Authors: Alessio Antonini (Knowledge Media Institute, The Open University, Milton Keynes, UK; alessio.antonini@open.ac.uk); Sam Brooker (Communications, Arts and Social Sciences, Richmond University, London, UK; sam.brooker@richmond.ac.uk)

This Seed edition’s formatting was converted from the ACM version of record under the supplied ACM publication authorization.

ACM reference format: Alessio Antonini and Sam Brooker. 2020. Mediation as Calibration: A Framework for Evaluating the Author/Reader Relation. In Proceedings of 2020 ACM HyperText (HT’20), July 13–15, 2020, Virtual Event, USA. ACM, NY, NY, USA, 9 pages. https://doi.org/10.1145/3372923.3404789

Abstract

Emerging communication technologies remediate and redefine relations between reader and author, but a comprehensive progressive framework for assessing this dynamic during the process of preparation, transmission, reception, and consumption of media remains elusive. Such a framework is of consequence for hypertext (and first generation electronic literature in particular). Speculative claims for its utility and equally reductive rejections of the reading experience it offers call for a model which assesses the calibration of the reader/author relationship from within the medium itself. This paper presents a first framework for assessing these dynamics both at the stage of authoring and reading. Within this analysis framework we identify eleven remediating factors conceived as scales between opposing tensions, and implement this model with reference to first generation electronic literature.

CCS Concepts: Human-centered computing; Human computer interaction (HCI); Interaction paradigms; Hypertext / hypermedia

Keywords: Author-Reader Interaction, Hypertext, Collaborative Technologies, Media Experience

1 Introduction

This paper focuses on medium as defining the conditions under which authors and readers negotiate their relationship. This recentering of the medium as the object of research is motivated by the proliferation of new solutions, and disintegration of content standards, precipitated by the digital revolution. The stability afforded by the domination of certain historically located media – the book, the canvas, the record – is eroded by the proliferation of digital technologies whose impact remains in many cases underdetermined. Technologies like hypertext reinvigorated practices more often associated with the pre-industrial era – collective reading, collaborative annotation, patronage. Equally the production processes and workflow that underpin the generation, distribution, and consumption of content is insufficiently addressed by existing practices.

We propose a new framework for evaluating the author/reader dual-relation that seeks to fully address the mediation of their relationship by content technologies. Specifically, this new paradigm proposes an alternative view of this power balance -as a “geometry” defined by a set of technology features which in turn restructure a related set of dynamic dimensions.

2 Background

A central challenge for any intermedial, interdisciplinary framework, however, is the aporia of accomodating affordances from different textualities under one epistemic umbrella. Attempts to contrast network textuality with linear print, for example, is a favoured target of early hypertext critics [1,2] and this disaggregation of digital technologies is a project shared for example by John Cayley. Any attempt to universalise the experience of reading inevitably risks appearing essentialist, even reductive of these very different experiences. Models of communication have, however, historically circulated around two agents and the means of communication bridging them: see Michael J. Toolan’s teller/tale/addressee [3]; Claude Shannon’s sender/channel/receiver [4]; Computer Science’s user/software/programmer; literature’s author/text/reader. While each term is deeply historicised (from which this paper has already adopted the author, reader, text approach widely used in the humanities) there is general recognition of a fundamental asymmetry between these two agents. An author inscribes an intended meaning into a particular medium; the reader interprets this material from their own perspective. Reader-response theorist Wolfgang Iser describes this separation of authorial intent and individual interpretation as being two equally valid polarities [5], in keeping with Jakki Spicer’s conclusion that the 20th century had on balance “jettisoned the notion of authorial intention as a primary focus in the interpretation of a text’s meaning” [6].

Definitions of reader and author are in a constant state of flux, a lasting consequence of poststructuralist undoings. Our refocussing on technology (defined as a temporary concretisation of a particular set of technological variables) allows reader and author to be defined from within the context created by this specific configuration. In the interests of expediency we might also reference Derrida in placing these terms sous rature [7], since they are so loaded yet essential that we must simply strike them through and move on, or Marie Laure Ryan’s observation that if we redefine such terms too broadly we will “need a new word for retrieving words from one’s mind to encode meanings, and the difference with reading will remain” [8]. Our goal is not to define the reader, but to define the reader created by this particular configuration, though a fuller exploration of historical reader relations would provide opportunities for future work. The computer as reading agent (as in distant reading) represents a potentially interesting avenue for future exploration, though arguably this represents a supplementary knowledge circuit which still places the human interpreter in the position of reader.

An asymmetrical relationship (as between reader and author) should not be seen as intrinsically antagonistic. As literary theorist Bo Eklund suggests “literary authority, like any other form of authority, exists only when it is recognized” [9]; it is the reader who consents to willing subordination, fitting themselves to the idiosyncratic material limitations of the medium in question. Whether one entity or many, lean-forward or lean-back, we can generally identify the three components that make up the communicative act. As Janet Murray notes, we must “distinguish derivative authorship from the originating authorship of the system itself” [10]. A driver is not a mechanic; a photographer is not a chemist. Absent the most purely collaborative project, in which both agents engage in real time using identical tools, an asymmetry between author and reader will naturally emerge.

Here, as elsewhere, we differentiate between the categories of author/reader/text and specific iterations thereof. Author should not be read as synonymous with writer, which carries an enlightenment valorisation of the individual at odds with the collaborative activities of software development (not to mention filmmaking or audiobook recording). Equally text should not be identified with any specific medium or reader with an individual – interpretative community might be considered equally valid as a referrant. Philosopher and critic Michel Foucault once argued that author should be redefined to be “equivalent of one, or a series, of definite descriptions” [11], which acknowledges the reader’s instinct to ascribe intentionality to a text without defining what or whom that author may be. Readers will inevitably ascribe motivation to choices made in the development of even the most profoundly collaborative project: as philosopher Alexander Nehamas puts it, “no reading can fail to generate an author” [12] -however you understand this term.

This contribution, however, focuses on the mediator which defines the terms of exchange between reader and author. Transmission between author and reader requires a set of “intersubjectively verifiable instructions for meaning- production” [13], which is to say an arrangement of common symbols organized according to rules understood by both parties; put simply, a medium. Professor of Information Paul Duguid rightly cautions us that it is “important to think not idealistically about information, but materially” [14], while poststructuralist critic Roland Barthes stated that it would be “useless to attempt a material separation” between the work and its interpretation [15]. Representation of thought as audiobook, television series, or novel represents a decision which will naturally result in the elision or privileging of certain material. This notion is perhaps better articulated by Oliver Postman’s argument that first and foremost “form excludes content” [16]. This is to say that the medium by which content is constructed and conveyed configures the material affordances of that content. While this argument may seem essentialist, it is perhaps better understood as a pragmatic approach to the potential and limitation of respective media.

Decisions taken at a software level should be considered instrumental in generating a particular experience, since it is these decisions that will determine the relationship the reader will have with the content. This notion -understanding software as an authorial element with whom the reader engages in dialogue -is well established [17-19] but a suitably wide-ranging framework has proven elusive.

2.1 The Study of Creative Work and Reading

One area which remains undefined here is the specifics of the reading act, a task which has eluded literary theory. Some early theorists [20,21] argued that narrative helps us understand our lives; others saw it as an emotional or aesthetic experience [22,23] others agreed on little save that literature could not be evaluated in terms of transmitting information [24,25].

Clearly a definition even of literary reading remains an open question, even before engaging with the contextual question of collective reading, digital reading, episodic reading, and so on. The study of reading is additionally a component in a variety of other fields, including reception studies, author studies, genre studies, computer science, philosophy, semiotics, cognitive science and beyond. Rather than focus on subjective interpretation, we can instead look at reading as phenomenon, an approach to reading which explores the act itself rather than the perceived outcome. Two integrative theories of reading studies present themselves. The first, proposed by Mangen and van der Weel [26], focuses on the preparation, act, and effects of reading, and provides a phenomenological perspective centred on the reader and the relation with the content. This integrative framework structures themes concerning reading into three macro phases with subsequent foci -preparation/text, act/reader, effects of reading/environment. The result is a mapping of research challenges highlighting potential convergences and collaborations which will underpin our ongoing approach.

The READ-IT reading experience ontology (REO) takes a relatedly phenomenological perspective of the reader, providing a model based on the analysis of sources of reading experiences [27]. Consistent with the previous framework, REO structures reading experience in terms of the reading agent, reading activity, and reading resource. Where REO differs is in supporting the description of complex reading experiences and the relations between agent, resources and activities, e.g. how reader’s dispositions influence the interaction with a digital media and a specific genre.

In summary, both approaches emphasise three phases of reading: preparation, activity, and outcome. Preparation includes all activities and factors concerning the creation of text, background of actors and environment in which a reading activity is decided. Activity includes the physical and cognitive actions required to acquire and elaborate a text and the environment and the human and social activities in which reading is embodied. Outcome includes the consequences of reading at personal and social scale, e.g. personal decisions and acquired knowledge and the overall impact on social groups. Present at each stage are the structural, material constraints placed upon both agents by the content technology itself.

2.2 Non-Neutrality of Technology

That a medium conveys and embodies conscious and unconscious values and assumptions, which are spread and replicated through use, is not a new idea. We find it expressed in different forms and from different perspectives in several foundational contributions: Galimberti’s Psyche and Techne [28], for example; the appropriate technology [29]; Floridi’s infraethics [30]. Being phenomenological theories with a keen interest in ethno-methodologies [31], the kernel of Human-Computer Interaction might equally be traced back to the position of non-neutrality, recognising as it does that effects are not general, but specific to the user and their environment. An emblematic example might be the use of critical theories in HCI [32] and, in particular, Feminist HCI [33], which addresses discriminating assumptions concerning the user and the use of technology. The field of adversarial machine learning [34] and studies about bias in data similarly address how bias or manipulation of data reflects and multiplies societal issues through data-driven technologies.

2.3 Non-Neutrality of Technology

Evaluation of this process can be undertaken from two perspectives (see Figure 1): action to response, in which the study focuses on how actors act in their socio-cultural- technological environment and which reaction they cause, e.g. response at an individual scale and reception at a broader scale from the other actors; 2) respose to action, the study of how the current socio-cultural-technological environment foster, mediate, and influence the actions of the different actors.

Figure 1: Action/Response and Response/Action perspective in studying content technologies, concerning the mediation of: Work, Manifestation, Expression and Item (IFLA, 2013).

To illustrate this approach, we might consider the hypertext paradigm, which is an emblematic example of how content technologies can be addressed either from the author and the reader viewpoint, and from one of these two perspectives.

Firstly, we might assess hypertext as an authoring technology. In this regard, hypertext permits authors to control the structural component of the content (from hidden to explicit), and reading flow through the design or selection of a system of intertextual connections between lexia. From an action-to-response perspective, then, the study might emphasise how author choices and use of hypertext influences the reading experience and the type of reading (e.g. skimming or deep reading, pathways between lexia).

Conversely, we might consider hypertext as a reader technology. In this regard, hypertext frees the reading experience from fixed linear structures, providing intuitive representation and control over conceptual interconnections during the process of interaction. From a response-to-action perspective, the reader reacts to the presence of intertextual elements, thereby developing new practices and exercising agency over the use of both text and paratext, shaping her experience in modalities not envisaged by the author.

Note 1. The editor function, it should be noted, can be located within either the reader or author function, dependent on whether it establishes the initial conditions of engagement or subsequent operations under these conditions.

It should be clear that assessment either from an action-response or response-action perspective is not exclusive -it is possible to assess this technology from both directions and achieve compatible if contradictory results. It is as possible for a technology to impose problematic or constraining conditions on both authors and readers, or to valorise both. Balance of power is not equilibrium, and empowerment of one agent does not necessarily require disempowerment of another. A framework of content technology should therefore support an understanding of the connection between design choices and a variety of outcomes, harmonious or otherwise. By way of illustration, we may consider critical research into social media “stalking”, and the relation between authors and readers on web-comics platforms. In the former case we find social media as a content technology creating a condition of conflict between author and reader whilst also acting detrimentally to both [35]; in the latter, as evidenced by Benatti’s work [36], we find a platform acting as enabler for the formation and emergence of new authors and creative content -supported by a fruitful engagement between authors and readers.

Mode of expression in each case carries the potential to act deterministically, mediating the dialogue between author and reader across temporal, cultural, and spatial gaps. From this perspective, we question how content technology configure this connection in terms of managing the input of different actors, their engagement through its output, and how it realises the conversion or translation between the instances of the different actors (throughput), see Figure 2.

Figure 2: Input, Output and Throughput connections mediated by content technology.

The impact of this technology-mediated connection is the principle target of our proposed framework.

3 Framework

This contribution provides a first framework for describing how technology re-mediates the author/reader relation. This framework provides a comprehensive (but not exhaustive) list of re-mediating factors described as scales between two opposite tensions. These scales are then used to analyse three different types of content technologies at the scale of system, media and feature. These indicative use cases consider the impact of the input, output, and throughput of a content technology on the phases on preparation, activity, and outcome, as identified by the presented integrative frameworks.

Figure 3: Analysis Framework of content technologies.

The following represents a limited number of dimensions within which technology operates to mediate the relationship between authors and readers (see Figure 3). As identified above this is not a comprehensive list, but one with a suitable level of interoperability between media:

Structure / Content, enhancing, changing or customising the structure or to improve, specify, deepen the content sense-making

Replicable / Unique, results in a replicable, sharable condition or configures a unique situation

Self-contained / Linked, supports the creation of self-contained contents or supports interconnectivity and interdependency with external sources

Ambiguous / Explicit, permits ambiguity and gaps (to be filled by other actors) or supports and encourages the addressing of ambiguities

Delegation / Control, supports devolving responsibilities and activities toward other parties or exerts control over activities and responsibilities more commonly addressed by other actors

Actionable / Reflective, oriented toward action and change or reflection and analysis

Legible / Invisible, foregrounds the nature of the interaction or hides steps and actors

Structured / Unstructured, provides a structured interaction or facilitates a more fluid configuration

Synchronous / Asynchronous, supports or fosters synchronous or asynchronous interactions between actors

On site / Displaced, supports or fosters interactions between actors sharing the same physical location or otherwise displaced across multiple locations

Mono-modal / Multi-modal, supports or fosters a single-or multiple genre of interaction

Mono-platform / Multi-platform, supports or fosters interactions between actors sharing a same digital location (platforms) or displaced across multiple platforms The unique calibration of these factors is generative of specific author/reader dynamics. Furthermore, as mentioned, a technology can have a variety of effects on both reader and author which could be better understood and analysed in terms of its unique calibration, and resonances between these factors.

4 Pre-assessment & Evaluation

The impact of content technology on the reader-author relation is to create circumstances consistent with (or contrary to) the needs and desires of these actors. Addressing the broad complexity and deepness of technology adoption are goals of several fields of research and therefore out of scope of this work’s scope. Furthermore, a technology has to be assessed within a specific applicative scenario and setting, as a failure or success are to be expressed in relation to a specific goal.

On the other hand, as discussed in section 2.1, the relation between agent and technology concerns at least:

The preparation for reading, how expectations and needs are addressed by a technology and barriers and issues concerning its introduction in a socio-technological context

The activity of reading, interaction between users and technology, the legibility of the system of how it operates, and the output produced and the embodiment in the users’ environment

The post reading, does the use of technology lead to positive or negative effects In this view, as discussed in section 2.3, a specific signature (the design choices) configures a baseline for adoption, user experience and impact concerning:

Input, the response of the user by the technology

Output, the response of the technology by the user

Throughput, translation of user-to-user effects of technology

Furthermore, it is worthwhile to highlight that the success of new technology is rare and the result of long-term processes of refinement and evolution: the precondition of success is the lack of critical errors.

Table 1: Pre-assessment matrix.

	Preparation	Activity	Post
Input
Output
Throughput

Therefore, we propose a pre-assessment matrix for the analysis of inconsistencies in technology profiles that could result in issues or failure of the author / reader interactions (see Table 1).

It is important to highlight, that the instantiation of the framework and the pre-assessment has to be done for each actor. Indeed, inconsistencies and, therefore, friction can be traced back to the misalignment between the design choices of input and output, among the different actors.

5 Case Study: First Generation Electronic Literature

In this section, we apply the framework and assessment matrix to analyse and discuss first generation electronic literature. Through this example, we aim to showcase the benefit of reframing content technologies in general (and first generation electronic literature specifically) by addressing technology as a remediation of the space where authors and readers interact.

The identification of a narrow object of study is intentional. This case study seeks primarily to assess the utility of our model. It does not seek to accommodate all nuances of hypertext as a medium, which is beyond the scope of this paper and will be developed when seeking edge cases or areas where this model does not apply. When demonstrating a model, it is sometimes necessary to ground it in a representative example, rather than fully interrogating edge cases. Demonstrating a model for literature would not seek to seek the most experimental or modernist variations on the form, for example, but the most representative. First generation electronic literature presents a useful object of study, as the forensic critical examination of its perhaps slim corpus provides scholarly familiarity as our unfamiliar model is explored. By way of definition we may turn to Leonardo Flores, who summarises the ELO’s N. Katherine Hayles’ definition of first generation electronic literature as “pre-web, text-heavy, link driven, mostly hypertext, that still operated with many paradigms established in print” [37]. While possible, expanding this definition to accommodate works of the second generation – which Flores considers ongoing – or protohypertext works would be to undermine the advantage of focussing on a form which has such clear material and historical boundaries, though one would anticipate resonances as one moves to such parallel intellectual projects.

5.1 Critics of First Generation Electronic Literature

While they disagree on the criteria or cause, a plurality of scholars consider this a failed paradigm. The reasons are varied [38] but for brevity we use Mangen & van der Weel’s tripartite explanation: a misapplication of the network metaphor as reflective of human cognitive processing; a view of asymmetry as antagonistic not reflected in reading preferences; an ideological approach which ignored the empirical study of reading [39].

While worthwhile, certain assumptions within this tripartite structure must be addressed. Firstly, it implies an authorship community motivated by dreams of commercial or mainstream success. Secondly, it carries an implicit distaste for the modernist digital poetics that were such a motivator, bringing to mind Diane Greco’s fear that the community would fragment into “anxious identification or phobic dissociation” [40] with established theoretical models. The Electronic Literature Organisation’s curated resource of experimental, innovative e-literature is evidence of an audience for this particular type of experience. Web literature, e-books, and games offer additional examples.

While aspects both of premise and explanation can be contested, Mangen & van der Weel’s findings are empirically correct and therefore worth of being taken into account. While this framing misrepresents (to an extent) both the motivations of this particular authorship community and the peripheral hypertext systems which strayed from the dominant paradigm, this phenomenological and materialist approach lends itself to the evaluation of the proposed framework.

Rather than seeing first generation electronic literature as in some way flawed, this section identifies the author/reader relation generated by this paradigm. This in turn provides an opportunity both for a recalibration of this content technology, and for a reorientation of criticism to acknowledge the unique affordances of the technology as is.

This approach emphasises the importance of seeing hypertext not in isolation, but as an iteration of a particular textual configuration generative of an associated reader experience. Web literature, e-Books, and certain video games have reproduced elements of the hypertext paradigm to greater commercial or mainstream success.

5.2 Analysis of First Generation Electronic Literature

First generation electronic literature may be defined as a non-linear narrative work; a set of interconnected narrative segments (lexia) within a logical structure, defined by the author, navigated through decision points controlled by the reader. A first discrimination concerns the type of logical structure: false-forking, latex structure where all branch converge to a single ending, or true-forking, where readers choices result in different endings (tree-based structure). Moreover, the hypertext structure could either be authored work (i.e. designed by the author) or autonomously generated (e.g. generators common in gaming systems).

The choices of false/true forking and control over the final structure have implications for the end experience of the reader. For instance, false-forking hypertext reflects only partially the readers’ choices, providing a misleading sense of control over the story. Pre-designed structure can be tested and evaluated by the author, while a generator could result in unwanted side effects.

First generation electronic literature is not precisely analogous to hypertext. Nevertheless, different systems do have certain features in common, family resemblances that can be used to evaluate them as a genre.

Structure, the narrative structure becomes part of the authorial work. The reader cannot rely on standards and therefore must be aware of the specific conceptual logic implemented in the novel.

Unique, novels offers a personalised experience, based on the reader choices. Furthermore, the reader’s specific configuration is not reflected in the medium. Thus, the experience is replicable in theory, but intimate and unique in practice.

Self-contained, a novel includes text, structure and navigation mechanisms configuring a stand-alone work. HyperText link to external resources, but the logic of narrative requires in general authors to create self-sufficient works.

Note 2. This position is grounded on practice rather than theory.

Ambiguous, Readers cannot anticipate how a novel is different, unique or similar with others, as the authorial work is extended from content to structure and navigation mechanisms. Furthermore, Readers cannot foreseen the relevance of their input and the impact on the narrative flow: they operated in uncertainty.

Control, author’s control is extended to the structure and navigation of the work, while the reader’s control is extended over the experience of the work

Reflective, narrative is reflective and the hypertext novel aims to support reflection on content by placing readers’ in front of decision points, and content variations (resulting from the different possible navigation flows). and scales: of the content and the structure.

Legible, the new interactions provided by the hypertext are visible to the reader through a specific paratext designed for the specific work.

Unstructured, the control of the reader over the navigation flow configures and fosters non canonical interactions. Readers can freely (to an extent) chose the navigation structure but also restart the experience to explore different pathways before reaching a conclusion. Thus, interaction can be any combination of in-depth and broad across the variety of alternatives.

Asynchronous, reading is done at different time.

Displaced, reading can be done in different locations. 11. Multi-modal, a novel can include music and audio tracks, graphics, images and video complementing text.

Mono-platform, it requires a specific software context.

5.3 Discussion

In this section, we address the three arguments raised by Mangen & van der Wheel. Specifically, we address their first two arguments concerning psychology and experience of reading by applying the pre-assessment matrix to the analysis outlined in the previous section. Finally, we will provide a more general comment concerning their third argument about narrative.

5.3.1 Psychology and Cognition. Concerning the first argument about psychology and cognition, two issues emerge from the combination of features (which we might call design choices):

Firstly, the ambiguity, control, and uniqueness of the interaction in the preparatory phase influences the throughput: how the author’s choices about the expected output are converted in the request of input from the reader. Specifically, from the author’s perspective, a hypertext novel is a collection of works presented to the reader as a single work. From the reader’s perspective, the act of choice generates a “unique” experience, though one legislated from within the structural configuration of the work. Such a configuration may be algorithmic (sculptural) or gated (guard fields) or more transparent (linkbase) but still represents an authorial rather than reader decision.

The cognitive issue concerns the mismatch between the reader’s perception and what is being provided by the author. Indeed, the reader’s control is not over the narrative itself, but the version of the narrative she is going to read (e.g., similarly as choosing a different translation). The author still retains full control over the narrative in its multiple versions.

From a psychological perspective, readers do not expect to contribute to the content of reading. While they perceive their input as relevant in shaping the content, this is not the case: the contents are set by the author, while the reader’s decisions concern a version among the available. This choice is not presented as such, and extended until the very end of the novel. The author wishes to provide personalized content (based on readers’ choices), but the result is extending the preparation phase (the choice of book) to the duration of the reading act.

The second issue concerns the combination of structure, ambiguity, and legibility of the interaction in the post-reading phase concerning the management of input provided readers. The reader cannot anticipate the effects of navigating the hypertext structure. This ambiguity contrasts with the manner in which decision points are presented to readers: structured and highlighted, as indicating a degree of importance that may or may not be reflected in the content.

From a cognitive perspective, readers cannot evaluate their choices, therefore resulting in the notorious sense of being lost in a labyrinth. The two-part source and destination is more often subversive or undermining of reader intention than directly reflective, with the effect of their choices often irrelevant, delayed, or contrary to their expectation. This operational context prevents readers from developing a mental model of the interaction (i.e., reflecting the logic of the technology).

From a psychological perspective, we might turn to Gerrig’s definition of narrative reading as “transportation” [41]. In this view, the reader is passively moved around the narrative world. Indeed, the reader also has an active part, but concerning the enactment of characters and situations: the performance giving life to, for instance, the characters. It is worth highlighting that hypertext, on the one hand, extends the readers’ responsibility on content navigation, but, on the other hand, it requires more engagement and commitment. A mismatch between agency and knowledge yields a chaotic, even unmotivated navigation which disempowers the reader.

5.3.2 Reading Experience. These issues serve to frame the impact of first generation electronic literature on the reading experience. In particular, it emphasises the mismatch between a readers’ natural passive disposition toward narrative (though not towards interpretation) and the requirement of choice-making.

Concerning the reading experience, there is at least a third issue related to the impact of the combination of structure, reflective, and unstructured interaction on the output of the reading act. Narrative reading is a reflective activity, during which we indulge in details and situations, acquiring knowledge and experience that does necessarily have a direct consequence for our future activity. On the other hand, access to content in hypertext is strongly mediated by its unique structure (i.e., non-standard or previously unknown by the reader). Furthermore, the control over the interaction with the structure is given to the reader, resulting in unstructured access to content.

In this scenario, the necessary interaction with the narrative structure is prominent and contests the centrality of content. This change of balance between structure and content may configure a radically different experience more similar to spatial exploration. Indeed, the structure of hypertext is central in the authorial work. This centrality results in a tension between the exploration of narrative “architecture” and the narrative.

In conclusion, a work can be experience architecturally without necessarily being seen as a reading act. On the other hand, the balancing between structure and content is a source of ambiguity as these two types of experiences cannot be combined. Thus, to support a positive experience, the work should facilitate the reader in deciding how they wish to address a novel, and be aware of their decision.

5.3.3 Narrative. In our opinion, the third argument concerns the aesthetics of first generation electronic literature, an area which presents significant challenges for a predominantly structural framework. A novel might be seen as a unique aesthetic structure in itself, while first generation electronic literature may be considered a medium imposing its own aesthetic qualities.

In this regard, we should place the focus on the extended capabilities of hypertext (as enhanced text). Indeed, more expressivity must be compensated by extended skills and capabilities, as authors must manage both narrative, architecture, and the unique synthesis of both generated by hypertext.

Concerning the quality and capabilities of hypertext authors, a recurrent argument, e.g., also raised during ACM Hypertext 2019, is the difficulty in evaluating hypertext, e.g., the appropriateness of links, references, navigation mechanisms and what value they provide to the reader. Indeed, we can evaluate the experience of a specific work but still lack a general theory of hypertext aesthetics and critics, which should support authors in placing their works either in the experimental or mainstream field.

In conclusion, hypertext as a broad range of technologies is more exposed than traditional linear works to undermine the reader. The degree of freedom provided by the technology is also hard to control for individual authors. This exposure is a flaw only in relation to the author’s capabilities in managing the complexity of hypertext and, therefore, the result of an unfortunate combination of a too-expressive technology and non-capable users. This flaw can hardly be seen as specific of the paradigm but a structural constraint of the technology more suitable for a team of authors rather than single individuals (e.g., as in the game industry).

6 Conclusions

In Bolter and Grusin’s Remediation the authors describe the double logic of remediation: that our culture “wants both to multiply its media and to erase all traces of mediation” [42]. This paradoxical desire to be both immediate and to celebrate medium as novelty necessitates a model that reflects more soberly on the consequences of the act of mediation. “Hypertext did not result in the revolution in authoring practices, the decentering of authority, or the empowerment of the reader that had been projected” argues Kathleen Fitzpatrick [43]. Too often discussions of first generation electronic literature likewise emphasises its failings, as if the unique affordances and properties of the medium were inadequate for some undefined, abstract ambition to which all media should supposedly aspire. To criticise first generation electronic literature for what it cannot do is to reject a hammer for breaking your coffee mug when used as a spoon. The value of the framework presented on the preceding pages is to identify the unique calibration of a particular medium, in this case hypertext; this is not done in the spirit of denigration or valorisation, but to foreground the unique affordances which might yet be tapped. If first generation electronic literature paradigm is to be rehabilitated for “mainstream consciousness” [44] then we should recognise it for what it is, not what it fails to be, and discuss it in those terms both with authors and readers.

More widely, the preceding model seeks to make visible the intrinsic features of media not in the spirit of celebration, but of reflection. It is the authors’ opinion that a revised paradigm and a specific framework about content technology is necessary for opening a discussion about the impact of technology, addressing wide-ranging issues as the socio-technological position on editorial systems, the impact of micro-advertisement on the publishing workflow, tailored advertising and content aggregation, and so on. These concerns are frequently and widely discussed in isolation, but a comprehensive model accommodating these factors provides an opportunity to see them in a new context.

This approach refocuses evaluation from an inductive to a deductive mode. Rather than extrapolating from specific technological cases toward general conclusions, this model allows general principles to be used in the assessment of specific cases. This approach lends itself to integration into the developmental workflow of new technological solutions, as it is not necessary for the technology to exist before assessing its potential advantages. For instance, in “Making Design Theory” [45], Redström outlines several roles of theories and theoretical frameworks in the pre-use stages of a design process. Indeed, pre-assessment of new solutions allows for reflection and conscious assessment of the developmental direction prior to full-scale development, and of the potential needs in relation to ethics assessment and user testing.

Framework development opportunities are of three types: firstly, whether the binaries identified represent genuine oppositions; secondly, whether these binaries are exhaustive; thirdly, how robustly this model can accommodate new and emerging technologies. The first will be the subject of the next development phase, in which small groups will be asked to map existing technologies to these binaries and assess their applicability. The second is not a claim made by this paper, and it is hoped that the community can assist in the development of other oppositions of consequence to the reader/author relationship – particularly as expressed within hypertext. Exposure to the humanities community is likely to yield additional commentary and criticism, which the authors welcome. The third represents an opportunity, but in foregrounding the dynamics of the reader-author relationship -one as old, perhaps, as communication itself -it is hoped that this model will be of continued utility. A systematic and comparable analysis of emerging technologies will form an additional component of this future work.

Acknowledgments

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).

References

[1] Liestøl, Gunnar. (1994). “Wittgenstein, Genette, and the Reader’s Narrative in Hypertext.” In Hyper/Text/Theory, edited by George Landow, 87-120. Baltimore, MA: Johns Hopkins University Press.

[2] Landow, G. (2006). Hypertext 3.0: Critical Theory and New Media in an Era of Globalization. Baltimore, MD: Johns Hopkins University Press.

[3] Toolan, M. (2001) Narrative: A Critical Linguistic Introduction. London: Routledge.

[4] Shannon, C. (1948) A Mathematical Theory of Communication. Bell System Technical Journal, 27, 379–423.

[5] Iser, W. (1974) The Implied Reader: Patterns of Communication in Prose Fiction from Bunyan to Beckett. Baltimore, MD: Johns Hopkins University Press.

[6] Spicer, J. (2005) The Author Is Dead, Long Live the Author: Autobiography and the Fantasy of the Individual. Criticism, 47, 387-403.

[7] Derrida, J. (1974). Of Grammatology (Baltimore: Johns Hopkins University Press).

[8] Ryan, M-L. (2001). Narrative as Virtual Reality: Immersion and Interactivity in Literature and Electronic Media (Baltimore, MA: Johns Hopkins University Press.

[9] Eklund, B. (2008). Authority and the Social Logic of Recognition: Poetics, Politics and Social Theory. In Donovan et al. (eds), Authority Matters: Rethinking the Theory and Practice of Authorship. New York, NY: Rodopi.

[10] Murray, J. (1997). Hamlet on the Holodeck. Cambridge, MA: MIT University Press.

[11] Foucault, M. (1977). What is an author? In Bouchard, D. (ed) Language, Counter-memory, Practice. Ithaca, NY: Cornell University Press.

[12] Nehamas, A. (1986). What an author is. The Journal of Philosophy 83, 685-91.

[13] Iser, W. (1978). The Act of Reading: A Theory of Aesthetic Response. Baltimore, MD: Johns Hopkins University Press.

[14] Duguid, P. (1996). Material Matters: Aspects of the Past and Futurology of the Book. In Nunberg, G. (Ed) The Future of the Book. Berkley, CA: University of California Press.

[15] Barthes, R. (1979). From Work to Text. In Harari, J.V. (ed) Textual Strategies: Perspectives in Poststructuralist Criticism. New York, NY: Cornell University Press.

[16] Postman, N. (2005). Amusing Ourselves to Death. 3rd ed. New York, NY: Penguin Books.

[17] Carey, J. (1989) Communication as Culture. London: Routledge.

[18] Heibach, C. (2000). The Distributed Author: Creativity in the Age of Computer Networks. Dichtung. Accessed February 29, 2020. http://www.dichtung.de/2000/Heibach/23-Aug

[19] Mazzali-Lurati, S. (2007). Here is the author! Hyperlinks as constitutive rules of hypertextual communication. Semiotica 167, 135-168.

[20] Richards, I. (1924). Principles of Literary Criticism. London: Routledge.

[21] James, H. 1934. The Art of the Novel. New York, NY: Scribners.

[22] Barthes, R. (1980). The Pleasure of the Text. New York, NY: Hill and Wang.

[23] Miall, D. & Don Kuiken. (1994). Beyond Text Theory. Discourse Processes 17, 337-52.

[24] Kivy, P. (1997). Philosophies of Art: An Essay in Differences. Cambridge: Cambridge University Press.

[25] Lamarque, P. & Stein Olsen. (1994). Truth, Fiction and Literature. Oxford: Clarendon Press.

[26] Mangen, A., & Van der Weel, A. (2016). The evolution of reading in the age of digitisation: an integrative framework for reading research. Literacy, 50(3), 116-124.

[27] Antonini, A., Vignale, F., Gravier, G., Ouvry-Vial, B. (2019). The Model of Reading: Modelling principles, Definitions, Schema, Alignments. HAL, https://hal-univ-lemans.archives-ouvertes.fr/hal-02301611.

[28] Galimberti, U. (2002). Psiche e techne: l’uomo nell’età della tecnica. Feltrinelli.

[29] Hazeltine, B. & Bull, C. (1999). Appropriate Technology: Tools, Choices, and Implications. New York: Academic Press. pp. 3, 270.

[30] Floridi, L. (2017). Infraethics–on the Conditions of Possibility of Morality. Philos. Technol. 30, 391–394. https://doi.org/10.1007/s13347-017-0291-1

[31] Dourish, P. (2004). Where the action is: the foundations of embodied interaction. MIT press.

[32] Bardzell, J., Bardzell, S., Blythe, M. (2018). Critical Theory and Interaction Design. The MIT Press.

[33] Bardzell, S. (2010, April). Feminist HCI: taking stock and outlining an agenda for design. In Proceedings of the SIGCHI conference on human factors in computing systems (pp. 1301-1310).

[34] Kurakin, A., Goodfellow, I., & Bengio, S. (2016). Adversarial machine learning at scale. arXiv preprint arXiv:1611.01236.

[35] Antonini, A., Gomez Mejia, G., & Lupi, L. (2019, September). All We Do is” Stalking” Studying New Forms of Reading in Social Networks. In Proceedings of the 30th ACM Conference on hypertext and Social Media (pp. 111-115).

[36] Benatti, F. (2019). Superhero comics and the digital communications circuit: a case study of Strong Female Protagonist. Journal of Graphic Novels and Comics, 10(3), 306-319.

[37] Flores, L. (2019). Third Generation Electronic Literature. Electronic Book Review. https://doi.org/10.7273/axyj-3574

[38] Brooker, S.. 2019. Man proposes, God disposes: Re-assessing Correspondences in hypertext and Anti-Authorist Literary Theory. In Proceedings of the 30th ACM Conference on hypertext and Social Media (HT ’19). Association for Computing Machinery, New York, NY, USA, 39–48. DOI:https://doi.org/10.1145/3342220.3343649

[39] Mangen, A., & Van der Weel, A. (2017). Why don’t we read hypertext novels?. Convergence, 23(2), 166-181.

[40] Greco, Diane. 2009. “Why are we still talking like this?”. In Reading Hypertext, edited by Bernstein, Mark & Diane Greco, 15-18. Watertown, MA: Eastgate Systems Inc.

[41] Gerrig, R. J. (1993). Experiencing narrative worlds: On the psychological activities of reading. Yale University Press.

[42] Bolter, J. & R. Grusin. 2000. Remediation. Cambridge, MA: MIT University Press.

[43] Fitzpatrick, Kathleen. 2014. Planned Obsolescence : Publishing, Technology, and the Future of the Academy. New York, NY: New York University Press.

[44] Pope, James. 2013. “The Way Ahead: The Teaching of Hyper-narrative at Bournemouth University.” In New Writing 10(2): 206-218.

[45] Redström, J. (2017). Making Design Theory. The MIT Press.