What Is Oral Formulaic Composition? Building a Poem From Ready-Made Phrases
By the BrainSnail editorial team. How these articles are written and checked, and how to tell us when one is wrong.
Long poems composed and performed without writing are not memorised word for word but assembled in performance from a stock of phrases that fit the metre. Working that out explained several things about ancient epic that had puzzled readers for centuries.
The problem it solved
Homeric epic contains a striking amount of repetition, with fixed phrases attached to particular characters and objects, whole lines and passages recurring, and descriptive epithets applied whether or not they suit the moment, so a character is swift-footed while standing still. Readers had noticed these features for centuries and had treated them as stylistic choices, as marks of a primitive stage of composition or as corruptions. The formulaic explanation accounts for them as a working method, since a performer composing in real time cannot pause to invent, and a stock of phrases of known metrical shape allows lines to be assembled at speed while the performer attends to the narrative. On that account the repetition is not a defect but the visible residue of how the poems were made.
How the system works
The stock of material is organised by metre and by meaning together:
- •Formulas are phrases of fixed metrical shape expressing a given idea, so a performer selects by what fits the space remaining in the line
- •Sets of formulas cover the same idea in different metrical shapes, giving options for different positions
- •Economy is characteristic, with generally only one formula available for a given idea in a given shape, which minimises what must be held
- •Themes are larger recurring units covering typical scenes such as arming, feasting, arrival and single combat
- •Story patterns organise whole narratives into shapes that can be filled with particular material
- •The system is learned by apprenticeship rather than taught explicitly, and performers frequently describe themselves as reproducing rather than composing
How it was established
The argument was made empirically rather than from the texts alone, which is what made it convincing. Milman Parry analysed the Homeric epithets statistically and demonstrated the economy and extension of the formula systems, showing a degree of organisation that deliberate literary choice would not produce. He then went to Yugoslavia in the 1930s with Albert Lord and recorded living performers of oral epic, who composed long narrative poems in performance without writing, and documented that they used exactly such systems, that their performances varied between tellings while the story remained, and that they described their own practice in terms of accurate reproduction despite the variation. That fieldwork converted a hypothesis about ancient texts into an observation about a living practice, and the recordings remain a major archive.
Memory without writing
The findings connect to a wider question about how much can be transmitted without writing, and the answer is more than modern readers expect. Oral traditions have preserved genealogies, legal material, ritual instruction and technical knowledge across many generations, and the techniques supporting that are identifiable, including metre, rhyme and alliteration which constrain what can be substituted, formulaic structure, narrative organisation, and specialist practitioners trained from childhood. Accuracy varies enormously by tradition and by what is being transmitted, since a tradition treating exact wording as sacred achieves far higher fidelity than one treating the story as the thing preserved, and the Vedic tradition in particular developed recitation methods producing transmission accurate enough to be verified against later written versions. Assuming oral transmission is unreliable is therefore as mistaken as assuming it is exact.
Where the argument goes
The implications reach well beyond Homer and have been contested throughout. Applying the criteria to other traditions produced a substantial comparative field, with Old English, Old French, South Asian, Central Asian, African and Native American traditions analysed the same way, and the results are uneven, since some texts show the expected density of formulas and others do not. Critics have argued that the presence of formulas does not establish oral composition, since a literate poet imitating the style produces them too, and that the approach undervalues the artistry of the texts by making the poet an assembler. Defenders respond that the method does not preclude artistry and that the best performers are distinguished precisely by what they do with the system. The question of how the Homeric poems reached written form remains open and is argued in these terms.
The takeaway
A performer composing in real time assembles lines from a stock of phrases of known metrical shape, which explains the repetition and the epithets applied regardless of the moment. Parry demonstrated the organisation statistically and then recorded living performers doing exactly that in Yugoslavia. Whether formulas prove oral composition is contested, since a literate imitator produces them too.