Digital Scholarly Editing Blog Post #2 – Observations on TEI marking-up Samuel Beckett’s Krapp’s Last Tape

This blog post continues the rationale of the first post written on the subject of annotating a selection from James Joyce’s novel Ulysses (1922). As was argued in that first post, it is important that digital humanities projects and projects involving TEI mark up, provide a statement of intent and rationale that can provide the researcher and ‘common reader’ alike with information about the text and the editorial processes that a text has been subject to. This should go beyond the information provided in the , or elements in the . This particular blog post is intended to serve as a gloss to the process of marking up an excerpt of Samuel Beckett’s short play Krapp’s Last Tape (1958) in TEI.

In script form, Krapp’s Last Tape is only pages long, but can, depending on the production, run between thirty and sixty minutes. This variation in production length can be attributed to how long each particular production spends on enacting the play’s many pauses, ellipses and mechanically repetitive stage directions.

This text was chosen for mark up because it is a dramatic text, and therefore requires a different range of elements and tags than the prose extract from Ulysses. Furthermore, the TEI mark up project that will be carried out, the ‘Circe’ episode from Ulysses, will be combining the prose elements and dramatic elements, so the marking up of a play script will provide an opportunity to become familiar with a different mark up methodology.

Krapp’s Last Tape provides an interesting challenge to the process of mark up because of its paradoxical use of voice. Most plays have more than one character, but Krapp’s Last Tape could be regarded as a partial exception. The play represents Krapp, a man who records his thoughts on the year that has passed on his birthday every year. The play takes place on his sixty-ninth birthday, a year in which he decides to listen to a recording he made on the same day thirty years ago. This tape, from when he was thirty-nine, appears as a voice in the piece, but the extent to which it can be regarded as a different character is uncertain, considering that it is Krapp who is speaking on it, albeit when he was thirty years younger.

This present the first difficulty to the TEI code, but it was decided that it the original text’s line on the matter be maintained, that Krapp be credited as a speaker and the Tape be credited as a different speaker. In the line of the code at least, there will be no overt link made between the two, to suggest that they are essentially the same person.

The source text for this TEI version of Krapp’s Last Tape is not from a copy of the play text that is published on its own. Instead, it is part of Faber & Faber’s Complete Dramatic Works of Samuel Beckett. This is why it is declared as the source text in the TEI Header.

There are frequent pauses in the dialogue and in the parts between stage directions in the play script. They appear as “[Pause.]” TEI provides a number of different @type attributes for the purposes of stage direction, such as ‘delivery,’ ‘business,’ or ‘modifier,’ to provide just three examples. It was ultimately decided that the use of the @type attribute ‘delivery’ would be most suitable for the conveying of the pauses. This was done despite the fact that the TEI-C website suggests that it be used for the way in which a character speaks as it represents a withholding of speech or action on the part of the performer.

The stage directions that describe Krapp’s repetitive movements, whether he be walking off stage, opening and closing drawers or switching tape reels are tagged with the element , with the @type attribute ‘instruction.’ It was decided not to formulate an excess of @type attributes for the element for fear of creating a sequence of disparate categories that would separate them out too finely from one another. Nesting <stage=’exit’> within <stage=’tech’> within <stage=’business’> in turn would seem counter-intuitive based on the kind of playwright that Beckett is. Furthermore, Beckett was well known for his micro-managing style of direction and high standard of perfectionism when producing his own plays. This frequently manifested itself as an absolute refusal to deviate even in the slightest way, as he saw it, from the text as it was written. Many performers of Beckett today speak of his stage directions with a kind of reverence. It is this scrupulousness that merits, for the purposes of this exercise, the use of the @type attribute ‘instruction,’ in order to emphasise the importance of accuracy in a staged iteration of Krapp’s Last Tape. This is, of course, not to suggest that this TEI encoded excerpt of the play is to be performed in any context, but merely to maintain a kind of integrity to the source material, which is in this case, contingent on the work of the editor.

This apparent digression is important to consider, as it seems that for the element that there is far more scope for the editor’s own interpretation and judgement than is usual in TEI. As was said earlier, the TEI-C website provides a number of potential @type attributes which can be considered quite productively in this context. The @type attribute ‘novelistic’ is particularly so. The TEI-C’s example of a novelistic stage direction in as follows: Having had enough, and embarrassed for the family. This is a stage direction from Lorraine Hansberry’s play A Raisin in the Sun (1959). It is obvious why this is read as a stage direction that is more characteristic of novelistic discourse than dramatic play-texts. It deals overtly in interiority and the character’s (Ruth, in this instance) motivations, which would typically be left out of a stage text and left for performers to articulate on their own terms. However, contained in this @type attribute is a closed-off definition of ‘novelistic’ discourse as realistic and transparent. While Krapp’s movements may be monotonous and not particularly illustrative of an inner state in the same way that the example provided above is, it is neither accurate, nor particularly discriminating to in turn suggest that novelistic discourse that does not partake of this particularly, readable language is only novelistic. A number of Beckett’s texts, or writers that have been influenced by him often describe at length a sequence of repetitive motions that try even the most patient reader’s attention, in order to parody and subvert the monolithic notion of what a novel ‘should’ be.  While this would seem a punctilious point to make at first, it is important that users of TEI recognise its occasional interpretative biases. As Jerome McGann writes: “it is not only difficult and time consuming to implement, but its hierarchical principles and other design characteristics set permanent and unacceptable limits on its usefulness with arts and humanities materials.”[1] The history of the TEI-C displays the fact that they have often functioned democratically as a consortium and adapted their methods of standardisation according to use, until they have come to formulate the open, almost endlessly malleable to develop the open, almost endlessly malleable means of marking up texts that it perpetuates today.

Despite the protestations of theorists such as McGann, the contention of this post is that from experience of using TEI, it is for the most part an objective framework. However, there are instances, as in the case of the @type novelistic that it throws up some contentious but interesting questions about naming and tagging. This post will once again assert the importance of interrogative critical apparatuses in conjunction with a mark-up project. Problematic tags such as ‘novelistic’ do not need to be expunged altogether from the TEI, but it is important that debates take place as to whether some tags are better than others at maintaining the objectivity that the TEI should be striving for. This is demonstrated by the writings of Susan Hockey on TEI, who points to the fact that when it was being developed initially, it emphasis was on the ordering of a text for ease of navigation. This was at the expense of interpretative or analytical mark up languages presumably because these would be far more problematic for the obvious reasons. Furthermore, TEI was developed in order to combat the proliferation of mark-up languages which risked limiting the interoperability of digital texts. The TEI-C’s evolutionary framework and ability to continually update its schema is a good reason why the process of its development should not come to an end, but rather continue. Critical analyses, debates and interpretations such as those contained within this post should help to further this.

[1] McGann, Jerome, Radiant Textuality: Literature After the World Wide Web (Palgrave: 2001), p.17


Beckett, Samuel, Complete Dramatic Works (Faber & Faber: 1986)

McGann, Jerome, Radiant Textuality: Literature After the World Wide Web (Palgrave: 2001)

Schreibman, Susan, Siemens, Ray and Unsworth, John, A Companion to Digital Humanities (Blackwell: 2004)

Schreibman, Susan and Siemens, Ray, A Companion to Digital Literary Studies (Blackwell: 2008)

Text Encoding Initiative Consortium Website,


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s