Bad Quartos and the Myth of the Memorial Reconstruction

Shakespearean scholars have long imagined that Elizabethan actors would help publish pirate copies of Shakespeare’s plays for a few shillings on the side by reconstructing them from memory, but could this have actually happened?


When the First Folio of Shakespeare’s plays was published in 1623, half of its thirty-six plays had previously been published individually in quarto or octavo format. The terms quarto and octavo refer to the size of the booklet, quarto being a larger page folded in four, while for octavo the page is folded into eight. However, in Shakespearean scholarship the term ‘Quarto’ has taken on a life of its own.

Some of the quartos are so close to the First Folio version that it is evident the Folio text was based directly on the quarto edition. Many of the plays are represented by several editions of quartos, some closer to the First Folio text than others. The range of differences between the First Folio text and the earlier quartos therefore varies considerably.

Sometimes a quarto may be a shorter, cut down version of the First Folio text. In some cases the differences between multiple editions of quartos can be attributed to changes made to the play over time. Often the differences between the quarto and the First Folio text are so slight they can be accounted for as typographical errors or minor amendments. However, a handful of quartos are so different to the First Folio text that Shakespearean scholars have dubbed them ‘Bad Quartos’.

The differences between these ‘Bad Quartos’ and the First Folio are so great that they cannot be attributed to accidental typographical errors or even extensive editing. Many scholars believe that the differences are in fact corruptions of the true text of the play which came about when the plays were pirated by unscrupulous stationers.

The theory is that these Bad Quartos were produced through either stenography or memorial reconstruction. In the case of stenography, a scrivener might attend a play and take it down in shorthand, then reconstruct it for the stationer. The concept of the memorial reconstruction has been one much favoured by Shakespearean scholars. In this method, it is theorised, one or more actors who performed in the play reconstruct it from their collective memory, in order to make a few shillings on the side by selling it to a stationer. Many editors of plays with supposed Bad Quartos go into extensive discussion as to which members of the cast were involved in the reconstruction and attribute to their lapses in memory garbled lines and inclusions of lines from other plays and even other playwrights.

Some scholars, Eric Sams amongst them, decry the whole concept of Bad Quartos and memorial reconstructions. Maintaining that Shakespeare started writing much earlier than generally believed and that he was a lifelong reviser, Sams believes that the so-called Bad Quartos are simply Shakespeare’s own earlier versions of his plays. Whether or not Sams is correct, there are enough flaws in the Bad Quarto theory to render it very much in doubt.

The ‘Bad Quarto’ theory is derived from the belief that not only was Shakespeare the pre-eminent playwright of his day, but in fact, the only one and one that was so original that every play that resembles one of his must necessarily be either by him or derived from one of his plays. However, this reasoning can lead to grave errors.

E.A.J. Honigmann, for example, the editor of the 1954 Arden Shakespeare edition of The Life and Death of King John, not only based his dating, but also his interpretation of the play, on the assumption that The Troublesome Reign of King John, first published in 1591, was a hastily produced Bad Quarto, thus dating Shakespeare’s play to 1590. This dating led Honigmann to posit that the play was being written at the same time as James Burbage was quarrelling with Margaret Baynes and that the action of the play owes more to what was happening at The Theatre than to King John’s actual biography. In actual fact, this could not have been the case as more recent editions give the play’s likely date of composition as 1595/6. (However, I am still grateful to Honigmann for bringing that fascinating story to my attention. See The Troublesome Reign of James Burbage – Part One )

However, The Troublesome Reign of King John has since been identified as a play in the repertoire of The Queen’s Men (see The Elizabethan Playing Companies). Several other plays similar to Shakespeare’s, such as The True Tragedy of Richard III (first published in 1594), The Famous Victories of Henry V (first published in 1598) and The True Chronicle History of King Leir and his three daughters, Gonorill, Ragan and Cordella (first entered in the stationers’ register in 1594), all previously considered Bad Quartos, have also been identified as plays belonging to the Queen’s Men. It is thought that Shakespeare was so familiar with these plays before they were published that he may have been a member of the company early in his career.

First Quarto title page of The Taming of the Shrew
First Quarto title page of The Taming of the Shrew

Another play that has been identified as a Bad Quarto is The Taming of a Shrew (first published 1594). While there is only an article differentiating this title from The Taming of the Shrew, the texts of the two plays are very different. A Shrew is set in Athens rather than Padua; except for Christopher Sly and Kate, all the characters have different names; Kate has two sisters instead of one and each has only one suitor; Christopher Sly reappears at the end of the play; and much of the dialogue is lifted from plays by Christopher Marlowe.

In his 1981 Arden Shakespeare edition of The Taming of the Shrew, Brian Morris proposes that the player who contributed the memorial reconstruction played Grumio, Petruchio’s servant, and that as he had such a small part he would have misremembered much of the play. However, even the player with the smallest part could not have missed the fact that The Shrew was set in Italy and that Baptista had only two daughters, or forgotten the names of almost all the characters including his own and that of his master.

Rather than a memorial reconstruction, it is much more likely, as Stephen Roy Miller proposes in his 1988 edition of The Taming of a Shrew, that it is a deliberate adaptation of The Shrew, made by an anonymous author who saw himself as a ‘play doctor’, improving the play to appeal to a more discerning audience and bringing it more in line with Commedia Erudita than Commedia dell’Arte (see The Taming of the Shrew and Commedia dell’Arte). Nonetheless, for the purposes of copyright, the two plays were considered one and the same, so that Shakespeare’s original play was prevented from being published until the First Folio and was not published in quarto until 1631.

Apart from the scholars’ speculation, there is, in fact, no real evidence of the practice of either stenography or memorial reconstruction. There is only one reference in the documents to the practice of reporters surreptitiously taking plays down from performance. However, if, as has been suggested, Shakespeare’s lines were delivered rapidly, it is hard to imagine an Elizabethan scrivener, using a frequently inked quill on rough paper, being able to get down anything more than the gist of the dialogue.

Again, as beloved as the memorial reconstruction theory may be, neither is there any contemporary evidence of that practice. The only recorded occurrence of anything similar was in the eighteenth century, and it was carried out by a theatrical entrepreneur, not a printer. Furthermore, as Marlovian scholar, A.D. Wraight has pointed out, the very idea that players, with their highly trained memories, would have garbled and mixed up their lines, or slipped in passages from other plays, is highly improbable.

Putting aside the issue of Bad Quartos, as we saw earlier, not only are most quartos different to one degree or the other to the First Folio text, some plays have several editions in quarto, each differing more or less from the First Folio text. If we discount the possibility that these versions are surreptitiously pirated editions drawn from memory or stenography, we must come to the conclusion that they were derived from a written text.

All plays would have gone through three written stages. Their original composition by the author/s could have generated several drafts, known as foul papers, before the final draft was produced. This final draft may well have become the official playbook which had to be submitted to the authorities for a licence. At most a second complete text may have been produced as a working copy for the bookholder (the equivalent of the modern stage-manager). The bookholder would then prepare scrolls for the players on which were written only their own lines introduced by their cues and their stage directions.

Any printed versions must have derived from one of these stages. The official playbook and the bookholder’s copy would have been closely guarded and by definition could not be the source of the more corrupted versions of the text. They could only have been delivered to the printer with the authority of the company.

Many scholars believe that the author’s foul papers were often used to produce quartos. I find this most unlikely as authors had no reason to keep their foul papers which would have been quite illegible. I would suggest that the most likely sources for the less accurate, or cut down quartos would be the players’ scrolls.

I would imagine that, once the players had their lines, their scrolls were collected, collated and stored for later use, in case, for example, a new player joined the company or changes needed to be made to the casting. There would also have been separate sets of scrolls for different version of the play, say the complete version and the cut down touring version. These scrolls would have been much easier to purloin and/or pass onto stationers than the playbooks. Such scrolls would most likely contain lots of cuts and amendments or be missing a page here and there, all of which would account for the corruption of the printed copy.

On the other hand, we sometimes come across plays where it is obvious that more than one version of a speech has been included in the final printed version. This may have occurred when the play has been printed from an amended playbook. As only one or two copies of a complete play were produced, any amendments, expansions, cuts or edits to the play script would have been made to the official copy. This might have been achieved by slipping new pages into the script, or cutting out and pasting the new version of a speech over the old one using paste which might eventually lose its grip, especially after many years of disuse. One can imagine a confused compositor finding loose bits of paper, losing some, or, not knowing which was the right one, printing two versions of a scene.

Errors could also have entered the text during the transition from the handwritten to the printed text. The fastest way to make a copy of a written text is by auditory transmission i.e. one person reads the text aloud to another who writes it down. The text could have been dictated by the bookholder to a scrivener who wrote a copy of the play for the printer. Alternatively, the text could have been dictated directly to the compositor as he set the type. In such cases errors could have crept into the spelling and lineation. Other errors might have occurred if the compositor had trouble reading the handwritten text.

One might even bring together the theory of memorial reconstruction with these two new suggestions and posit that in order to make some money on the side either a player, or the company itself, takes the scrolls of a popular play to the stationer. Where the scrolls are difficult to read, the printer might ask the player to dictate the lines either from the scroll or from memory. Where later, more accurate quartos exist, it may have been that the company itself made the playbook available to the stationer.

As unethical as the publication of these quartos might have been, after The Globe burned down in 1613, the editors of the First Folio may well have been grateful for them, as they may have been the only surviving copies of some of Shakespeare’s plays.


© Pauline Montagna 2021



Honigmann, E. A. J. (edited by) The Life and Death of King John, Arden Shakespeare, Methuen (1954)

McMillan, Scott and Sally-Beth McLean, The Queen’s Men and Their Plays, Cambridge University Press (1998)Miller, Stephen Roy (edited by) The Taming of a Shrew: the 1594 Quarto, Cambridge University Press (1988)

Morris, Brian (edited by) The Taming of the Shrew, Arden Shakespeare, Methuen (1981)

Sams, Eric The Real Shakespeare: Retrieving the Early Years, 1564–1594, Yale University Press (1995)

Wraight, A.D., Christopher Marlowe and Edward Alleyn, Adam Hart (1993)

Back to Top