics of the Names:

the Four Book of Mormon Groups and the Five Tolkien Groups.

[Page 116]Methods and Results

Given the exploratory nature of this study, we performed only two analyses to compare these two name sources in both consistency and variety of names. We included phonotactic probabilities and identification by language recognition software.

Phonotactic Probabilities

Our first analysis examined 313 names, 183 from Tolkien and 130 from the Book of Mormon, to determine how much the five groups of Tolkien names differed from one another in their phonotactic probabilities in comparison with the four groups of Book of Mormon names. The phonotactic probability calculator developed by Vitevitch and Luce, which is available online, provides probabilities by comparing the phonemes in each ordinal position of a given word to the standard American English frequencies.

Klattese, a computer-readable transcription method developed by Dennis Klatt, was used to enter each name phonemically into the calculator.

We examined individual phonemes because some vowel sounds in English, /ĭ/ and /ī/ for example, are more common than diphthongs such as /ai/ or /oi/. Similarly, the consonant sounds /l/, /t/, /k/, and /n/ are more common than /j/ or /w/. We also looked at sequences of phonemes (bifones). The calculator gives probabilities of co-occurrence of phonemes: how frequently two given phonemes occur next to one another. For example, /bi/ occurs less frequently in standard American English than /br/.

In addition to phonemes and bifones, we examined overall word probabilities, which are also provided by the calculator. We emphasize that the phonotactic probability of an entire name doesn’t reflect how often the name is used but how English-like it is. For example, regardless of how often the names Bilbo and Frodo appear in Tolkien’s works, the average phonotactic probability of Frodo is lower (.0433) than that of Bilbo (.0518). We also emphasize that the purpose of the study was not to determine how English-like these names were. Rather, the phonotactic probabilities were used as a way to make quantitative comparisons between the names and examine the differences.

A nested two-way MANOVA was used to examine the overall difference between the two name sources, which was not significant, and also the differences among the name groups within each source, which were significant almost entirely due to differences in mean word length. [Page 117]Dwarf names are shorter than other Tolkien names, and Jaredite names are shorter than other Book of Mormon names.

There were notable differences among the groups in their variety of name lengths, phoneme probabilities, and bifone probabilities. The Book of Mormon name groups have much greater in-group divergence than Tolkien names, which are more homogeneous. In this study we examined the nine name groups and differences are apparent between Tolkien names and Book of Mormon names — especially Jaradite names — which would be expected when comparing any two distinct sources.

The within-source comparisons, the language groups within each book, reveal additional differences between the texts. Although Tolkien claimed his character names were primarily based on different languages — real or invented — the phonotactic probabilities did not differ significantly. In the within-source comparisons among the five Tolkien languages, only one of the 10 language group comparisons (10%) is statistically significant (man versus elf). Among the four Book of Mormon name groups, four of the six variance ratios (67%) were significant (Nephite versus Jaredite; Mulekite versus Jaredite, Lamanite versus Jaredite, and Lamanite versus Nephite).

Results, shown in Table 2, present the averages of overall name probabilities of phonemes and bifones for the Book of Mormon and for Tolkien as well as the average word lengths for each.

Table 2. Average Phoneme and Bifone Probabilities and
Word Lengths of the Names in the Four Book of Mormon Groups
Compared to the Names in the Five Tolkien Groups.

The between-source comparisons and the within-source comparisons for bifones showed a similar pattern to the phonemes, but the contrast between the Tolkien name groups and the Book of Mormon name groups was not as strong.

Language Recognition Software

The second analysis utilized language identification software. The first analyses focused on word parts — sounds and sound combinations. In this examination we looked at words as whole units, using IBM name recognition software to identify which languages seemed to be indicated [Page 119]by each of the names. The classifier algorithm analyzes the spelling patterns in the names but also checks to see how closely they match IBM’s archive of almost 800 million names. The software identification is based on modern languages, which are not relevant for this study since many of Tolkien’s names are supposedly based on ancient languages and Book of Mormon believers claim that its names are also based on ancient languages. However, the software provides a “generic” response when a specific name shows no match with any known language.

Figure 1 gives a combined tabular and bar graph comparison of the number of generic and non-generic names for each name group, with their relative percentages. In Tolkien’s works the highest percentage of generic names was 39.1% (Dwarfs), and the lowest percentage of generic names was 27.8% (Hobbits), a difference of 11.3%. In the Book of Mormon, the highest percentage of generic names was 50.0% (Mulekites), and the lowest was 20.0% (Jaredites), a difference of 30%. The chi square value for the Book of Mormon names was 5.189 with 3 degrees of freedom, and the chi square value for the Tolkien names was 0.850 with 4 degrees of freedom. Thus the Book of Mormon name groups were significantly more diverse than Tolkien’s.

Discussion

Results suggest that Tolkien was unable to entirely escape his phonoprint when selecting or creating character names, even though he claimed he based them on or found them within distinct languages. The two analyses showed little differentiation involving his five major naming groups. A possible phonoprint of sorts seems to have surfaced in his names regardless of the language groups in which he placed them. This is consistent with the results of earlier research on Tolkien’s possible phonoprint⁴⁷ and similar to the results from wordprint studies involving other authors of diverse characters such as Mark Twain. Even when Twain intentionally tried to create words (names) to represent different worlds, he was unable to change his own free-flow noncontextual word patterns successfully enough to simulate wordprints representing other peoples or cultures.⁴⁸

[Page 120]Figure 1. Bargraphs with Accompanying Frequency Tables Comparing the Percent of Generic Names for the Five Groups of Tolkien Names and the Four Groups of Book of Mormon Names.

In contrast, in Book of Mormon names, a single author’s phonoprint does not emerge. Lamanite, Nephite, Mulekite, and Jaredite names have varied by group in the way one would expect names from different cultures to vary when looking at sounds within names and names as whole units. If the Book of Mormon names were created by an individual, they were created by a very different process or based on languages more different from each other and consistent within themselves than those created by Tolkien.

This article is not the first time aspects of Tolkien’s work have been compared to the Book of Mormon. Noel B. Reynolds considered the geography of the two sources and asked if the Book of Mormon describes “real places” or “fantasy geographies and civilizations such as Tolkien’s Middle Earth.” He then wrote of several places in the Book of Mormon [Page 121]that can be positively identified in the middle east, explaining how they are described with an accuracy and detail that would have been beyond Joseph Smith’s knowledge base.⁴⁹ Similarly, Randal A. Wright has asked, “If Tolkien could write Lord of the Rings, then why couldn’t Joseph Smith write the Book of Mormon?”⁵⁰ He then answered his own question by detailing the many years that Tolkien spent writing his book compared to the relatively short time in which Joseph Smith produced the Book of Mormon. He also compared Tolkien’s maturity and advanced education to Smith’s youth and lack of formal education. Although the comparison between Tolkien’s works and the Book of Mormon is not original, this study is the first in-depth comparison of the names found in these two sources.

Hugh Nibley demonstrated differences between Jaredite names in the Book of Mormon and other names that seem to have “Hebrew and Egyptian roots.” He wrote, “the Jaredites and Nephites spoke entirely different languages, and even a cursory search will show that Jaredite proper names have a peculiar ring of their own.”⁵¹ Although this study did not examine possible connections with ancient languages, it did reveal that Jaredite names are not the only group with a “ring of their own”; names from other Book of Mormon cultures have this as well. If Joseph Smith authored the book, he created phonetically independent name groups — something which Tolkien apparently was unable to do.

As linguistically talented as Tolkien was, he was not able to use different sounds and whole name units consistently according to the different invented languages from which they were derived (i.e., Quenya, Sindarin, Westron, etc.)⁵² — even though he was the creator of these languages. Consciously or subconsciously, it appears Tolkien could not escape his own natural leanings toward some phonemes over others as he invented or selected names. If Joseph Smith authored the [Page 122]Book of Mormon, he invented an incredible number of unique names without leaving evidence of a possible phonoprint — a flexibility that Tolkien, despite his incredible repertoire of sounds and symbols, was not able to accomplish.

Authentic names emerge from a variety of geographic, cultural, and ethnic origins, particularly within nations. Thus they show greater variance than fictional names created or selected by a single author. Although much more research needs to be done to establish the validity and reliability of using phonoprints for author identification, this exploratory study may open a door to a fresh line of research that could merit further investigation.

Conclusion: Phonics, Structure, and Meaning

Hugh Nibley wrote, “The evidence that will prove or disprove the Book of Mormon does not exist.”⁵³ Some questions cannot be answered accurately or reliably by tests and analyses. An international variety of linguists, cryptographers, botanists, physicists, and historians have been trying for 600 years to reason out explanations behind the Voynich Manuscript, sometimes referred to as the most mysterious book in the world.⁵⁴ Theories of its origin and authorship range from one of the lost tribes of Israel to the Aztecs⁵⁵, to a group of medieval scholars in an area near Europe — perhaps in Iran or Turkey.⁵⁶

Stephen Bax, a British professor of applied linguistics, used the methods of historic decoders of Egyptian hieroglyphics, who began with proper names: They searched for the known names of pharaohs to put symbols against sounds. Since Bax did not know the culture or geography affecting the manuscript, he began with words that by their relationship to the drawings seemed to name them — including seven plants and a constellation that he could identify. From names he moved to logically connected linguistic and semantic principles. He was eventually able to work out 10 words of text and sounds for [Page 123]13 symbols.⁵⁷ Worldwide headlines hailed Bax as having decoded the manuscript — but many of us would not accept 10 words and 13 sounds as manuscript translation. Fortunately, our understanding of the Book of Mormon is based on a complete translation made possible by the gift and power of God.

Elaborate technological-statistical analyses have discerned linguistic, semantic, and cultural patterns as well as some author factors in the Voynich Manuscript, but they haven’t come together into a complete explanation that will reveal either the language or meaning of this book. Similarly, analysis of linguistic, geographic, and cultural factors in the Book of Mormon, including phonemic patterns in its unique names, cannot prove that it is or is not a translation of an ancient record. Nevertheless, results of such investigations are interesting and can be meaningful to those who value the book.

Today’s technological tools enable us to answer many questions and solve many mysteries. But the truthfulness of the Book of Mormon must be conveyed with experiences of faith. For some, the Book of Mormon rivals the Voynich Manuscript as the most mysterious book in the world, but the Voynich Manuscript is unreadable. The Book of Mormon is not. Its meaning has been accessible for many years and has been life- changing for millions of Latter-day Saints throughout the world.