Love poetry, explicit and unembarrassed, with no mention of God, no law, no covenant and no history. Its presence in the canon is the most interesting fact about it.
What the book is
The Song of Songs is a sequence of love poems, voiced mainly by a woman, addressed to and about a man, with an intermittent chorus. The body is described at length and with pleasure. There is no plot, no divine speech, and no reference to Israel’s history, law or worship. God is not named in the book.
Its canonical status was disputed in antiquity and the dispute is recorded. What secured it was the allegorical reading, in which the lovers are God and Israel or, later, Christ and the church. That reading is old, but it is not the plain sense of the text, and it was applied to the book rather than found in it.
How it is put together
The unity of the Song is argued over. Some read it as a single composition with a developing structure, others as an anthology of separate lyrics collected around a common theme. The repeated refrains — the adjuration not to awaken love before it is ready, the invitations to the garden — are the main evidence for design.
The superscription attributes it to Solomon, but the Hebrew phrase can mean by, for, or concerning Solomon, and Solomon appears in the poems as a figure spoken about rather than a speaker.
The dating question
The linguistic evidence points late. The Song contains a Persian loanword — the word behind orchard or park — and features of vocabulary and syntax that align with post-exilic Hebrew and with Aramaic. On that basis a Persian or early Hellenistic date, roughly the fourth to third century BCE, is the common position.
Against that, some scholars argue that the northern dialect features and the close resemblance to much older Egyptian love poetry point to an early core, later edited. The debate has not been settled, and this is a book where the internal evidence is all there is.
Where the external record touches it
There is no external attestation of the Song as a text or of anything it describes, and there could not be: it narrates nothing datable.
The comparative record, though, is unusually good. Egyptian love poetry of the New Kingdom, roughly the thirteenth to twelfth centuries BCE, survives in several collections and uses the same conventions: the lovers as brother and sister, the garden as the body, the description of the beloved part by part, the obstacles and the go-betweens. Mesopotamian sacred marriage texts supply a second set of parallels.
What the comparison shows is that the Song belongs to a well-attested regional genre of secular love poetry. That is an argument about what kind of text it is, and it cuts against the allegorical reading being original to it.
A note on how it reads
The woman speaks more than the man, initiates more often, and is the one who searches the city at night. In a canon where women rarely hold the voice, that is worth noticing.
The Song is also the source of a large amount of later devotional literature, particularly medieval, which read it allegorically in enormous detail. That reception is historically important and it is a separate question from what the poems meant when they were written.
Related: the overviews of Ecclesiastes and Proverbs, the other books attributed to Solomon, and the method note on what kind of text this is.