The Authorship of the Song of Songs

Introduction

The Song of Songs (or Canticles) is a collection of passionate, sometimes erotic love poems, ostensibly celebrating the love between a man and a woman. Over the centuries, Song of Songs has been treated with various hermeneutical approaches— allegorical, dramatic, literal-historical, cultic or ritualistic, dream view, typological etc. The lack of consensus on authorship and the large number of interpretive approaches are evidence that Songs is not understood very well.[1] These two articles will adopt an intertextual approach and attempt to contextualize Songs within its historical setting.

Part 1 of this study examines the question of authorship and what the superscription tells us, and it considers whether the Hebrew of the book indicates a date of composition. The next article (Part 2) will aim to demonstrate that that Song of Songs belongs to the reign of Hezekiah and that “Northern Israelite” linguistic features are a consequence of Hezekiah’s courtship of the northern tribes. Comparisons will also be made with Psalm 45 (‘Song of Loves’) identified by Rendsburg as ‘Israelian Hebrew’ (northern origin). The fact that Ecclesiastes shares many of the same linguistic features suggests a similar provenance (i.e. an era of Uzziah-Hezekiah) but that will be left for a future article.

What is the historical setting of Songs?

Many would answer that it was written by Solomon and comes from the period of the early Monarchy (ca. 900 BCE)—however, numerous scholars (based on linguistic evidence) opt for a much later date, even as late the Persian period (Achaemenid rule) 539-323 BCE.[2]   This would seem (to me) to be far too late, but how do we account for the unusual linguistic features?  Moreover, what do we make of the superscription ascribing Solomonic authorship to Songs?  Is there perhaps evidence that points to a different period of composition, somewhere between early monarchy (i.e. Solomon) and late post-exilic (i.e. Ezra/Nehemiah)?   Solomonic authorship is not an intractable problem but the linguistic question is far more difficult to resolve, as we are of necessity reliant on conclusions reached by specialists in ancient near eastern linguistics.

Superscription

Our first subject for analysis is the superscription. J. Reese asserts that,

…the attribution ‘to Solomon’ affixed to the Song is an editorial superscription that links this poetry to Israel’s famous poet and sage rather than a declaration of authorship. No hint of actual author or authors appears in the text.[3]

However, scholars such as Delitzsch, Raven, Steimmueller, and Young have all held to Solomonic authorship;[4] others (Smith, Driver)[5] date it to the late Monarchic period (pre-exilic, before 600 BCE); and T. Longman believes that the superscription should read “which concerns Solomon” rather than “by Solomon”. He makes the following points about the superscription hmlvl rva ~yryvh ryv,[6]

1. The relative pronoun ’ašer occurs only Song 1:1; elsewhere in the book we have the relative particle še (e.g. Song 3:7). This supports the idea that the superscription was a later addition to the book.
2. The preposition l used with Solomon’s name does not necessarily indicate authorship.[7] The semantic range of the preposition is wide, and it theoretically could be translated “of”, “to,” or “concerning”.
3. Hunt offers the interesting observation,

The use of ’ašer only in the opening verse might suggest an archaizing post facto pseudepigraphic device for the book’s title, perhaps in order to render a paronomasic construct on Solomon thus: ašer lî- and šelomoh: š-l: š-l-m [this also appears in the opening of Proverbs [1:1]: mishlê šelomoh: m- š-l: š-l-m], a possible poetic opening gambit.[8]

But even if the superscription is artistic, as Hunt suggests, this does not in itself give us sufficient grounds for pseudepigraphy—we need to analyze the whole book to support that conclusion.

The scholarly discussion is inconclusive. The Targum understands the superscription as authorial but the Vulgate omits it altogether. The superscription could be editorial, but this does not mean the editor is not a contemporary of Solomon, perhaps an amanuensis. The relative pronoun is unique to the superscription but then this could be nothing more than a reflection of the fact that it is a superscription and not part of the poetry; the relative particle could be used for poetic reasons. However, Hunt’s suggestion of a paronomasic construct in v. 1 is suggestive of artistry in the superscription over and beyond a pedestrian editorial superscription.

The LXX translation (o[ evstin tw/| Salwmwn) does not support ‘concerning’ as the earliest interpretation of the Hebrew that we have; were this correct, we would have had the preposition peri.;[9] but, as it stands, the Hebrew preposition could be taken as an indication of authorship or simply that the Song concerns Solomon.

The Language Problem

All languages develop over time; the English language developed from Anglo-Frisian dialects and Anglo-Norman and picked up French, Latin and Greek at a later stage. Writing can be dated if the history of the language is understood. Such things as syntax, vocabulary, phonetics and morphology require analysis and this is a rather specialised field.  Scholars have identified several strands:

• ABH (Archaic Biblical Hebrew) sometimes called “Old Hebrew” or “Paleo Hebrew”—examples are the Song of Moses (Exodus 15) and the Song of Deborah (Judges 5).
• SBH (Standard Biblical Hebrew) sometimes called BH (Biblical Hebrew) or EBH (Early Biblical Hebrew) 8th to 6th centuries BCE; this forms the bulk of the Hebrew Bible.
• LBH (Late Biblical Hebrew) from the 5th to the 3rd centuries BCE corresponding to the Persian Period (post-exilic)—examples are Ezra and Nehemiah.
• DSS Hebrew, (Dead Sea Scrolls Hebrew) sometimes called Qumran Hebrew from the 3rd century BCE to the 1st century CE corresponding to the Hellenistic and Roman Periods before the destruction of the Temple.
• MH (Mishnaic Hebrew) from the 1st to the 3rd or 4th century CE, corresponding to the Roman Period, from example, the Talmud.

W. Dobbs-Allsopp is one of a group of scholars, who after careful analysis of word-order, syntax, Aramaisms and loan-words (mainly Persian) assigns Songs of Songs to post-exilic LBH, nevertheless even Dobbs-Allsopp concludes his monograph with a remark from Pope (Pope 1977:27):

This position is not incontrovertible, of course, as Pope rightly observes,

The dating game as played with biblical books like Job and Song of Songs, as well as many of the psalms, remains imprecise and the score is difficult to compute.[10]

Adding to the complex mix of linguistic identifiers is the recently proposed Israelian Hebrew (or IH), a proposed northern dialect of biblical Hebrew (BH). It is offered as an alternative explanation for various irregular linguistic features of the Masoretic Text (MT) of the Hebrew Bible. It competes with the suggestion that such features are Aramaisms, indicative either of late dates of composition, or of editorial emendations. Although IH is not a new proposal,[11] it only started gaining ground as a challenge to older arguments for late dates for some biblical texts since about a decade before the turn of the 21st century: linguistic variation in the Hebrew Bible might be better explained in synchronic rather than diachronic terms, meaning that various biblical texts could be significantly older than many 20th century scholars supposed. The work done by Rendsburg is particularly interesting; C. L. Seow offers the following review:

Gary Rendsburg has recently distinguished himself as a Hebrew dialectologist. In this new monograph, he proposes a set of linguistic criteria by which one may identify psalms that are written in the northern Hebrew dialect, which he calls ‘Israelian Hebrew’ (IH). By these criteria he concludes that Psalms 9–10, 16, 29, 36, 45, 53, 58, 74, 116, 132, 133, 140, and 141 are all of northern origin, and as a by-product of his investigation, he isolates a list of linguistic features of IH, representing by far the boldest attempt yet to define such a dialect.

Rendsburg begins by identifying the sources for reconstructing IH. These include not only texts that are universally accepted as originating in the north (e.g., Hosea), but also accounts which concern northern figures (various narratives, tribal blessings, Balaam’s Oracles), various passages that scholars have from time to time attributed to northern sources or are said to manifest an abundance of northern linguistic features (Nehemiah 9, Deuteronomy 32, Proverbs, Job, Ecclesiastes, and Song of Songs). Texts that have linguistic or poetic affinities with Phoenician, Aramaic, or the Transjordanian dialects are all said to reflect IH.…By identifying clusters of “IH features,” then, Rendsburg concludes that the psalms in question are all of northern provenance.…

Rendsburg’s contribution cannot be gainsaid. Some linguistic features that have hereinto been regarded too conveniently as archaisms, or as late and foreign intrusions into Hebrew, must now be reevaluated as such. More importantly one is reminded not to harmonize or standardize Biblical Hebrew too readily.[12]

The following summary of the language of Song of Songs is offered by J. A. Cook:

The language of Song of Songs is striking in several ways. Most notable is its use of the relative conjugation še to the almost complete exclusion of ’ášer (which occurs only in Song 1:1). Only the book of Ecclesiastes has more occurrences of še; however it employs ’ášer with almost equal frequency. This feature has been taken to be indicative of Aramaic influence or late date, especially in light of the similar, almost exclusive use of še in Mishnaic Hebrew. However, such conclusions are problematic in light of the appearance of še in archaic Hebrew poetry (Judg 5:7), and the judgement that še alternatively may be indicative of a Northern Hebrew dialect.

Evidence of foreign influence in Song of Songs is found in its occasional Aramaic spellings and foreign loanwords. The Aramaic spelling of n†r (“keep guard”), verses Hebrew ncr, is found in several places in the book (Song 1:6; 8:11, 12), and in one passage the Aramaic spelling BérôT (“juniper”), versus Hebrew Bérôš (Song 1:17), is used. The word ParDës (Song 4:13), often translated “park” or “enclosures,” is a Persian loanword; aPPiryôn (Song 3:9), variously translated “palanquin” (NSRV) or “chariot” (NIV), is possibly a Greek loanword.

The other striking linguistic feature is the number of hapax legomena, thirty-seven in all. Given the size of the book, it contains the highest proportion of such terms of any book in the OT. In addition, F. E. Greenspahn…classifies fourteen of these as “absolute hapax legomena” – that is, forms built on roots that are not used anywhere else in the Bible…[13]

Cook’s conclusion is especially enlightening:

Alternative explanations must take into account variation arising from possible diglossia (i.e., the coexistence of a literary language and a colloquial language) and other differences in registry, the affect of genre and subject matter on language, and the difficulty of distinguishing between archaic language (i.e., the use of older forms of language) and archaisms (i.e., the intentional use of older forms of language in mimicry of earlier compositions). Similarly, judgements on “loanwords” are notoriously difficult to make, in terms of what sort of influence one language might have had on another and in terms of dating.

These questions are often driven by ideological concerns, such as what are acceptable and unacceptable dates for biblical compositions, an undue pessimism about finding answers to these questions, or an a priori commitment to the lateness of all biblical culture and writings.[14]

Alter also expresses reservations about late dating based on language:

When it was more the scholarly fashion to date the book late, either in the Persian period (W. F. Albright) or well into the Hellenistic period (H. L. Ginsberg), these differences might have been attributed to changing poetic practices in the last centuries of biblical literary activity. Several recent analyses, however, have persuasively argued that all the supposed stylistic and lexical evidence for a late date is ambiguous, and it is quite possible, though not demonstrable, that these poems originated, whatever subsequent modifications they may have undergone, early in the First Commonwealth period.[15]

Young contends that Persian loanwords employed in Qoheleth (Ecclesiastes) cannot be used to date the book to the Persian period, as “Persianisms” also occur in pre-exilic books.[16] Young suggests that a probable route for such words was Assyrian deportations of Iranians to the vicinity of Judah in the late eighth century BCE. Young concludes his paper with the words,

Let me sum up the argument of this paper. Linguistic evidence is just that: evidence. It is permissible to use it as one of a series of arguments in attempting to date biblical texts. However, linguistic evidence cannot be decisive.[17]

Conclusion

(1) The superscription could just be an indication of the subject of the poem (Solomon) and not an ascription of authorship.

(2) Hebrew lexicography and syntax alone cannot determine the dating of a book; alternative hypotheses are possible. It needs to be supported by other evidence—socio-historical and intertextual.

---