Janus Parallelism:
Speculation on a Possible Poetic Wordplay in the Book of Mormon

  • Article Formats:
  • PDF
  • MOBI
  • ePub
  • Kindle store
  • NOOK store
  • Order Print Copy

[Page 61]Abstract: In this article, Paul Hoskisson discusses the question of whether Janus parallelism, a sophisticated literary form found in the Hebrew Bible and elsewhere in manuscripts of the ancient Near East, might also be detected in the Book of Mormon. Because the Book of Mormon exists only in translation, answering this question is not a simple matter. Hoskisson makes the case that 1 Nephi 18:16 may provide the first plausible example of Janus parallelism in the Book of Mormon.


[Editor’s Note: Part of our book chapter reprint series, this article is reprinted here as a service to the LDS community. Original pagination and page numbers have necessarily changed, otherwise the reprint has the same content as the original.

See Paul Hoskisson, “Janus Parallelism: Speculation on a Possible Poetic Wordplay in the Book of Mormon,” in “To Seek the Law of the Lord”: Essays in Honor of John W. Welch, ed. Paul Y. Hoskisson and Daniel C. Peterson (Orem, UT: The Interpreter Foundation, 2017), 151–60. Further information at https://interpreterfoundation.org/books/to-seek-the-law-of-the-lord-essays-in-honor-of-john-w-welch-2/.]

John W. Welch’s discovery in 1967 of chiastic structures in the Book of Mormon was to that date the most important literary discovery regarding latter-day scripture since the beginning of the Restoration, and it continues to be a singular event in the scholarship of the text of the Book of Mormon. He not only opened up hitherto unplowed and [Page 62]fertile research fields, but he also pointed the way forward to previously untouched approaches to the text. All faithful scholars have benefitted from this discovery. It is therefore a daunting task to write a requisite article to thank and honor a great scholar and a dear friend.

While serving a mission for The Church of Jesus Christ of Latter-day Saints in Germany, John, or Jack as his friends call him, began his discovery serendipitously by attending a class in the Regensburg Priester Seminar on the New Testament. This led him to a book that explained the chiastic structures in and of the Book of Matthew. He became intrigued with the concept of chiasmus, for, “if it is evidence of Hebrew style in the Bible, it must be evidence of Hebrew style in the Book of Mormon.” He soon enough found chiasms in the Book of Mormon. After returning home he wrote his master’s thesis on chiasmus and has never looked back.1

Janus Parallelism in the Hebrew Bible

In 1978, eleven years after Jack’s discovery, Cyrus H. Gordon published an article defining and outlining the literary form he termed Janus parallelism, and about which he had lectured previously.2 He chose the name because the Roman god Janus had two faces, one looking forward and one looking back. Thus, this label describes “a literary device in which a middle stich of poetry parallels in a polysemous manner both the line that precedes it and the line which follows it.”3 Gordon came upon the concept one day when reading Song of Solomon 2:12, which reads in my own translation,4

The blossoms are seen on the earth;
the time of pruning/singing has come,
and the voice of the turtle dove is heard in our land.

[Page 63]The key to the polysemous parallel is the Hebrew word zāmîr in the middle stich, and whose translation I have rendered with two of its meanings, “pruning” and “singing.” The meaning “pruning” looks back to the time of blossoms in the first stich, and the meaning “singing” looks forward to the sounds of the voice (of the dove) in the third stich.

Gordon discovered other scattered examples of Janus parallelism in the Hebrew Bible, none of which, like Song of Solomon 2:12, are recognizable in English translations (see below). Others followed his lead. Gary Rendsburg, a student of Gordon, found examples in the book of Genesis.5 Another scholar, Scott Noegel, published a fairly exhaustive study with numerous examples from the book of Job.6 In fact, Janus parallels are attested in a wide array of Near Eastern languages, including “Akkadian, Arabic, Hebrew, Hieroglyphic Egyptian, and Ugaritic.”7

One of the reasons that Janus parallels in the Hebrew Bible went relatively unnoticed for so long is that the context tends to mask the polysemy. Unlike chiasmus, which depends on the mirrored inversion of a poetic structure that creates a balanced structural configuration of semantic parallels, Janus parallels depend on polysemy. The polysemous nature of the key word allows it to parallel at least two other words that by themselves are not parallel. For example, in the case of Song of Solomon 2:12, the word blossoms in the first stich is not a semantic parallel, antithetical or synonymous, with voice in the third stich. Adding to the difficulty is that zāmîr with the meaning “singing” occurs much more often in the Hebrew Bible than with the meaning “pruning.” In fact, while neither is common, the meaning “singing” occurs six times,8 and the meaning “pruning” occurs only three times, including in the [Page 64]verse in question.9 Only because Gordon knew Hebrew, Arabic, Ugaritic, Akkadian and Egyptian lexicography so well was he sensitive enough to recognize the polysemic nature of homophonic and homographic zāmîr.

Most examples, however, of Janus parallelism in the Hebrew Bible are even more difficult to recognize than homographic and homophonic zāmîr. The monographic, polyphonic, and polysemic example in Genesis 15:1 of the unvowelled Hebrew text illustrates this difficulty.10 In the King James translation, this tristich reads,

Fear not, Abram:
I am thy shield,
and thy exceeding great reward

The consonant-only text of the middle stich reads, ʾnky mgn lk. With no vowel markings, several different standard vowel patterns can be applied to the consonantal structure, resulting in different ways to read the middle stich. When the Masoretes supplied the vowels to the previously consonant-only text, of the theoretically possible patterns they could supply only one, namely, ʾānōkī māgēn lāk, “I am thy shield,” for the middle stich. With this vowel pattern, the polyphonic and polysemic nature of the consonant-only text is obscured. That is, with the vowels it can only be pronounced in one way and mean, in its literal sense, only one thing, so that the reader is less likely to think of other options for vocalization and meaning. Furthermore, the meaning “shield” for māgēn is well known from the Hebrew Bible (and even in modern Hebrew) and provides a strong parallel to “fear not” in the first stich.11 Therefore, most readers of Hebrew would rarely consider other less well-known meanings that are possible through different vowelling.

If, however, the vowels of the Masoretic text are ignored, then other readings and meanings can emerge, thus creating the possibility of polysemy. By reading the text silently, the reader can ponder which of the acceptable and normative vowel patterns fits the consonant-only text. And if the reader is astute, the polysemous nature of the middle [Page 65]stich can be discovered. For instance, if mgn is vocalized as the Hebrew noun megen, which means “gift,” the parallel with “reward” in the third stich is strong.12 Notice, though, that because reciting the consonant-only text out loud requires inserting vowels, reading out loud eliminates all but one of the meanings of mgn.13

To summarize, Janus parallels are constructed in two slightly different ways. First, as in the example in Song of Solomon 2:12, a homographic and homophonic word has different meanings, allowing for a word play. Second, the consonant-only text can be read polyphonically and therefore polysemously, as in Genesis 15:1. It should also go without saying that the polysemous nature would be lost in any translation because the translator, unless extremely gifted at finding polysemous words in the target language, would have to settle on only one of the possible meanings of the source text word.

To give the reader an idea of how Janus parallelism might work in English, I have constructed the following example. Note that it only works because I have mixed modern and older English spelling conventions with imaginative forms, which forces the reader to ponder which meaning to assign to the unusual spelling tyme.

Of all mortalle spyces
that temporally entyce us
it is tyme
that orders every season
with eternal, divine reason.

Admittedly, my attempted English Janus parallel is rather clumsy. The task of constructing valid examples is extremely difficult because English requires vowels, and both vowels and consonants are usually phonemic in English. Most homographic polysemous English words would not force the reader to consider any meaning other than the most common one.

[Page 66]Having provided a basic, if technical, explanation of Janus parallelism in the Hebrew Bible, I now turn to the Book of Mormon.

A Possible Janus Parallelism in the Book of Mormon

Using the same basic logic with which Jack began his study of chiasmus in 1967, I suggest that if Janus parallelism is a feature of Hebrew writing, which it undoubtedly is despite being relatively unknown in the wider academic community, then it is possible that it exists in the Book of Mormon.

The main stumbling block to recognizing polysemous parallels in the Book of Mormon is that the book is a translation, and no translation of the Janus parallels in the Hebrew Bible (or other languages where they exist) survive when translated into English or almost any other modern language.14 If translators become aware of polysemy in the source language, they would still be faced with the nearly impossible task of finding in the target language an equally polysemous word that replicates the same intentional meanings of the word in the source language. In the original example above from Song of Solomon 2:12, what single English word could mean both “sing” and “prune”? Therefore, it would appear that searching for English lexical polysemy in the Book of Mormon would be arduous at best and impossible at worst.

A better approach would be to look for tristich forms. That is, in its simplest form, any Janus parallel translated into English would probably appear as a tristich. But since not every tristich in the Book of Mormon (or in the Hebrew Bible for that matter) contains polysemy, the acid test would be to back translate the English into Hebrew.15 Admittedly, any back translation introduces a level of subjectivism that proscribes any claim to certainty. Assuming, however, that the back translation is reasonably valid, an absence of any polysemy in the back translation would greatly diminish the plausibility of a Janus parallel. All attempts, [Page 67]therefore, to demonstrate polysemous parallels in the Book of Mormon should elicit high levels of skepticism.

Nevertheless, I believe that a plausible case can be made that a Janus parallel exists in 1 Nephi 18:16. The relevant words of the tristich read,

I did look unto my God,
and I did praise him all the day long;
and I did not murmur against the Lord because of mine afflictions.

This verse,16 with its three part parallelism, is manifestly poetic, even in English. And the artful and progressive lengthening of each stich reinforces the poetic nature of the verse. Additionally, the fact that this verse seems only loosely attached to the previous verse and relatively unattached to the following material would appear to call attention to its stand-alone nature as a tristich.

The polysemy would center on the verb in the second stich, praise. Immediately it can be seen that praise Him in the second stich is a strong antithetical parallel to murmur against the Lord in the third stich. But praise Him is not entirely parallel synonymously with look to God in the first stich. That is, in English, there is at best a weak semantic parallel between the first and second stich. If this verse is a candidate for Janus parallelism, a back translation into Hebrew would have to provide a better semantic parallel between the first and second stich while polysemously preserving the strong parallel between the second and third stich.

The most common verb in Hebrew for praise is hallēl.17 This is the Hebrew word behind the English expression hallelujah, meaning, “Praise ye the Lord” (Ps. 147:1). It is also the verb used in Hebrew in “I will praise him among the multitude” (Ps. 109:30). The consonantal structure underlying hallēl in Hebrew is polysemous. In addition to “to praise” it can also mean “to shine,” as in Job 31:26, “If I beheld the sun when it shined;” and “to give light” in Isaiah 13:10, “For the stars of heaven and the constellations thereof shall not give their light.” An extended meaning of hallēl could be “to cause to shine.”18 Therefore, in 1 Nephi 18:16 it could mean something similar to “I did cause him [my [Page 68]God] to shine all the day long,” in the sense that the actions of the author made his God to shine forth in glory. But in my mind, these semantic gymnastics using hallēl do not provide a stronger parallel with the first stich than the monosemous praise.

A better candidate in this verse for a back translation of “praise” into Hebrew would be the piel verb zammēr, the same root translated as “song” in Song of Solomon 2:12 that started the discovery of Janus parallelism. Besides “to sing” and “to prune” as explained above, this verb in Hebrew also means “to praise.” For example in Psalm 57:7 (verse 8 in Hebrew), zammēr is poetically paired with šîr, “to sing”: “I will sing [ʾāšîrâ] and give praise [ʾăzammērâ].” In Psalm 147:1 zammēr is paralleled with hallēl, “to praise”: “Praise ye the Lord: for it is good to sing praises unto our God.” But literally, the verse reads, “Praise Yah! For it is good to sing19 our God,” making halělû parallel to zamměrâ. The difference between zammēr and hallēl seems to lie in the nature of the intended action; zammēr is sung or chanted out loud musically and therefore could be a more public expression of praise, while hallēl could be spoken aloud but could also be spoken in the mind, privately, silently.

With these two not totally unrelated meanings of zammēr, it is possible to construct in plain English an admittedly speculative case that 1 Nephi 18:16 could contain a Janus parallelism. I therefore suggest that the first and second stichs could be read as

I did look unto my God,
and I did sing unto him all the day long;

and that the second and third stichs be read as they now appear in the 2013 edition,

and I did praise him all the day long
and I did not murmur against the Lord because of mine afflictions.

The polysemous parallel succeeds because the Hebrew verb zammēr can mean both “praise” and “sing.”

[Page 69]Conclusion

In our English Book of Mormon 1 Nephi 18:16 reads nicely as it stands. The tristich works well enough that most readers probably never stop to contemplate the verse or examine its poetic structure. The first and second stichs exhibit a weak but recognizable parallel. The second and third stichs form a demonstrably strong antithetical parallel. And if it were not for the possibility of a Janus parallel hiding behind the English text, no further discussion would be necessary.

But knowing that Janus parallels are an occasional feature of Hebrew, and other ancient poetry, is enough of an enticement to brave a speculative back translation of the verse into Hebrew to see if a stronger synonymous parallel might exist in a posited Hebrew urtext. The back translation that I offer seems to suggest a stronger synonymous parallel between the first and the second stich while maintaining the strong antithetical parallel with the third stich.20

Might this verse be an example of a Janus parallel? I can only claim that it is plausible. We will never know for sure until the urtext becomes available. In the meantime, perhaps other verses might prove to be potential candidates. But no matter how many plausible instances might be observed, no one of them alone will be totally convincing without the witness of the urtext.

1. For Jack’s own account of his discovery see John W. Welch, “The Discovery of Chiasmus in the Book of Mormon: Forty Years Later,” Journal of Book of Mormon Studies 16, no. 2 (2007): 74-87, 99.
2. Cyrus H. Gordon, “New Directions,” Bulletin of the American Society of Papyrologists 15, no. 1/2 (1978): 59-66. I remember hearing Gordon talk of Janus parallelism during the academic year 1972-73 in classes I took from him at Brandeis University.
3. Scott B. Noegel, Janus Parallelism in the Book of Job, (Sheffield, England: Sheffield Academic Press, 1996), 12, citing Gordon, 59-66.
4. Unless otherwise noted, all Old Testament quotes are from the King James version. This verse reads in the King James translation, “The flowers appear on the earth; the time of the singing of birds is come, and the voice of the turtle is heard in our land.”
5. Gary Rendsburg, “Janus Parallelism in Gen. 49.26,” Journal of Biblical Literature 99 (1980): 291-93; and “Notes on Genesis XV,” Vetus Testamentum 42.2 (1992), 266-72.
6. Noegel, Janus Parallelism. For examples in literatures other than Hebrew, see Zackary M. Wainer, “Janus Parallelism in Šulgi V,” Bible Lands e-review, 2013, no. S2, and, as cited in Wainer, Wayne Horowitz and Shalom Paul, “Two Proposed Janus Parallelisms in Akkadian Literature,” NABU, 1995, no. 1 (Mars): 11-12; Noegel, “Yasîm-El’s Sophisticated Rhetoric: A Janus Cluster in ARMT XXVI, 419, 1. 10’, NABU 1995, no. 4 (Décembre): 81-82; and Shalom Paul, “Polysensuous Polyvalency in Poetic Parallelism,” in Sha’arei Talmon: Studies in the Bible, Qumran, and the Ancient Near East Presented to Shemaryahu Talmon, ed. M. Fishbane and E. Tov (Winona Lake, IN: Eisenbrauns, 1992), 147-163.
7. Wainer, “Šulgi V,” in the second paragraph.
8. 2 Sam. 23:1; Isa. 24:16; 25:5; Ps. 95:2; 119:54; and Job 35:10.
9. As a verb, zmr is attested many times with the meaning “to sing or praise,” but with the meaning “to prune or trim” it occurs in only two passages, Isa. 5:6 and Lev. 25:3-4. As zmr it is attested in line 6 of the Gezer tablet.
10. For a discussion of the Janus parallelism in this verse see Noegel, Janus Parallelism, 12, which is based on Rendsburg, “Gen. 49.26” What follows is my own analysis of this verse.
11. The meaing “shield” in this verse may hark back to Gen. 14:20 where the piel perfect verb miggēn occurs in the phrase “delivered thine enemies into thy hand” (KJV). I thank my friend and colleague Dana Pike for this suggestion.
12. Another way to read mgn is as the piel infinitive verb maggēn, meaning “to gift.” But an infinitive in this middle stich does not make a lot of sense.
13. The polysemous nature of the consonant-only texts suggests, if not demands, that these texts did not have their origin in oral recitation, but rather were composed in written form and were designed to serve a literate clientele. As interesting as this topic is, it is, unfortunately, beyond the scope of this paper to explore the origin of the written text of the Hebrew Bible. For further study into this issue, consult the works of the Israeli scholars Aaron Demsky and Shalom Paul. See also Brian B. Schmidt, Contextualizing Israel’s Sacred Writings: Ancient Literacy, Orality, and Literary Production (Atlanta: Society of Biblical Literature, 2015).
14. An exception might be languages that use the Arabic script without vowels.
15. This is not the place to quibble over whether the language on the plates that Joseph transmitted to us was Hebrew or Egyptian. If it was not Hebrew, then it was Egyptian, and the Egyptian hieroglyphic script is capable of producing polysemous words, a prerequisite for constructing a Janus parallel. (See the discussion above with notes 6 and 7.) Unfortunately, not knowing any Egyptian, I am unaware of any specific examples of Janus parallels in Egyptian. For a general discussion of related literary forms in Egyptian, see Antonio Loprieno, “Puns and Word Play in Ancient Egyptian,” in Puns and Pundits: Word Play in the Hebrew Bible and Ancient Near Eastern Literature, ed. Scott B. Noegel (Bethesda, MD: CDL Press, 2000), 3-20.
16. When I say “verse” here I am not referring to the versification introduced into the Book of Mormon with the 1879 edition, but rather to the seeming independence of the form of this tristich from the surrounding material.
17. There is no qal form of this verb, just the piel, pual and hithpael forms.
18. Koehler, L., Baumgartner, W., Richardson, M. E. J., & Stamm, The Hebrew and Aramaic lexicon of the Old Testament (Leiden: E.J. Brill, 1994-2000), electronic ed., 249, הלל I. Hereafter, HALOT.
19. My own literal translation. It should be noted here that zammēr, neither in its meaning, “to sing” nor in its meaning “to praise” takes a preposition; both senses are transitive in Hebrew. Thus, unlike English, which requires a preposition if sing is used to “sing to someone,” in Hebrew there is no preposition.
20. Given the context of verse 16, that is, Nephi had just suffered tremendously from the cruelty of his brothers, another polysemous parallel suggests itself. The root zmr, besides meaning “to sing” and “to praise,” also means “protection” in Amorite (a North-west Semitic language that predates Hebrew) and survives in Hebrew personal names such as Zimri. See HALOT, זמרי. As tempting as this lexeme is, it would require more modification of the tristich than I am comfortable with. In the first place, zmr would have to be read as a verb. But zmr as a verb meaning “to protect” is not attested in biblical Hebrew, though it does exist as a verb in Ugaritic. Also, in order to parallel the first stich, the second stich would have to read “I am protected by Him all the day long,” which would be hard to reconcile homographically with the extant reading, “I did praise him all the day long.” A phonological intimation of such a polysemy however cannot be ruled out (ʾăzammērâ might suggest zimrî, “my protection” or the hypothetical form *ʾezzāmēr “I am protected”).

Go here to see the 3 thoughts on ““Janus Parallelism: Speculation on a Possible Poetic Wordplay in the Book of Mormon”” or to comment on it.