There is a singular privilege that accompanies living in this first century of digital computation: the opportunity to discern what it means. The task is nowhere near complete. We know, or think we know, that computation is having a bearing on our capacities to locate pattern, create and replicate pattern, and disseminate pattern. There are even some who believe its advent will ultimately impinge on the trajectories of natural and human history. Welcome to the Noosphere [Teilhard de Chardin 1959][1]. The ambition of this exercise in meaning-making is relatively more modest. We seek to explore how computation can be used to support the human penchant to adorn one's surrounds. More specifically, our purpose here is to explore how Augmented Reality (AR) Objects can be used as constituents for Landscape Architecture. In the aftermath of Ivan Sutherland's foundational work in the late 1960s establishing the field of computer graphics, the prospect that users might seek to integrate computer-generated form with their surrounds became a plausible one, one realized in the decades since via the medium of Augmented, and now Mixed, Reality [ACM 1988]. In the ensuing decades, AR has emerged as an actual or potential support for multiple fields, including education, construction, engineering, computer gaming and firefighting. In this contribution, we present the efforts of The DataScapes Project to explore how AR's capacity to situate and register digital content might be leveraged for artistic purposes: Augmented Reality content, combined with constituents from Data Art, can be leveraged as raw materials for Landscape Architecture.

The first challenge faced by members of the project was the fundamental one of determining how to proceed. We were presented with a blank canvass and had to select the location and method that, so to speak, would constitute our "paint." Choosing a location proved to be the more straightforward of the two decisions. We were looking for a readily accessible space at one of our two universities, a locale that at once presented a sense of grand scale and a sense of enclosure. We were looking for a tract of land that would serve as a frame and a complement for our AR artwork. Based on that criteria, we selected the traffic circle at Brock University in St. Catharines, Ontario Canada. The circle is 140 metres in diameter, is centrally located on campus, and until recently was bounded by willow trees.

Selecting the method and content that would constitute the matter of our creation, by contrast, proved more difficult. To be sure, there were multiple domains of art that anticipated – in part – what we were seeking to do. The fields of Landscape Art, Earthworks, and Digital Art, through their creation of anamorph and trompe l'oeil artworks, for example, were suggestive of what an AR-based Landscape Architecture might accomplish. But the three fields also differed significantly in their practice from what we were aiming to accomplish. To start, while Landscape Art and Earthworks both make location a central constituent of artistic production, artists in both fields sculpt material objects – constituents from a local landscape – to generate artistic form, not digital content [Thompson 2014] [Herrington 2017]. In the domain of Digital Art, artists such as Joe Crossley have sought to conflate the digital with the material by transforming static empty spaces – generally building surfaces – into multi-media surfaces using projection mapping. The content that is disseminated, however, is 2D, not 3D [Crossley 2017].

Landscape Architecture, the primary field to which we sought to contribute, also did not present an obvious theoretical frame of reference, method or aesthetic to assist our efforts. One dimension of the problem here is that practitioners still tend to view digital tools and content as instruments to support the design of physical landscapes. A second problem is that the field's ontology of place is not sufficiently broad to encompass the fusion of digital and physical that we sought to construct. Since the Renaissance, landscape designers and gardeners, historian John Dixon Hunt argues, have typically divided space into three constructs or "natures," with first nature referring to wilderness, second nature to cultivated land, and third nature to landscapes shaped with aesthetic intent [Thompson 2014, 38]. Some designers suggest that Landscape Architects should consider adding a fourth nature to the list, one encompassing reclaimed landscapes and restored habitats, while we in turn wondered if we had stumbled on a fifth: landscapes containing digital annotations, objects or complements.

A third issue centered on the field's self-definition. While Landscape Architecture traces its origins to Landscape Gardening, the profession has maintained a complicated relationship with the arts over the course of the 20th century. Then, many designers, particularly in the wake of the profession's 1966 Declaration of Concern [2], suggested the focus of landscape architects should be on environmental conservation, not the translation of artistic or philosophical trends, be it transcendentalism or post-modernism, into landscape form. Other practitioners have taken the stance that the primary focus of the field should be on design, here understood as landscape construction that is optimized for the needs of a site's visitors and users. The artistic ambitions of the architect are deemed to be a secondary concern, if they are considered at all. On the 50th anniversary of the 1966 statement, however, the Landscape Architecture Foundation published a New Landscape Declaration, one in which architects called for a renewed emphasis on aesthetics [Landscape Architecture Foundation 2017]. One reason for the call was the realization that aesthetic concerns are not a luxury: poor design of buildings and cityscapes – here understood as the mindless prioritization of utility over beauty or meaning – has had a deleterious social impact. Landscape architects have also stressed artistic concerns because of a widespread sense that the field has never been a site of aesthetic innovation, and that it has missed important opportunities, such as the emergence of avant-garde art in the early 20th century [Corner 2017] [Fajardo 2017] [Jellicoe 1995] [Thompson 2014]. While the field has not been oblivious to important trends such as Modernism and Post-Modernism, "there is surprisingly little discussion," Richard Weller writes, "of what contemporary landscape aesthetics are and what they might yet become" [Weller 2017, 10].

Given this lack of definition, our team opted for an exploratory approach to inform the design of our project, one akin in the Digital Humanities to Stephen Ramsay's Screwmeneutics and Kevin Ferguson's Digital Surrealism. In his seminal essay, Ramsay proposes a method of scholarly activity that eschews the conduct of what he refers to as the "search." The "search" in this context is the practice of identifying a canon – the state-of-the field for a given domain of research – and then identifying a gap in that same canon. As scholars, we contribute to our respective fields by conducting research to fill that gap. But there are times, Ramsay writes, when this formulaic approach does not work, particularly when the researcher is attempting something new. How can the "search" "help me find what I'm looking for," he writes, "when (a) I don't know what's here and (b) I don't know what I'm looking for?" [Ramsay 2014, 115]. For Ramsay, the answer to this dilemma is to engage in what he refers to as "browsing." Drawing on the ideas of Roland Barthes, Ramsay proposes that we engage in the construction of the writerly text. In contrast to the readerly text, which involves a passive form of reading, engagement with a writerly text presupposes a process of composition, an immersion of the reader into the flux of existence, and initiation of a process of discovery and systematization: the connection by the reader of the content from the given text with a context, be it another item of content, a person, or a compelling idea. The theory of the writerly text, Barthes writes, "is a practice (that of the writer), not a science, a method, a research ... this theory can produce only theoreticians or practitioners, not specialists (critics, researchers, professors, students)" [Ramsay 2014, 119].

Via this process, Ramsay argues, the humanist defines a path through culture, thereby providing the Humanist – or, if you will, the Screwmeneuticist – with perspective, the platform necessary to support discovery, desire and questioning. One useful first step in defining such a pathway was undertaken by David Ferguson in his recent visual analysis of Disney animations, where he brings his artistic artefact of interest – the film – into relation with the concept of complex. Any artefact of interest, he suggests, can be dissected and its components re-articulated into new combinations. However, while useful, Ferguson's method for applying Screwmeneutics is not appropriate for our purposes for two reasons. First, its aim is ultimately analytic, not generative. It seeks, like all Structuralist methodologies, to locate the intrinsic properties – the hidden archetypal structures – of a given art-form. Its purpose is not scenario exploration. It does not seek to conduct an exercise exploring the potential ways an art-form might exist. It seeks instead to define the way a given form already exists. Second, and by extension, Ferguson's methodology, and structuralist methodologies in general, are ones that collapse the dimension of time. They presuppose objects of analysis that operate in a synchronic fashion [Ferguson 2017] [Merrell 1975]. We, by contrast, were seeking to create Augmented Reality, multi-media landscape complexes that could incorporate the dimension of time, and more specifically would incorporate temporal art forms such as music. Our purpose, in the end, was to generate an artwork that would serve, in Seymour Papert's words, as an object-to-think-with, and in Barthes' words, as a writerly text [Papert 1980]. Given that we were aiming to create something that was new, we sought to construct an artefact that would provide definition, suggest potential, and present questions and connections for future artists and humanists to explore.

For these reasons, we opted to employ a different Screwmeneutic method to conceive and develop our landscape complex, namely Harold Innis' concept of the Oral Tradition. For those unfamiliar with his work, Harold Innis was a media theorist and one of the founders of the Toronto School of Communication. Between 1940 and 1952 he produced a set of works, including Empire and Communications and The Bias of Communication that were dedicated to exploring the physical, formal and cognitive effects of communication media. A political economist by training, Innis turned to the study of communications in 1940 because he perceived, not surprisingly, that the world was falling apart. How was it possible, he asked in the final decade of his life, that the world, after enduring the insanity and carnage of the First World War, should so readily plunge itself into a Second? The answer, he believed, was to be found in the grand sweep of global history. What factors, he wondered, enabled societies to thrive? And what, by contrast, led to their dysfunction and eventual collapse? His lived experience, combined with his studies of the past, convinced him that human collectives functioned much like biological organisms. The vitality, adaptability, indeed the sanity, of institutions, nations and empires were dependent on the environmental circumstances in which they found themselves. Since the inputs and outputs of human-environmental interaction were largely mediated by communication devices, Innis set high store on their importance in global history. The political and cultural periodization of history, Innis believed, could be correlated with the invention or adoption of new communication technologies. So could the pathologies that afflicted cultures. There was a price to be paid for the unthinking use of technology, and in his writings Innis pointed to two: cognitive rigidity and cognitive flux [Innis 1946] [Bonnett 2017]. Technology historically had produced pathological forms of groupthink in which cultures either focused on the wrong thing (due to cognitive rigidity, or bias), or no thing at all (due to cognitive flux). With its judgment impaired by one of the two pathologies, a given culture would lose its ability to discern the opportunities and threats latent in its environment, fail to innovate, fail to compete and as a result fall prey to its competitors [Bonnett 2013, 191–198].

It was a pessimistic construction of history, but it was not fatalist. While most regimes inevitably went the way of Nineveh and Tyre [3], Innis believed it was possible to construct a culture that retained its vitality and resilience via an ethic he referred to as the Oral Tradition. The Oral Tradition, as Innis conceived it, was not a call to forsake writing and adopt pre-literate Greek communication practices that relied on memory and voice. Instead, like Barthes, Innis called on his readers to adopt an active stance toward knowledge, a willingness to alter constructs and formalisms to meet the needs and experience of the present. Such a stance, Innis argued, had enabled the Greeks to make their innovations in theology, philosophy, science, politics and art. The Oral Tradition was also distinguished by its reliance on linguistic and aesthetic formalisms that were characterized by internal complexity and hierarchy. The power of such formalisms lay in their latent potential. They contained multiple constituents that could be arranged and re-arranged at will, enabling artists to explore new aesthetic constructs, and philosophers new intellectual possibilities, through the separation, interpolation and translation of content. In its essence, it was a serio-comic method, one that enabled Greece to shift its conception of nature from one based on myth to one shaped by science [Bonnett 2013, 209–217].[4]

A third feature of the Oral Tradition, one deriving from its commitment to active manipulation of knowledge, was its use of multi-modal forms of expression. The reason Innis named his communication ethic the Oral Tradition in the first place was due to the influence of linguist Edward Sapir, who noted the "formal richness" of ancient communication practices, practices that contained "a latent luxuriance of expression that eclipses anything known to modern civilization" [Innis 1950, 8–9]. Such luxuriance of expression, and with it enhanced expressive potential, could be purchased in the present by creating formalisms that combined written, vocal and visual forms of representation. Such forms in principle presented possibilities for information visualization – which would assist in the location of significant environmental patterns – and information translation – which would enhance viewer understanding of the construct's content [Bonnett 2013, 166–181] [Innis 1946]. Referring to Italian artistic practice in the 15th and 16th centuries, for example, Innis noted Andrea Alciati's invention of the emblem book, a construct in which poetry, "one of the oldest arts, was combined with engraving, one of the newest" [Innis 2015, 101] The rationale for so doing was provided by Francis Bacon, who argued that emblems "reduce intellectual conceptions to sensible images and that which is sensible strikes the memory and is more easily imprinted on it than that which is intellectual'" [Innis 2015, 101] [Bacon 1654].

The commitment to multi-modality in turn presented a fourth characteristic of the Oral Tradition: a commitment to integrate both spatial and temporal forms of representation. Because of its present and historic relationship with the art of poetry, Innis argued, the Oral Tradition "implied a concern with time and religion. 'The artist represents coexistence in space, the poet succession in time' (Lessing at the University of Berlin, 1810)" [Innis 1951, 102]. In his own economic work, Innis would apply this characteristic of the Oral Tradition by interpolating time series of price data with maps to track the emergent patterns of North American economic history, years before the practice become common via Geographic Information Systems [Bonnett 2013]. The Oral Tradition was finally characterized by what in modern parlance can be characterized as an open-source ethic, one in which content and form was routinely lifted from one work and integrated into another. "The great epics," Innis writes, "were probably developed out of lays constantly retold and amplified. Old ballads were replaced by combinations of a number of episodes into a unity of action. The epic was characterized by extreme complexity and unity" [Innis 1950, 72].

In its essence then, the Innis' Oral Tradition is a method for intellectual inquiry and aesthetic innovation that rests on a willingness to explore the possibility spaces afforded by complex, spatio-temporal, multi-modal form, and to reconstitute that form through the integration of content brought in from the outside. To assist our exploration of the possibility space associated with digital Landscape Architecture, we opted to use the methods and modes associated with Data Art. Also referred to as Information Art, Data Art rests on the premise that the world is replete with forms that can be harvested for artistic purposes. It chiefly emerged from scientific efforts to use sight and sound analogues as a method to locate significant patterns and relationships in large data sets. From that effort, multiple intriguing genres of art have emerged, particularly in the field of music, such as DNA music [Gena and Strom 1995], protein music [Dunn and Clark 1999], microbial and meteorological music [Larsen and Gilbert 2013], and even a music of the spheres, music compositions derived from astronomical data [Ballora and Smoot 2013]. With Data Art, we were presented in principle with a method that would enable us to translate a given data set at once into visual and sonic form.

Once we had a mode for our AR landscape, the next, more difficult step was determining its matter, the data that would provide the pattern, the source of serial distinctions for our artworks. The team's decision-making here was determined by the following factors: our decision to produce two works, contingent circumstance, and the intellectual interest of project team members. The contingent circumstance centered on our need to locate sonification software. While Bill Ralph, a mathematician, and Mark Anderson, a computer scientist, respectively possessed the skills required to visualize our data, we had no one with the requisite ability to translate raw data into musical form. After a fairly protracted search (there is not a lot of proprietary or open-source software dedicated to supporting the generation of Data Art) we were able to locate MusicWonk, an application developed by John Dunn, a pioneer in computer music and art since the 1970s. While MusicWonk is purportedly able to able to work with any data set, the software specializes in supporting the creation of Genetic Music from DNA and Protein Data. The application's website provides tutorials and links to multiple libraries of genetic sequences, including the one we used: the U.S. National Institutes of Health GenBank [Algorithmic 2015]. Based on this specialization, we opted to use protein data and to create an artwork called The Five Senses, a composition of five movements which would respectively be constructed from protein data supporting Sight, Smell, Touch, Taste and Hearing in humans.

Our data selection was also prompted by the team's interest in the construct of change known as self-organization or emergent change, a ubiquitous phenomenon in the natural and social sciences studied by the Science of Complexity [Waldrop 1992] [Gell 1994]. It was also prompted by the observation of scholars such as Werner Jaeger and Harold Innis that different domains of human activity – such as science, philosophy and theology – often produce constructs of change that are very similar [Innis 1950, 93]. It was finally prompted by the observation of John Bonnett that a number of parallels could be found between the core concepts of emergent change – such as positive feedback, the governance of formal cause, and the teleological governance of system attractors – and the philosophy of history associated with the Christian Bible. Economists, for example, often characterize positive feedback as the Matthew Effect, given that the process of cumulative change it describes matches the dynamic Jesus describes in the Parable of the Talents, where the rich became richer and poor poorer [Rigney 2010]. Given these parallels, we opted to create a second work simply titled Emergence based on source text from the Bible. Like The Five Senses, we planned a composition featuring five movements, with each piece named after a constituent concept of self-organization, respectively "Emergence," "Differentiation," "Regulation," "Selection" and "Attractor." Each piece, in turn, would rest on data taken from the King James version of the Bible, text that provided narrative or mythic analogues to the concepts of positive feedback, formal cause, and so on.

With the terms of composition settled, the team's next step was to identify the specific data sets we wanted to use for each composition and then begin the process of sonification. The proteins selected for The Five Senses are shown in Figure 1, while the biblical texts selected for Emergence are shown in Figure 2. The process of sonification for The Five Senses began by harvesting the letter data associated with each protein record, such as that shown in Figure 3. In many ways, NIH protein records resemble what any researcher might find in a library catalogue for a book listing, with metadata describing authors, key words, versions and the like. However, NIH records also feature a sequence of letters where one typically would find Library of Congress or equivalent subject headings describing a book's contents. Those letters collectively constitute the protein described in the given record. In Figure 3, the highlighted letter sequence refers to the amino acids making Medium Wave sensitive opsin 1 (Homo sapiens), a protein contained in green cone photopigment that we used to generate the work Sight. These letter sequences provided the patterned sequences that would be translated into notes for Sight. Similar letter sequences generated the notation for the four remaining movements of The Five Senses, while the biblical passages provided the initial basis for Emergence.

The process of sonification for both compositions was initiated in MusicWonk first by converting every letter from every dataset into a counterpart number, with A equaling 1, B equaling 2, Z equaling 26, and so on. Once an entire letter set from a given dataset had been translated, the software then aligned the numeric series with a specified music scale, producing a raw music string, as shown in Figure 4. From that point, the project's two music composers – Erin MacAfee and Amy Legault – had two methods of composition and two modes of operating with their source data open to them. The first method of composition – selected by neither – was to use the algorithmic method of music composition supported by MusicWonk. There, the raw music string is transformed by feeding it through a set of components that control features such as tempo, chords, instrumentation and the like. The process, shown in Figure 5, is akin to creating an algorithm using a visual programming language, or creating an electronic circuit.

The second method – selected by both – required the two to export the raw music string to a *.midi file, and then open the exported file in Finale, a music composition software package. The appeal of Finale for Legault and MacAfee is that it presented methods for music notation that were familiar to them, most notably by featuring an interface and tools that supported viewing and direct manipulation of music string notation. With Finale, the two were able to alter note duration, inscribe chords of their own devising, and repair aberrations in pitch.

Finale also enabled each composer to determine how she wanted to use the musical sequences inherited from her source data. MacAfee's approach in The Five Senses was minimalist. While she was willing to alter select attributes in her file, such as rhythm, instrumentation and the duration of individual notes, she was not inclined to alter the tonal sequence inherited from her datasets. Legault's approach, by contrast, was more interventionist. Instead of viewing her inherited notation series as a fixed object, she opted to view her raw music strings as libraries from which she could splice identified musical components. The appropriated sections were – much like the lays in Innis' Oral Tradition – placed in new sequences and were integrated into established musical genres selected by Legault, such as the fugue.

While Legault and MacAfee were generating their respective sonifications, team members Bill Ralph, a mathematician, and Mark Anderson, a computer scientist, were working on their visual correlates. Ralph is also an algorithmic artist, one who describes his work as "an attempt to make a visual connection with the enormous complexity and unity that lies within the rich mathematical objects that inspire my images" [Ralph 2018]. He is particularly interested in mathematical objects that can generate chaotic, dynamic systems, and has used them to generate dynamic and static works distinguished by their complex topologies and colour. From the standpoint of this project, Ralph's approach was ideal because it often relied on source data to drive the generation of a given work. For The Five Senses, he used proprietary algorithms driven by two inputs. The first source of pattern was the set of distinctions obtained via a sequential progression through the data. The second source was the set of relationships detected by the algorithm between different, non-proximate strings contained in the data. Both were leveraged to provide a visual focus to our landscape composition. For Emergence, Mark Anderson used a different approach. Instead of deriving his images from the data, he selected Screen Vector Graphic images that related conceptually with concepts such as Differentiation and Attractor, and then used project data and the application NodeBox to animate each image and prompt transformations in colour.

With the sonification and visualization of our two works complete, two further tasks remained, the first being the construction and overlay of a digital set to display The Five Senses and Emergence. Given that we conceived our two works as Landscape Architecture, and that the Brock Traffic Circle – our selected first venue – was 140 metres in diameter, the parameters of our project mandated that the components of our digital set be relatively large in size for the visualizations to be observable and for them in turn to serve as artistic complements to the surrounding landscape. That need was further compounded by the presence of the physical statue, known as She-Wolf, situated in the midst of the circle, and the scaffolding and the four 4.6 metre-high QR code signs we installed to activate and situate our digital works. All three objects are shown in Figure 6, and all three required effacing by the digital set to prevent their disruption of the impact of our AR Landscape works. Accordingly, the set shown in Figure 7 was designed by John Bonnett using SketchUp to display The Five Senses and Emergence. The set is bounded by eight 9-metre-high monoliths, with the front of each monolith displaying a graphic representing the title of the movement currently under display. Figure 8, for example, shows the graphics for the movement Touch on the left and Differentiation on the right. The centre of the set is composed of two components. The cylinder at the base is used to cover the statue, scaffold and QR code signs mentioned above. Floating above it is a 9 metre object used to display the given movement's visualization. For The Five Senses, a cube was used to display the artwork, while a sphere was used for Emergence.

Our final task was the creation of an Android app to support user viewing of the two works via a tablet or phone. Our requirements for the DataScapes app were straightforward. We wanted to simultaneously see and hear the respective visualizations and sonifications associated with each movement. We further wanted multiple users to the site to experience the performance in sync. If one viewer was viewing and hearing the movement Taste, we wanted their counterparts to see and hear the same thing at the same time. We also wanted user entry into the experience to be simple, involving little more than user activation of the app, and direction of his

or her device camera at the QR signs situated in the middle of the traffic circle. The app only worked at a distance of 6 metres or less from the QR code signs, to prevent unsafe viewing of the exhibit by, for example, drivers circumnavigating the circle. The DataScapes application was developed using Vuforia, an SDK (Software Development Kit) designed to facilitate the development of AR applications on mobile devices. Leveraging 3D objects, planar surfaces and patterns situated in the environment, Vuforia is able to determine the user's position and camera orientation relative to objects such as the project's QR Code signs, and in turn to situate and register computer-generated 3D objects. These capabilities enabled team programmers Joe Bolton and Mark Anderson to fulfill another requirement for the DataScapes app: the capacity to move around and within the artwork. Vuforia's capabilities, combined with GPS and gyroscopic data, enabled the DataScapes app to track the movement of viewers, and in response to change the orientation and rendering of displayed objects in real-time, to make our digital objects seem as if they were integrated and locked into the natural landscape.

The final results of our efforts are shown in Figure 9 and Figure 10, along with a video showing the operation of the app in Figure 11 (For an earlier treatment of The DataScapes Project, see Bonnett et al. 2018). The first version of the DataScapes app in our estimation was a partial success, both technically, and with respect to audience reception. With respect to the app's functionality, our initial trials of the two works were hindered by two difficulties. The first and perhaps most disappointing setback was the inadequate processing power of the two Asus 12" tablets we had on hand for viewing the display. The tablets did not have the capacity to simultaneously render and register our set while simultaneously displaying our data visualizations in dynamic form. Despite our best efforts to find work-around solutions on site, we had to settle for the necessary compromise of presenting Bill Ralph's and Mark Anderson's respective visualizations in static form. A second difficulty centered on the stability of the set. While the first version of the app adequately anchored our set to the Brock traffic circle, several of its constituents, most notably its monoliths, tended to vibrate or jitter in a way that was at once distracting and unpleasant. These deficiencies have been partially remedied in the second and third versions of our app. Our intention in these two iterations was to develop portable versions that would display The Five Senses and Emergence indoors, while remedying some of the performance failures identified in the first. Version two, shown in the animation featured in Figure 12, was designed to support Mark Anderson for an Edge Hill University lecture dedicated to the Art of Computing, and presents the project as a wall display. Version three, shown in Figure 13, transformed the orientation and scale of the project yet again, by adapting it for display as a table-top installation. In both, we were able to better stabilize the display of our set. And in both, we were able to add some dynamism to the display, albeit by introducing movement into the set rather than our visualizations. Both iterations required changes to the application, most notably by changing the way Vuforia searched for targets and changing the orientation of the 3D content it displayed in relation to those targets. In addition, we also removed the initial application's use of GPS tracking code, as that functionality was specifically designed to support the geo-location of 3D content on the Brock traffic circle, and further designed to hinder the display of content to users not on campus. Finally, we reduced the size of QR code displays used with versions 2 and 3 of the DataScapes app and designed each version to generate displays proportional in size to the QR code target. Via this step, the user can print a code on a piece of paper and generate a display proportional in size to the underlying table.

As for audience assessment, we report here anecdotal evidence gathered during our first trial. Our rationale for this approach as opposed to more extensive and expensive methods used in Human Computer Interaction (HCI) and elsewhere is due, in part, to concerns regarding the reliability of user assessment methods pioneered in cognitive psychology and HCI. These methods have been drawn into question due to, on a general level, the replicability crisis in the social sciences and, more specifically, widespread criticisms of the two fields pointing to errors in statistical method, small sample sizes and large confidence intervals [Lundh 2019] [Greenberg and Thimbleby 1992] [Cairns 2007] [Gigerenzer 2004] [Cohen 1994]. Putting the matter simply, user assessment via this route seemed to impose a good deal of effort for results that did not necessarily convince.

For the type of exploratory exercise that we were pursuing, it seemed more useful to take a craft-based approach to user assessment. In fields such as web design, argues Gerd Waloszek, a former interface designer with SAP, the metric for success for a given design is not its authentication via an HCI analysis of design consumers. The real metric rests on whether other design producers copy the design, making it a convention [Waloszek 2003]. In a different design context, Elizabeth Eisenstein and Hellmut Lehmann-Haupt note a similar process in the history of the print book. Through experimentation and appropriation of designs from competitors, printers such as Peter Schoeffer collectively generated the conventions of the modern book, such as tables of contents, title pages, running heads and footnotes [Eisenstein 1979] [Lehmann 1950]. A craft-based approach, then, suggests that the "success" of our effort will depend on whether other artists, landscape architects and digital humanists choose to emulate our approach to landscape design, because they like it, and because they believe it will induce a favorable reaction from their respective audiences. The best we can do is report what we learned from our viewers and suggest the questions and implications that stem from those observations. We will be "successful" if colleagues choose to mimic those efforts and address the questions we have raised.

Based on that framework, we noted two classes of respondents – general viewers and artists – and important questions that arose from each. General viewers for the most part found our various pieces engaging, reporting an intrinsic fascination with the fact that data, particularly protein data, could be translated into art forms, particularly music. That feedback raised an important question for us, however. Is Data Art intrinsically compelling on its own? Is it capable of generating an intellectual or affective response from viewers without annotation, or must it be presented with an accompanying spoken or written context to maintain viewer interest? This is, of course, not a new question in the history of art and Landscape Architecture, where practitioners continue to argue whether art should attempt to communicate a concept or message – such as the need for environmental sustainability – or instead focus on generating an emotional response from viewers [Thompson 2014]. A related question is whether artists and Landscape Architects can generate the sensory "languages," be it visual, sonic, olfactory or something else, required to successfully communicate the artist's message to an audience without textual assistance.

The response of artists was, perhaps not surprisingly, more searching and more critical. While most viewers found the project's content and approach very interesting, some took exception to our temporary appropriation of the Brock traffic circle (we used it for one week), particularly since it already contained a physical statue that was surrounded by the project's scaffolding and QR codes. One artist deemed the metal framing and the statue's digital effacement problematic. Another deemed it a violation. This feedback raised fascinating ethical and political questions for us. To be sure, the infrastructure we constructed around the physical She-Wolf statute did hinder proper viewing of the statue. Whether it constituted a violation depends on your point of view. Until artists raised the issue with us, the construction of the scaffolding and QR codes prompted no concern from project participants, nor did it for the university officials who granted us permission to use the space. We meant no disrespect. It was not our intention to make a comment on the statue, nor to damage it. But some artists did see our temporary occupation of the space as an unacceptable abridgement on the statue's integrity and the sculptor Ilan Averbuch's rights.

One response to these concerns would be to note that the scaffolding and QR codes will not be a permanent component of AR Land Architecture, or more generally, AR-based artforms in future. We used them because current technology requires their use to situate and register content. In future, artists will be able to use RTK GPS (Real-time Kinematic Geographic Positioning System) data to position content with centimetre and even millimetre level accuracy. That expedient will resolve the issue of the art form's physical intrusion into a given space, but not its digital annotation. Is it ethical for one artist to use digital methods to efface, comment upon, add to, mock, or alter the context of a physical artwork created by another, even if the digital annotation is not visible to the naked eye? To that ethical question we can add a political one: who gets to add that annotation? The complexity and sensitivity of this question was anticipated in 2017 in New York when the statue Fearless Girl (sculpted by Kristen Visbal) was placed in proximity to the statue Charging Bull, a work situated on Broadway and long associated with Wall Street and the city's financial district. To many, Fearless Girl presented an important message of female empowerment. Charging Bull's sculptor Arturo Di Modica, however, was outraged, arguing that Fearless Girl's proximity and positioning transformed the meaning of his statue, from one symbolizing prosperity and strength, to that of villain [Dobnik 2017] [Mettler 2017]. The controversy raised multiple questions, ones that we now face with our project. When an artist makes an intrusion into a space, does that space become sacrosanct? Artists have long made interventions into spaces that they deem to be playful, even provocative. Are those intrusions to be exempted from three-dimensional annotations that some will label as legitimate super-positions or rejoinders, others as graffiti, and others as outrageous violations? Who gets to decide the issue? These issues will become increasingly central as AR becomes a ubiquitous communication medium.

In addition to the questions posed above, what other observations and questions might we offer to conclude this exercise in Screwmeneutics via the Oral Tradition? In our view, the best way to finish would be to explicitly address the questions that constitute the heart of this issue of DHQ.