• Oral Tradition and Internet Technology by John Miles Foley


In medieval times, parchment, produced in heavily labor-intensive fashion from the hides of calves, sheep, or goats, was understandably hard to come by, and scribes responded to this inconvenient scarcity by repurposing their precious commodity.  The method they employed was straightforward enough: they simply rescraped the surface of existing manuscript pages, removing most traces of whatever was encoded there, and then wrote the new text on top.  Whereas it was the original skin-scraping that converted animal hide into medieval desktop, this fresh excising readied an already used folio for a new textual life.  Thus arose the term palimpsest, derived ultimately from Greek palin (“again”) and psao (“I scrape”).  

Naturally enough, this relatively crude method of erasure didn’t entirely remove the earlier layer of writing, and over the years clever investigators developed chemical and light-sensitive tests to detect the left-over words, to ferret out whatever escaped the rescraping protocol.  In the process they managed to recover crucial biblical, classical, and scientific works still extant underneath the overwritten surface.

One famous example of such repurposing is the so-called Archimedes Palimpsest, which retains decipherable traces of works by Archimedes of Syracuse, an ancient Greek mathematician and physicist who lived in the third century BCE.  As fate would have it, a tenth-century copy of several of his works was incompletely scraped and overwritten by a twelfth-century monk as he sought to construct a Byzantine liturgical book, or euchologion.  One of the treatises preserved below the new surface, called “The Method,” survives only in this palimpsest.  Here is a sample page, with both layers of inscription visible:

A sample page of the Archimedes palimpsest from Wikimedia Commons

Even though repurposed, then, parchment palimpsests retain the remnants of prior text-encoding, which persists in shadow-form because of the nature of the medium – scribal intention notwithstanding.  Lurking behind the topmost message is a hidden message, just waiting to be received.  Palimpsests show how hard it can be to remove a once-textualized work from its textual vehicle.


A variorum, Wikipedia advises, is “a work that collates all known variants of a text. It is a work of textual criticism whereby all variations and emendations are set side by side so that a reader can track how textual decisions have been made in the preparation of a text for publication.”  The term derives from the Latin varius, and in this genitive plural inflection means “of the different [people].”

Variorum editions are in the business of presenting alternatives – at least to the extent that texts can accomplish that feat.  vEditions of ancient and medieval works, for instance, display alternative readings from different manuscripts.  Variorums of some modern works, such as the poetry of William Butler Yeats, allow the reader to track the poet’s revisions through successive versions, laying bare an important dimension of the creative process. 

Immediately below is a sample page from the Yeats variorum1. It presents the beginning of the poem entitled “The Hosting of the Sidhe,” documenting the different printings (with bold numbers keyed to a bibliography of items published between 1885 and 1949) as well as furnishing a line-by-line record of how the poet revised for different printings.  In other words, it offers a fixed history of composition.

“The Hosting of the Sidhe,” lines 1-9

Such compendia strive to accommodate variation and change via tAgora technology, to hit a target that did not and will not stand still with a single textual arrow.  They aim to resolve the inherent plurality of (equivalent) multiple witnesses by creating a static constellation of options.  And all that the carefully fashioned product can do is to sit there, static and unchangeable; as always with textual media, it’s up to the reader to introduce or restore the missing movement.  For all of their virtues, texts aren’t inherently kinetic.


So texts can indeed harbor multiple works, either by accident or by intention.  And, correspondingly, a reader can discover and experience their built-in plurality, either by taking the initiative to peer behind an incompletely drawn curtain or by accepting an open invitation to look beyond singularity.  

Palimpsests are accidental multitexts, since at no point in their history as tAgora objects was there any active intention of presenting more than a single work concurrently. Only because rescraping technology left behind a legible residue do we have an opportunity to recover the first – and putatively cancelled – layer of communicative code. And of course that’s no easy task.  Analysts must often take extraordinary measures to decipher what such texts were after all never meant to convey.

Variorums, on the other hand, are intentional.  They exist in order to accomplish one well-defined goal: to present multiple texts concurrently.  Variorums thus amount to an album of photographs, or a series of stills culled from a film, or a flip book.  In their composite presentation they imply movement by providing a sequence of realities, but, being texts, they can’t mimic that movement by themselves.  What they can do is offer the reader an opportunity to investigate the relationships among the multiple concrete instances they cite.  It’s the reader who must flip through the instances to simulate motion.

In both cases, then, multitexts are still and always texts, still and always creatures confined to the communicative ecology of the tAgora.

Unsupported in the oAgora

The oAgora supports neither palimpsests nor variorums, at least as they’re conventionally configured.  Why?  Because the rule-governed variation that characterizes oral tradition never congeals into a solid, invariable text – unless it’s recorded in some tAgora medium and as a result forfeits its identity as OT.  If a work exists only in its continual morphing, if it actively depends upon co-creation and emergence, then a text can never be anything more than a dead end.  Imprisoning Proteus may be the ideological default for we consumers of the book and page, but the moment we scrape the hide and consign living reality to its surface we’ve already exited the oAgora.  The oral marketplace can’t support any kind of encoding on parchment, never mind the rescraping that leads to repurposing with new code.

Likewise with variorums.  The variants that occur in an oral tradition – if that’s even the correct term for forced hesitations along a continuous trajectory of variability – never manifest themselves as finite things to be compared to other finite things.  In a living tradition we can compare recorded versions, of course, but in doing so we’re analyzing not the OT itself but rather arbitrary instances of its textual codification – and therefore of its traces, not its substance.  

This is a crucial distinction for ancient and medieval oral-derived texts as well, since variations among manuscripts are not only evidence of an author’s revisions or of fixed versions created by different scribes, but also of alternatives within the rule-governed flux of the OT web.  Identifying the “best” reading among surviving manuscripts of such works is inevitably a Pyrrhic exercise; no matter how carefully we proceed, the work cannot be wholly or optimally contained in a single text.  It’s larger, deeper, and richer than that.  For such reasons, then, variorum editions, which assume concrete things as their basis, just aren’t suitable as vehicles for oAgora phenomena.  Electronic hypertext projects like the Homer edition pioneered by the Center for Hellenic Studies suit such situations far better than the variorum.

Unsupported in the eAgora

Palimpsests and pathways don’t mix in the eAgora, either. The Word file you revise innumerable times leaves no trace of its step-by-step formulation (unless you save it as different items after each modification or engage the track changes feature, both of these strategies amounting to tAgora interventions).  Electronic “rescraping” can leave no traces of what lies “underneath” because there is no fixed layer to scrape off, no preexisting and still discoverable substrate.  What’s gone is gone.

Nor can the eAgora, which morphs to live and lives to morph, establish the constellation of fixed, invariable versions that constitutes a variorum edition.  Why?  Because, as in the oAgora, such data don’t exist.  So how about option-driven navigation using hypertext, you might wonder; isn’t that a kind of variorum?  Well, not really.  The IT (Internet Technology)-enabled marketplace supports a categorically different kind of experience for the user – one that involves co-creation, a web of options, and ongoing variation within limits, one that reflects the oAgora situation much better than a conventional text.  It might recall brick-and-mortar collections of items, but this facsimile experience is worlds away from listing manuscript-based options in a variorum edition.

The websites you visit can and do evolve without notice or a backward glance, so that you’ll need an RSS feed reader like NetNewsWire to keep abreast of new developments.  Oh, you can search the Wayback Machine to unearth previous versions of a site, but each version qualifies as a unique eWorld unto itself. You visit one of them, and you navigate within its universe; then you visit another, and so forth.  In effect, you’re using a tAgora-type library in order to access disparate, Balkanized eAgora itineraries.

What OT/IT supports and TT does not

Let’s close by turning the tables and looking at the situation from the opposite direction.  Here’s a simple adage for determining supported enterprises in the oAgora and eAgora: everywhere there is true interactivity there is also rule-governed morphing.  Change is driven by co-creativity and emergence, and is so continuous and enabling that it’s effectively invisible.  In other words, such morphing can’t be grasped while reality remains in play, and it can’t be halted without derailing the interactive process.  Finally, it’s a matter of what marketplaces permit and what they demand: oPathways and ePathways require users to surf them.  Not to mindlessly reduce their potential options to hard-and-fast singularity, but to participate in the making of what is happening right now and will vanish forever as soon as it happens.  Of course, the ready consolation for “lost” experience is the ever-present opportunity to generate another, parallel, in-the-moment experience.  And so it goes.

Neither palimpsests nor variorums are possible in the oAgora or eAgora because the process of composing, performing, and receiving remains forever undocumented within these agoras.  It can be, and often is, sampled and concretized by employing tAgora technology to manufacture an item or series of items.  But such items are responses to textual ideology, to conducting verbal business in a marketplace that values the brick-and-mortar to the exclusion of the virtual. In fact, the viability of the co-creative, emergent process—of surfing through a linked web of OT or IT—actively depends on that process remaining forever undocumentable: fluid instead of fixed, ongoing instead of frozen.  From the perspective of the oAgora and eAgora, documents are merely artificial interruptions of the flow.

In summary, then, here are some dynamics that OT and IT support and textual technology (TT) doesn’t:

1. Navigating through evolving webs of pathways.
2. Morphing according to the rules in force.
3. Managing systems rather than swapping tangible things.
4. Co-creating reality, which is jointly shaped and experienced even as it evolves.


1 Peter Allt and Russell K. Alspach, The Variorum Edition of the Poems of W. B. Yeats (New York: Macmillan, 1957).