Arden of Faversham contains an unusual number of characters, both main and minor, who tell stories. The main storyteller, Franklin, is the play’s only fictitious character – one who has affinities with another story–teller of the same name in a collection of tales, Chaucer’s Franklin. This essay argues that the authors of Arden use the story-telling characters to explore the function of narrative within drama.

None of Marlowe's plays or poems exist in manuscript (for one partial exception, see the discussion of The Massacre at Paris, below). Our earliest witnesses are printed. Printed texts reveal a great deal about the circumstances of printing; but they can also be encouraged to speak about the circumstances of composition and consumption. A chapter about Marlovian texts and authorship is thus also a chapter about critics and readers, about tastes and preferences: not just about what Marlowe wrote but about how it was received.

The first of Marlowe's texts to reach print was Tamburlaine, possibly his first play. On 14 August 1590 the publisher Richard Jones made an entry in the Stationers' Register (the register in which publishers entered their right to a work) for the two parts of Tamburlaine. In the same year he published both parts as a single volume, in a small octavo format.

The title page is an endearing example of early modern advertising. It provides a racy plot summary, boasts of recent stage success, and promotes the quarto as hot off the press:

THE POEM

With suspect texts containing between 757 and 3,416 lines of dialogue (the totals for A Yorkshire Tragedy and Q Richard 3, respectively), the Poem provides our largest catchment area of material for evaluation. This verbal matter can be affected by a number of agents: revising or indecisive authors, playbook annotators, actors, scribes, and compositors. The amorphous category of repetitions, for instance, which critics cite as strong evidence of memorial reconstruction, is easily attributable to all five of the above. How, then, do we distinguish a memorial repetition from any other?

REPETITION

External Echoes/Recollection

Harry Hoppe claimed to have identified twenty-four ‘true inter-play borrowings’ (also known as ‘echoes’, ‘parallels’, and ‘external recollections’) in Q1 Romeo and Juliet. All of his ‘borrowings’ come from Shakespearean texts, twenty of them from plays written in the same period as Romeo and Juliet (Richard 2, The Taming of the Shrew, 3 Henry 6, Titus Andronicus, The Two Gentlemen of Verona, and Richard 3); thus, the possibility of authorial self-borrowing is strong, and undermines any value Hoppe's parallels may have.

His parallels are of negligible value anyway; they contain proverbial phrases (e.g. The Merry Wives of Windsor: ‘we burn daylight’), or common sentiments (Richard 2: ‘And then be gone’) or refer to parallel situations (The Two Gentlemen of Verona: ‘a ladder quaintly made of cords’).

It should be apparent from the previous chapters that much New Bibliographic analysis of suspect texts is confirmatory rather than diagnostic. The features of suspect texts (texts suspected of being reconstructed from memory) are measured against a priori expectations of the features a memorially reconstructed text should exhibit. When the features of the material suspect text agree with those of the hypothetical model of memorial reconstruction, this congruence is taken to ‘prove’ memorial reconstruction. Small wonder that so many texts have been designated memorial reconstructions.

In articles and introductions to editions, New Bibliographers regularly list the features allegedly typical of, and confined to, suspect texts. The following comments on Q King Lear are representative:

In his presidential address to the Bibliographical Society in 1930, W. W. Greg reviewed the achievements of the textual colleagues with whom he was associated ‘in what will perhaps one day be recognised as a significant critical movement’. That critical movement came to be known as ‘New Bibliography’ and, as represented in editorial practice and textual journals in Britain and North America, assumed a significance and supremacy which Greg's modest prolepsis could scarcely have envisaged. With its emphasis on the material book rather than its literary content, and its formidable array of technical tools (to say nothing of vocabulary) for analysing textual evidence, the New Bibliography heralded a brave new world in editorial epistemology and procedure. By analysing all aspects of textual transmission, and formulating principles for editing, Greg and his colleagues aimed to systematise the editing of Elizabethan drama, thus delivering Shakespeare and his readers from the inaccuracies, eclecticism, and taste of previous editors.

In the fifteen-year period 1594–1609, versions of eighteen recently performed Shakespearean plays reached print in inexpensive quarto or octavo editions. From the beginning of the Shakespeare editorial tradition in the eighteenth century, editors and textual critics have been particularly perplexed by five of these playtexts: Q1 Romeo and Juliet (1597), Q1 Henry 5 (1600), Q1 The Merry Wives of Windsor (1602), Q1 Hamlet (1603), Q1 Pericles (1609).

This perplexity was caused by the suspicion that the London playreading public was being offered a product very different from that given the London play-going public (even though all of these texts advertised recent performance on their title-pages). Certain unusual features in these first editions (imperfect metre, blunt dialogue, and allegedly jejune depiction of characters and development of situations) were deemed incompatible with composition by a proficient commercial playwright and performance by a leading company. Furthermore, these five early playtexts were all suspiciously brief (particularly in comparison with subsequent quarto versions released during Shakespeare's lifetime, and with the Folio collection of Shakespeare's works prepared by his theatre colleagues, Heminge and Condell, after his death).

The response to the New Bibliography was enthusiastic. New Bibliography's ‘scientific method’ represented ‘a long-overdue revolt against the sterile warfare of theory-mongers, and cleared the air with an invigorating breeze of good sense’; it gave ‘Shakespearean scholarship a new confidence, infusing into it a new spirit of optimism’; the result was ‘that the history of literature will have to be largely re-written’. A specific focus of praise was the identification of suspect texts as memorial reconstructions by actors. ‘The establishment of the Orlando class of “bad” Quartos is the most important development… in the study of pirated dramatic texts’; ‘the chief achievement of the Greg school of… scientific bibliography [is] the distinction between “good” and “bad” quartos’.

Theories of memorial reconstruction did not begin with the New Bibliographers. The use of memory in reporting playtexts had been suggested by editors of Shakespeare from the nineteenth century. In 1842 and 1843, respectively, Collier attributed Q1 Romeo and Juliet and Q1 Hamlet to shorthand notes, supplemented by the compiler's memory, or by assistance from an ‘inferior writer’.

ORAL CULTURE

Elizabethan England was not a ‘primary oral’ culture in the sense that Homeric Greece was with its dependence on transmitting material (legal, historical, poetic) memorially from one mind to another. Sixteenth-century England was ‘residually oral’; a society in transition, en route to becoming a documentary society, it nonetheless had very strong roots in the oral world, whether through habit, illiteracy, or expense of paper. Aides-mémoire abounded, but they were, literally, what the noun suggests: aids to memory, not a substitute for it. So important was memory to the culture that numerous systems existed for stretching its capacity and enhancing its accuracy. In dealing with the subject of memorial reconstruction we must acknowledge that actors were not the only people with trained and efficient memories.

Walter Ong has suggested that we can measure the degree of residual orality in a chirographic culture from the emphasis accorded memory in its education system.

Up to now my investigation of memorial reconstruction has concentrated on the twentieth century and the theories offered by textual critics to explain suspect texts. I now turn my attention to the sixteenth and seventeenth centuries, to ascertain what external evidence exists for invoking memory, reconstruction, and actors, independently or together, to explain suspect texts.

As we shall see, what evidence there is supports the theory that plays could be, and were, reconstructed from memory. However, the evidence suggests that the concept of play reconstruction is not a single sequence of processes with a single objective (reconstruction from memory by actors for the purpose of performance) but a variety of processes involving different kinds of agents (auditors, actors, or other playhouse personnel), different purposes (performance, publication, private transcript), and different combinations of memory and other aids to recall (longhand notes, shorthand notes, reports of narrative outline) often supplemented by new composition.

The brevity of this chapter testifies to the paucity of external evidence. The evidence, such as it is, is nonetheless complex, drawing on a wide range of genres, periods, and cultures. My departure point is not Elizabethan drama, but its parallel in terms of popular and edifying entertainment: church sermons.