This issue is not completely new; the topic of annotation has been examined by Marshall [1] and that of the rhetoric of links by, among others, Landow and Bernstein [8][9], and as well in Literary and Digital Literary Studies. In the latter case, scholars aim to reconstruct the thought processes of a specific author in order to understand how they created significant works at a given point in time. From a technological point of view, the goal is to create technologies that facilitate the production of content for authors, by task automation (for example, versioning) and by supporting cooperative authoring.

Intra-document technologies support the management of forward connections1 within a single document or collection owned by the author, such as changelogs, versions, and live documents2. An

1 Diachronic connections from older documents toward newer documents. 2 Such as Etherpad, WattPad, Google Docs, Dropbox.

emblematic example is Google Wave, which attempted to connect preliminary ideas, notes, diagrams, meetings and messages between collaborators to a finalised work in a seamless process. On a parallel note, literary scholars have been studying for several decades how authors write new works, in particular from the Romantic period (late eighteenth century) onwards, where intermediate process documents such as authorial drafts and versions have been preserved in research libraries and are used to reconstruct the complex development of a literary text [10]. Increasingly, marginalia in books owned by authors of the Romantic age are perceived as valuable in reconstructing their thought processes and lives [11]. Summarizing, intra-document technologies provide an evolutive representation of works, generating links to the past from intermediate process documents.

Inter-document technologies support the management of forward connections among external sources toward the author’s work. Two emblematic technologies in this area are bookmarking (e.g. the del.icio.us social bookmarking platform) and reference management systems (RMS), as a specialized evolution of bookmarking. Bookmarking first and RMS later give authors the means to archive references for future use, connecting the phase of studying and collecting ideas with the actual production of content. These technologies had two orders of issues. Firstly, we have a problem of misalignment between archiving systems, based on objective description of a resource, and the rationale for archiving, which is based on the potential role and function of a resource for future work. Secondly, as a result of this mismatch, these tools do not allow reflection-in-curation [12] which provides interpretation of the information value of resources, in particular in relation to future works. Summarizing, inter-document technologies provide a buffer, a temporary memory space where to hold resources that may be of use in the future.

In both cases, the limitations are dependent upon the absence of a language of connections: symbols and semantics of, for instance, roles, functions, uses and judgments expressed by the author: a clear universal semantic of these “links to be”. On the one hand, within intra-document technologies the lack of semantics is replaced by an extensive representation of connections. On the other hand, within inter-document technologies the lack of semantics can be addressed only through loosely structured or unstructured user-defined information, such as folksonomies and notes.

Authoring tools already incorporate some form of annotation to support the writing process. For example, in Data-driven storytelling [13], the annotation of data charts has an important role in supporting the narrative which will be based on annotation of data charts. Platforms that implement the sculptural hypertext paradigm provide support to the author that is comparable to that of marginalia for the iterative construction of narrative content. This approach provides a workflow and tools to support the authorial work [14][15]. The limits of these platforms are in confining and binding the authors to a specific environment and approach to content creation, which is reflected in the final work

[16][17]. A platform-independent language of marginalia is needed

-----

to support authorial work in general, leaving the choice of approach and tools to the author.

The definition of semantics is not trivial either to remember or reconstruct afterwards, not even in the case of extensive studies conducted on significant authors. For example, literary scholars study authors’ libraries, the personal book collections of authors, e.g. Thomas Moore’s library, which is preserved in the Royal Irish Academy in Dublin. Moore made annotations in certain books, such as texts on Indian History and language while composing his oriental romance Lalla Rookh. He then composed the poem, where he added footnotes citing the books he had read as a way to reconstruct his composition process [18]. Contemporary authors still undertake the same process, even if with digital tools. Physical and digital libraries still contain strictly personal annotations, for which we lack a stable language enabling reliable interpretations based on rules and conventions that challenge the passage of time between note taking and content creation3.

When reading and writing digitally, we have boundless availability of sources and can read much more extensively, continuing the trend that occurred in the transition from manuscript to print. On the other hand, all the activities we do leave traces and can enable the formulation of links to the future. The classical approach to authors’ libraries is a brute force process, but even in a contemporary digital context, we cannot fully reconstruct what goes on in an author’s mind. Software platforms such as Liquid4 and Scrivener5 are striving toward a continuous navigation of the author from reading to writing to reading again. The framing of reading and writing as a continuous space is a significant paradigm shift that is, however, still locked in the personal sphere of the author. To open this space to a sharable understanding, a definition of a language and relative standards are still necessary steps.

Focusing on commonly-available authorial tools, an important distinction must be made between the private work of the author and its relationship with the wider field. On a practical level, this distinction can be found in the classes of intra-document and inter-document technologies.

Since readers undertake reading experiences with varying motivations, we are aware that not all marginalia are produced with the goal of enabling future content creation. In this analysis we consider marginalia that are traces of the evolution of thought processes that may or may not result in the authoring of new work. It is our belief that an encoding of such marginalia could both facilitate the task of the author and consequently enhance the experience of the reader.

3 In marginalia and personal notes, the association between sign, signifier and semantic is often contingent and unstable. 4 http://liquid.info/ 5 https://www.literatureandlatte.com/

Figure 1: On the left, screenshot of HoL. On the right, manuscript of Vernon Lee’s ‘Oke of Okehurst’ (1886), source digitised by Holographic Vernon Lee Project

As Marshall [1], we ground our article on the analysis of examples, in this case from the library of a specific author. Indeed, to shed light on the rationale of annotation in authorial workflow, we need to investigate not only how authors take annotations but how they are reflected in their future work. In this regard, given the systematic and in-depth research on the subject, we consider studies on authors’ libraries a significant source for compiling a first list of the different types of links to the future. In this contribution, we refer to examples identified through the Holographic Vernon Lee Project (HoL)6, focusing on a highly influential and productive author of the early twentieth century, who left an extensive archive of sources, notes, letters and diaries about the production process of her works spanning from scientific treatises to theatre, narrative, politics and travel writings. Furthermore, the Holograph Lee Project provides access to the digitised manuscript corpora of Lee, manually annotated by experts, providing transcriptions, hyperlinks and metadata to assist scholars in the study of her (often heavily intertextual) works7, see Fig. 1.

Figure 2: Left, Photograph of a structural annotation by Vernon Lee in the work of William McDougall. Right, William McDougall, The Group Mind (1920). Source from the Cultural Institute of Palmarino.

In contrast to Marshall [1], we do not attempt a systematisation of traces but focus instead on the underlying semantics of annotations (its rationale in the authorial workflow). The rationale is not reflected in the traces of annotation but in the relationship between traces and future work. In this regard, we take as example the Exercises in Style approach of Mason and Bernstein’s “On

6 http://eman-archives.org/EMAN/exhibits/show/parcours-dans-les-projetseman/holographical-lee-hol 7 There are currently twenty-five of Lee’s handwritten notebooks digitized, alongside 1,000 letters, and a small selection of manuscript works.

-----

Links: Exercises in Style” [2], using the Exercises in Style to test and exemplify the result of our analysis.

In the following section we introduce nine examples to illustrate the most significant uses of links to be. Then, we apply the concepts emerging from the examples to an exercise in style on the fragment employed by Mason and Bernstein [2], from Scott McCloud’s Understanding Comics [19]

“I crossed the street to the convenience store. The rain soaked my boots. I found the last pint of chocolate chip in the freezer. The clerk tried to pick me up. I said, “No thanks." He gave me this creepy look. I went back to the apartment, and finished it all in one hour. Alone at last”

The authors undertook this exercise as if they were preparing to read this text in order to write lecture material for university students.

Example 1. This example is from Lee’s copy of William McDougall’s The Group Mind (1920) which is held at the Collezione Associazione Culturale il Palmerino Firenze, see Figure 2 (left).

Lee marks ‘complimentary’ with a strike through and adds an additional marginal note to highlight a small (and incorrect) spelling mistake. This inter-document annotation is a trace of Lee’s writerly and editorial practice at work concerning the text in its form e.g. its syntax.

“I crossed the street to the convenience store. !e rain soaked my boots. I found the last pint of chocolate chip in the freezer. !e clerk tried to pick me up. I said, “No thanks." He gave me this creepy look. I went back to the apartment, and finished it all in one hour. Alone at last”

Threatening Masculinity

Links: Exercises in Style” [2], using the Exercises in Style to test and exemplify the result of our analysis.

In the following section we introduce nine examples to illustrate the most significant uses of links to be. Then, we apply the

“I crossed the street to the convenience store. !e rain soaked my boots. I found the last pint of chocolate chip in the freezer. !e clerk tried to pick me up. I said, “No thanks." He gave me this creepy look. I went back to the apartment, and finished it all in one hour. Alone at last”

Links: Exercises in Style” [2], using the Exercises in Style to test and exemplify the result of our analysis.

corner tub assistant

In this exercise, we highlighted three US textual expressions that could be unfamiliar to UK students and provided an alternative in the margin, which could be used to increase the readability of the text.

Example 2. In Vernon Lee’s Carnet VII, July - August 1916, Lee quotes the conceptual idea that: ‘The low class of white men who are the instruments of this decay themselves die out. The ‘beachcomber’, the trapper, the convict, boundary-rider, leave few progeny in spite of their strong sexual impulses.’ This concept of the ‘beachcomber’ is still of interest to Lee five years later when she reads Man and His Past (Appendices A1, Figure 3), where you can see from the document a clear note ‘119 Seaboard + “Beachcombers.”’ This conceptual inter-document link connects two texts separated temporally, but which reveal a repetition of interest and thought process, e.g. a focus on concepts.

In this exercise, we noticed expressions that identified masculinity as a threatening concept (having learnt from McCloud’s original that the narrator is female), leading the narrator to escape into loneliness.

Example 3. The page illustrated in Annex1 Figure 4 collates Lee’s thoughts on the writer Henri Barbusse, in particular, his work ‘Le Feu’. She writes: ‘H. Barbusse, is the most honest piece of fiction this war has produced, draws with an immense sympathy the moral & intellectual degradation of battle. X Individualism is lost. Politicians may talk of the devils at stake, but to the soldier their fine phrases are without meaning.

"Tonight he’s in the pink, But but soon he’ll die; and still the war goes on, he don’t know why It is in that spirit that men have gone to their death, dazed & miserable.”

Lee is providing an inter-documentary commentary on the First World War through the figures of speech or what we term the metaphor of Barbusse, exposing that during this period of heightened nationalism, fiction often provides greater truths than the political powers.

“I crossed the street to the convenience store. !e rain soaked my boots. I found the last pint “ : resort of chocolate chip in the freezer. !e clerk tried ” to pick me up. I said, “No thanks." He gave me this creepy look. I went back to the apartment, and finished it all in one hour. Alone at last”

In this exercise, we focused on the locution “the last pint of chocolate chip in the freezer” as symbolic of the narrator’s seeking food as a “last resort” consolation from her current situation.

Example 4. The notebook below (Appendices A1, Figure 5.) collates quotes from British and pan-European newspapers during

This image shows Lee quoting from the Daily News: ‘Oct 2

(from an advertisement of Rev. R.J Campbell’s “her Xty” collapsed) “Where was God when my only boy was bayoneted in the face & fell to bleed to death in agony cries heartbroken mother.” This inter-document trace informs Lee’s writing of The Ballet of the Nations: A Present-Day Morality (1915) and First World War trilogy, Satan the Waster (1920); quotes and expressions infiltrate the thinking, and writing of the author.

-----

“I crossed the street to the convenience store. !e rain soaked my boots. I found the last pint of chocolate chip in the freezer. !e clerk tried to pick me up. I said, “No thanks." He gave me this creepy look. I went back to the apartment, and finished it all in one hour. Alone at last”

In my dark room…

With my thoughts…

“I crossed the street to the convenience store. !e rain soaked my boots. I found the last pint of chocolate chip in the freezer. !e clerk tried to pick me up. I said, “No thanks." He gave me this creepy look. I went back to the apartment, and finished it all in one hour. Alone at last”

Independent!

To explore

In this exercise, we explored how reading can stimulate the creation of further content. We focused on the ambiguous expression at the end of the text and sketched out how it could prompt two possible continuations to the narrative.

Example 5. The two quotes below are taken from Lee’s published works, forty-three years apart. Both consider the process of memory and recall on the process of reading and are tied together by the use of the phrase ‘helter-skelter’. In utilising the same term

but adapting the meaning - Lee is creating an intertextual

resonance for her own reader, with ‘helter-skelter’ being the connective thread. In doing so, the term ‘helter-skelter’ summarizes the content meaning of a document which inspired the use of that particular phrasing.

From Belcaro, 1880: “This volume BELCARO is the first fruit of these attempts at knowing: it is not the Sir-Oracle manual of a professor […] but rather the scholar’s copy book, the fragmentary and somewhat helter-skelter notes of what, in his listenings and questionings, he has been able to understand.

From ‘The Nature of the Writer’, 1923: “Memory is not a helterskelter gathering together, since everything new becomes at once connected by similarity or significance with something old.”

In this exercise, we highlighted expressions that appear to us as relevant exemplars of an attitude to female independence and rejection of assigned gender roles that could repay further exploration.

Example 7. This copy of Brewster’s The Statuette and the Background (Appendices A1, Figure 6) was given to Lee by the author (and her competitor for the affections of Ethyl Smith), which she promptly annotated and gave back to the writer. In her marginalia, Lee critiques Brewster’s theory of aesthetics in a way which does not physically alter or highlight the text itself, but rather she provides a commentary of her contrary argument around the edges of the text. In doing so, Lee responds with almost as many words as are published on the page. She writes: ‘He does not make allowance for the ideal – the conception of a certain pattern of thinking and feeling - being or becoming in certain natures…’ This inter-document marginalia positions Lee as a figure within the aesthetic field, one who is able to interact and respond from a position of authority, e.g. exemplifying an argument against or in support of her own theoretical position.

“I crossed the street to the convenience store. !e rain soaked my boots. I found the last pint of chocolate chip in the freezer. !e clerk tried to pick me up. I said, “No thanks." He gave me this creepy look. I went back to the apartment, and finished it all in one hour. Alone at last”

Run!

Do something!

She hits the

clerk!

“I crossed the street to the convenience store. !e rain soaked my boots. I found the last pint of chocolate chip in the freezer. !e clerk tried to pick me up. I said, “No thanks." He gave me this creepy look. I went back to the apartment, and finished it all in one hour. Alone at last”

Trip to store

Reject clerk

Back home

In this exercise, we summarized the main events in the text, providing a schematic representation of the narration in the margin. This type of notation could enable the reader to quickly glimpse the essential points of the story without needing to read the entire text again.

Example 6. Figure 2 (right) shows a second example taken from Lee’s copy of The Group Mind in which she underlines and marks in the margins areas of particular interest. Lee places emphasis on single phrases, whole passages, and furthermore, underlines her own marginalia. Lee is rather cantankerous as she reads this introduction – asking questions of the author, as well as declaiming ‘And this is a psychologist!’ Lee’s marginalia highlights examples of material that show relevance or that are seen as exemplars of McDougal’s argument.

In this exercise, we underlined terms that suggest the opposite interpretation to the previous example. We argue that the narrator behaves in a passive manner, being soaked by the rain, weakly rejecting the clerk’s advances and displacing her need of attention onto the chocolate ice cream.

Example 8. The page below (Appendices A1, Figure 7) is taken from one of Lee’s notebooks from 1919, and is a written response to James Frazer’s works on anthropology. She writes: ‘I have myself applied this "evolutional" criterion to art, instead of trying to find out how art has had to come about & continue. There is however a sub-criterion allied to utility: viz. competition for attention & energy. Given the limited supply of both, it would tend to continue to go where it’s doing so was life pressuring.’ In this extract, Lee shows that she is able to position Frazer’s document within his academic discipline of anthropology, as well as in the field of aesthetics. Lee shows an acute understanding of both areas of study and is subsequently able to apply it to her own practice in aesthetics., e.g. representing positioning.

-----

“I crossed the street to the convenience store. The rain soaked my boots. I found the last pint of chocolate chip in the freezer. The clerk tried to pick me up. I said, “No thanks." He gave me this creepy look. I went back to the apartment, and finished it all in one hour. Alone at last”

Three stereotypes in five lines: 1 chocolate 2 creepy 3 consolation food

In this exercise, we compared this text to other examples from popular culture texts such as comics and television, evidencing the presence of certain recurring stereotypes, which we listed at the bottom.

Example 9. Lee’s edition of Kulpe’s work (Appendices A1, Figure

was donated to the British Institute in Florence, along with four

hundred and twenty copies of works owned by Lee, most with some form of marginalia. The name ‘Semon’ in the margin refers to biologist and psychologist Richard Semon, for whom Lee wrote an introduction to his posthumous work Mnemic Psychology (1923). Lee measures all theories of mnemonic function and physiological inheritance by his theories (which were very unpopular at the time). Lee became something of a public authority, who found Semon’s work a translator (Bella Duffy) and she provided an introduction to this work. In this inter- and intra document, Lee is showing, and extending her position within the field, whilst simultaneously producing notes with which she writes and publishes a comprehensive introduction to these complex psychological works. These marginalia suggest a branching from or an extension to previous work.

“I crossed the street to the convenience store. The rain soaked my boots. I found the last pint of chocolate chip in the freezer. The clerk tried to pick me up. I said, “No thanks." He gave me this creepy look. I went back to the apartment, and finished it all in one hour. Alone at last”

TODO: Swap gender ROLES

In this exercise, we sketched out an extension to the text by hypothesising a gender reversal, with the narrator as active agent in the attempted seduction of the shop assistant.

The way in which we mark texts while we read is, on the one hand, considered temporary and provisional, and on the other, is seen as a personal practice. Firstly, with the examples from Vernon Lee and secondly, with the exercises in style, we demonstrated that even these personal practices are based on common needs and activities. Highlighting these common needs and activities is a first step towards extending text and enabling a shared interpretation of these forms of paratext [20]. We consider these examples as a core set of common cases to guide the development of a general framework of the use of links to be.

As defined in the introduction, we consider marginalia as links to be because they point to future work. Therefore, for the purpose of systematising a common semantic of links to be, we focus on their functional role in the process of developing new content, rather

than on their appearance, their position in the text or other ontological facets.

For instance, we could consider the scope of links to be as being internal, within the document, or external, either relating a fragment to external works or relating the whole document to external works (external relations of the fragment or of the whole).

To understand how links to be affect future work, we must consider the components of the work, which are its form, its content and its reference field. In this framing, links to be must therefore be a carrier of information relevant to the definition of the form, content and position within the field of the future work.

Therefore, the relevant feature that links to be capture results from the assessment of the reader (future author). Firstly, readers engage with the aesthetic of the work encoding with link-to-be a critique, which can address 1) the appearance of the work, or 2) what the work represents, respectively the “flagranza” and “astanza” of a work [21] or its inscription or documental value

[22]. Then, the reader engages with the actionability of a work: its potential use, e.g. as a point of reference, complement, or difference. As mentioned, these functional factors can concern the different components of a work, see Table 1.

Table 1. Semantics of link-to-be.

Structure Content Field Appearance Text Expression Argument Representation Concept Summary Positioning Actionability Figure of Relevance Connecting speech

In this contribution, we have focused on the semantics of marginalia and annotations, rather than their aesthetics [7], their traces [1] or their positioning [23], which other scholars have examined. This convergence in the motivations for marginalia means that, while authors all annotate in different ways, they do so for the same reasons, which is the starting point for the development of a common language of links to be. If we consider hypertext as a system of thought of the reader, then supporting marginalia means supporting the elaboration of thought to conceive and produce new works.

There are three orders of reflections that can be made.

Firstly, it is worth reflecting on why links to be are still not defined and formalized. We believe one reason is related to the fact that links to be are still seen as personal content not meant for public sharing. Therefore, it is not necessary for us as authors of marginalia to make them futureproof, under the false assumption that we will always be able to interpret and recover their meaning. From personal experience, their interpretation is not a trivial task even for the author of the annotation. Furthermore, without a standard language, we cannot rely on basic tools used for managing contents, such as file search engines.

-----

Secondly, concerning digital remediation of cultural heritage, the adoption of an explicit framework for links to be can enhance the digitization efforts of authorial libraries by producing interoperable and comparable digital sources. One of the immediate consequences would be a paradigm shift, enabling for the first time a standardized approach to marginalia in critical editions, and therefore supporting practices such as “social text critical editions” [10]. For example, the extension of the TEI format (https://tei-c.org/) to include links to be is indeed on the agenda of several research projects, such as Melville’s Marginalia [24].

Thirdly, a formal language of links to be has strong implications for authorial workflows and the design of authoring tools. Beyond historical cultural heritage sources, the provision of tools with native support of links to be is a necessary step for preparing ourselves for the future of the “textual condition” [25]. In this regard, this framework can be used as the kernel of a design theory for a new generation of tools, extending the accessibility and the functionalities of reference management systems to marginalia.

This is a preliminary work that wishes to be a contribution to the effort of the hypertext community toward a definition of a semantics and formal language of links. This first step addressed what could be a core semantics of links to be. Future works should address the grammar of links to be, taking the responsibility to define a visual language of marginalia reflecting the core semantics. Last but not least, it will be necessary to instantiate this language in tools according to the different practices of the many authorial communities, such as creative writers, scholars, students, editors, journalists, legislators, and judges.

This work was partially supported by Reading Europe – Advanced Data Investigation Tool (READ-IT), a project funded by the JPI Cultural Heritage under the European Union Horizon 2020 Research and Innovation programme (grant agreement No 699523).

A special thanks to the Holographical-Lee project, dir. Sophie Geoffroy, Université de la Réunion & Institut des textes et manuscrits modernes, CNRS-ENS, is part of the Projet EMAN, supported by the International Vernon Lee Society (IVLS), Université de Paris; the Vernon Lee Archive, Miller Library, Colby College, Maine (USA); Archives Langweil-Halévy-Noufflard, Paris; Bibliothèque Nationale de France, Manuscrits, Paris; Laboratoire LCF, Université de La Réunion; UFR Lettres et Sciences Humaines (Université de La Réunion); Conseil Régional de La Réunion; Associazione Culturale Palmerino, Florence.

Images reproduced with permission of the University of Liverpool, British Institute of Florence Harold Acton Library, the Holographical-Lee project (CC BY-SA 3.0 FR), the Cultural Insitute of Palmarino and with the kind permission of Geneviève Noufflard.

[1] Marshall, C. C. (1998). Toward an ecology of hypertext annotation. Proceedings of the Ninth ACM Conference on Hypertext and Hypermedia (p. 40).

[2] Mason, S., & Bernstein, M. (2019, September). On Links: Exercises in Style. In Proceedings of the 30th ACM Conference on Hypertext and Social Media (pp. 103-110).

[3] Coleridge, S. T. (1830). On the constitution of the Church and State, according to the idea of each. Hurst, Chance & Company (p. 199).

[4] Queneau, R. (1947). Exercises de Style. Gallimard.

[5] Sherman, W. H. (2009). Used Books: Marking Readers in Renaissance England. University of Pennsylvania Press (p. 26-27).

[6] Piper, A. (2018). Enumerations: Data and Literary Study. University Of Chicago Press (pp. 10-11).

[7] Jackson, H. J. (2001). Marginalia: Readers Writing in Books. Yale University Press (p. 5).

[8] Landow, G. (1987). Relationally encoded links and the rhetoric of hypertext. (p. 335-7).

[9] Bernstein, M. (1998). Patterns of hypertext (p. 21-29).

[10] McGann, J. (2006). From Text to Work: Digital Tools and the Emergence of the

Social Text. Romanticism on the Net, 41–42. http://id.erudit.org/iderudit/013153ar.

[11] Jackson, H. J. (2005). Romantic Readers: The Evidence of Marginalia. Yale

University Press (p. 269).

[12] Webb, A. M., Linder, R., Kerne, A., Lupfer, N., Qu, Y., Poffenberger, B., & Revia,

C. (2013). Promoting reflection and interpretation in education: Curating rich bookmarks as information composition. Proceedings of the 9th ACM Conference on Creativity & Cognition, 53–62.

[13] Ren, D., Brehmer, M., Lee, B., Höllerer, T., & Choe, E. K. (2017, April).

Chartaccent: Annotation for data-driven storytelling. In 2017 IEEE Pacific Visualization Symposium (PacificVis) (pp. 230-239). Ieee.

[14] Bernstein, M., Millard, D. E., & Weal, M. J. (2002). On writing sculptural

Hypertext. Proceedings of the Thirteenth ACM Conference on Hypertext and Hypermedia, 65–66.

[15] Hargood, C., Hunt, V., Weal, M. J., & Millard, D. E. (2016). Patterns of Sculptural

Hypertext in Location Based Narratives. Proceedings of the 27th ACM Conference on Hypertext and Social Media, 61–70.

[16] Kitromili, S., Jordan, J., & Millard, D. E. (2018). How Do Writing Tools Shape

Interactive Stories? In R. Rouse, H. Koenitz, & M. Haahr (Eds.), Interactive Storytelling (Vol. 11318, pp. 514–522). Springer International Publishing.

[17] Green, D., Hargood, C., & Charles, F. (2018). Contemporary Issues in Interactive

Storytelling Authoring Systems. In R. Rouse, H. Koenitz, & M. Haahr (Eds.), Interactive Storytelling (Vol. 11318, pp. 501–513). Springer International Publishing.

[18] Roberts, D. S. (2018). Moore’s Oriental Artifice: Mughal History, Irish

Antiquarianism, and Romance in Lalla Rookh. In S. McCleave & B. G. Caraher (Eds.), Thomas Moore and Romantic Inspiration (pp. 244–259). Routledge.

[19] McCloud, S. (1994). Understanding comics. Harper Perennial (p. 159).

[20] Genette, G. (1997). Paratexts: Thresholds of interpretation (J. E. Lewin, Trans.).

Cambridge University Press.

[21] Brandi, C. (1975). Teoria generale della critica. Einaudi.

[22] Ferraris, M. (2012). Documentality: why it is necessary to leave traces. Fordham

University Press.

[23] Nakakoji, K., Yamamoto, Y., Takada, S., & Reeves, B. N. (2000). Two-dimensional

spatial positioning as a means for reflection in design. Proceedings of the Conference on Designing Interactive Systems Processes, Practices, Methods, and Techniques - DIS ’00, 145–154

[24] Ohge, C., Olsen-Smith, S., Smith, E. B., Brimhall, A., Howley, B., Shanks, L., &

Smith, L. (2018). At the Axis of Reality: Melville’s Marginalia in The Dramatic Works of William Shakespeare. Leviathan, 20(2) (pp. 37–67).

[25] Kirschenbaum, M. (2013). The .txtual Condition: Digital Humanities, Born Digital Archives, and the Future Literary. Digital Humanities Quarterly, 7(1).

-----

A.1 Lee’s annotations

Figure 3: Left, Flypaper showing Vernon Lee’s personal index to Osbert G. S. Crawford, Man and His Past (Oxford: Oxford University Press, 1921) Harold Acton Archive VL 930.1 CRA. (HoL)

Figure 4: digitised page from Lee’s Carnet XXIV (HoL).

Figure 5: Page from Carnet VI - November 1915 (HoL).

Figure 6: H. B. Brewster, The Statuette and the Background (1896), by courtesy of the University of Liverpool Library, SPEC ZANIA E.5

-----

Figure 9: Text.

Figure 10: Content.

Figure 7: digitised page from Lee’s Carnet XXIV, HoL.

Figure 8: Lee’s margin notes in Oswald Kulpe’s Outlines of Psychology. Harold Acton Archive VL 150.1 (HoL)

Figure 9: Text.

Figure 10: Content.

Figure 11: Figure of speech.

Figure 12: Expression.

Figure 13: Expression.

Figure 14: Relevance.

Figure 15: Argument.

Figure 12: Expression.

Figure 13: Expression.

Figure 9: Text.

Figure 14: Relevance.

-----

Figure 16: Positioning.

Figure 17: Connecting works.

-----