In the past ten years, there have been notes like the above on the imprint pages within the major editions of the London Plan, pointing readers to the online version of each document—links to their digital selves that, contrary to the statements, are now no longer available at the web addresses indicated. An imprint page is the page of a book containing details of the book’s publication, such as author, publisher, copyright and contact info, etc. These links do not exist anymore. Attempts to visit them are likely to result in either, 1) if lucky, a redirection to what may be the closest matched webpage under a different address link, or 2) if unlucky, to an error page advising that “the requested page could not be found.” In the case of searching for documents produced during the draft London Plan consultation from 2018-19, it leads to an “unpublished” page. These links are outdated as soon as they have been committed to print on paper.

For all the efforts to produce documents for the corporate memory as discussed in the last chapter, and despite the GLA’s organisational imperative to preserve the public record and corporate identity for future access, the “not found” status of documents throws up roadblocks in accessing the memory banks. The PDF, with its media affordance for wide circulation, has opened up the avenues by which the GLA and the Mayor can reach their audience and universally share their work. But this universality has also given rise to a complicated ‘document flow’ situation—the process of where and how documents show up, if they do at all—because when documents can be anywhere and everywhere, sometimes they are found to be nowhere.

Lisa Gitelman has defined a document in distinct terms of having a know-show function, but her theory is based on an assumed appearance. What happens when a document doesn’t show? What becomes of its public status? How does one find a way in and around the London Plan that appears to be on the constant move and that may face an uncertain, unpublished future? This is the situation that confronts online readers (present and future) of the London Plan, challenged to carve a clear path through the London Plans’ digital traces and plethora of lists and links of associated documents. Navigating the plan entails contending with the histories of the GLA’s (often transitory and fugitive) paperwork, the dis/appearance of text, the temporality of documents, the discontinuity of administrative practices, and the instability of databases and public statuses.

One consideration is the condition of text in today’s accelerated pace of information movement. The 1990s saw the rise of the Web, which, in writer and artist Kenneth Goldsmith’s estimation, “functions both as a site for reading and writing.” ¹ ⊕ Goldsmith, Kenneth (2011). Uncreative Writing: Managing Language in the Digital Age. Excerpt from Chapter 1 Revenge of the Text. Our new engagement with digital words brought in what he calls a “morass of information,” and new opportunities and complications surfaced for personal computer users and government workers alike in relation to “the enormity of the Internet” and the task of managing language and information in the digital age. ² For Goldsmith, such vastness and pervasiveness of the computer and the Internet altered not only work patterns of how knowledge is shared and exchanged, but of how we individually and differently make our way through “this thicket of information”—how we manage it, parse it, organize and distribute it. ³ It irrevocably changed how information is written, read and processed, by both human and machine. Text became hyper, going above and beyond what ordinary text can do. Writing became “non-sequential.” ⁴

Hypertext has a long history that predates the internet and the electric Information Age. A term first used by Ted Nelson in 1965 to mean text which contains links to other texts, it has literary roots in the late 19th and early 20th century underpinning its bibliographic format and concept of ‘links of association’. ⁵ In the Secret History of Hypertext, Alex Wright discusses the well-known influence of inventor and engineer Vannevar Bush’s hypothetical machine called the Memex (1945) on the development of the Internet. The Memex, short for memory extension, was a device conceived to allow users to comb through a large set of documents stored on microfilm, connected via a network of “links” and “associative trails.” Before the Memex, however, Wright notes, there was Belgian bibliographer Paul Otlet’s plan for a global network of “electric telescopes” that would more presciently anticipate the arrival of the World Wide Web permitting anyone in the world access to a vast library of books, articles, photographs, audio recordings, and films. ⁶ Along with Henri La Fontaine, Otlet had developed the Universal Bibliography project aiming to catalogue all of the world’s published information. It used a system of classification called the Universal Decimal Classification they had adapted from the Dewey Decimal System.

According to Steven Johnson, there was also Charles Dickens and Great Expectations. ⁷ Writing in a special 1998 issue of Columbia University’s online journal 21stC on the Future of the Written Word, Johnson argues Dickens set a new storytelling way of reading relationships between normally disparate characters and divided social orders: ⁸

Their “links of association” find equivalent purpose in high-tech descendants. The link, Johnson attributes, supplies a sense of coherence to the “teeming complexity” and “the imaginative crisis [that] comes from having too much information at our fingertips”. ⁹ Through linking, and in praise of Bush’s Memex machine for its potential to provide such continuities, he saw the promise of the Web as “a way of seeing new relationships, connecting things that might have otherwise been kept separate.” ¹⁰ Johnson saw the link as a profound linguistic achievement, “the first significant new form of punctuation to emerge in centuries.” ¹¹ With one click, hyperlinks have the potential to bring together scattered words/worlds across the infosphere, for the fragmented to find each other anew or be reunited again.

Bush and Otlet’s different yet overlapping ideas about information storage and retrieval, centred on a notion of ‘searchable’ data housed in a web of documents connected via a sophisticated linking system, would eventually lead to the creation of hypertext markup language, or HTML, the linguistic, computational system that undergirds the way we see, transmit and interact with information online today. Their work precipitated the future of documentation and recorded information, laying the conceptual foundation for current mechanisms of collecting, annotating, and sharing documents. This technological development is especially important to understanding the London Plan as a public document. Where the last chapter considered the London Plan as portable text, in equal measure, it is also significant to recognise the London Plan’s performance as hypertext. The ways in which the plan works as hyperlinks and associative trails determine its in/accessibility and in/visibility.

The GLA makes as many of their documents publicly available as possible on the GLA website in continuous efforts to widen participation in and understanding of the work of City Hall. From Google Search results indicative of the content composition and distribution of london.gov.uk, of the total content Google indexed, around 25-30% is PDF documents, and of these, 30% relate to the London Plan. A quick Google search (2021) limited to ‘site:london.gov.uk’ yields about 376,000 results, the number of webpages available with that URL. A further search constrained to only ‘filetype:PDF’ show about 81,400 to 102,000 results. Adding “London Plan” (enclosed in quotation marks) to refine the search criteria to only PDFs containing these precise keywords, narrows it down to 27,900 results. Dynamic and constantly changing, these numbers represent what Google Search can index at the moment of searching, not the accurate quantity published and available on the GLA’s website. What this suggests is that a third of the website contains PDFs, and a third of those PDFs have some info about the London Plan. The PDF is greatly relied upon for information dispersal.

Delivered through thousands of hyperlinks, each trail leading ultimately to a PDF, readers can find info on what past plans did, what the new one will do, and how the GLA went and will go about doing it. Intermediary pages act as signposts—navigational midpoints—to the document destination. At the bottom of almost every london.gov.uk page is a section called, “useful links,” which does the heavy lifting of connecting content.

The London Plan is typically presented as a single document, in reality, it is only the tip of a very large iceberg of innumerable hyperlinked documents. The 2017-2021 Examination in Public Library (EIP), for example, contains: the draft plan and supporting evidence base; links to all representations on the plan; published changes to the draft; EIP panel notes and other documents issued by the inspector panel; written statements on the matters to be discussed; and other documents requested or accepted by the panel. More than a ‘thicket’ or ‘morass’ of information, this central online repository totals a staggering 3,654 documents. The EIP Library is one long webpage listing a series of paths to follow, signposting to other texts and other paths. It lists hundreds upon hundreds of documents (via hyperlinks to more PDFs or to more webpages) that must be taken into account for a full picture review of the London Plan in its totality. Working in an indexical, listicle way, the EIP Library is the GLA’s solution to the administrative (rather than imaginative) crisis of having too much information that needs to be at Londoners’ fingertips. They use the webpage as a kind of memory machine, like the Memex, relying on it to provide document continuity, an extended way to record and to remember. ‘Memex’ is short for memory extension. But, much like how memory functions, it is shaky at best. These graphic illustrations form part of my design work analysing the London Plan and other related publications. I used software tools, Adobe Illustrator and inDesign, to visually analyse and map the documents relationship to each other. They revealed to me the breadth of the plan, the enormity of links involved, and the ways in which the document grows as the plan develops from draft to final.

The density of London Plan 2021 documents is overwhelming. Making a way through the density is overwhelming. Per the GLA website, the draft London Plan is a material consideration in planning decisions and gains more weight as it moves through the process to final publication and adoption. ¹² But moving with the document through the process also increases in hardship, weighing down and requiring a fair amount of nimble manoeuvring. Over time, the London Plan webpage and the EIP Library grows in length alongside the progress of the plan’s development—information increases in density, more text, links and webpages are added, others get taken off-line. The webpage is designed as a page listing and reflects progress of the London Plan 2021’s development. It looks like a printed page put online, not very interactive. The page organisation is a series of paths to follow. Images are minimal, used for illustrative purpose. Body text is descriptive, for information purpose, signposting and hyperlinking to other texts. Information is presented chronologically with no clear hierarchy of pages or documents linked. Over time, the webpage grows in length, information increases in density, and more links are added. Information gaps frequently arise out of ongoing changes to the status of documents, presenting unique challenges to those trying to follow the winding, shortened, or sometimes altogether cut-off paths of London Plan documentation. It reflects a complicated documentation situation, which started with the first draft London Plan, that has become increasingly difficult to manage.

Since 2002, there has been a section dedicated to the London Plan on the GLA’s london.gov.uk website, and some form of a dedicated Examination in Public Library page. Where the London Plan sits on the website has shifted across mayorships, located first directly under ‘Strategies’ during Livingstone’s time, then moved to the ‘Priorities’ section with Johnson then ‘What We Do’ with Khan for a while, usually three mouse clicks away from the homepage. As of this writing (November 2022), the London Plan recently moved again, now under “Programmes and Strategies,” returning almost full circle to Livingstone. Reaching the actual documents and information involves several, and sometimes many, more clicks. With each new plan, more steps in between have been added, more hyperlinks aggregated, more information to take into account. A sitemap sample capture of london.gov.uk from 2001–2008 illustrates how much the GLA website has grown overall in complexity from year to year, and how progressively further away from the homepage London Plan documents get. ⊕ A sitemap is a structured list of the pages of a website, typically itemised by hierarchy. Human-visible and computer-indexable, it is used by human and machine to navigate a site’s content. Radial graphs, produced using the Wayback Machine of the Internet Archive, comparatively show the tangents that have multiplied annually is exponential. The Internet Archive is an American digital library that provides free public access (“universal access to all knowledge”) to collections of digitized materials, including websites; the Wayback Machine is an Internet Archive initiative that contains 20+ years of web history and content captured by internet users and algorithms.These screenshots form part of my design methods of analysing the London Plan and other related publications. I used the digital tool, the Internet Archive, to screen capture historical versions of the GLA’s website in order to visually analyse and map the plan’s online presence across the years. They revealed to me the published state of the plan at different times, the enormity of links involved, and the ways in which documents move on the web as old London Plans accede to new ones.

The exponential growth of hyperlinked documents has consequence for document management—an issue that was obliquely identified during the time of the second London Plan, first off-handily acknowledged in Boris Johnson’s opening speech for the EIP for his Draft Replacement Plan 2010, in which he remarked that “our administrative machinery has creaked occasionally under the load.” ¹³ Then, after the publication of Johnson’s London Plan 2011, the Mayor published a Supplementary Planning Guidance, London Planning Statement, to clarify his planning functions in light of major changes to the planning system in 2012 whereby the new National Planning Policy Framework (NPPF) had revoked the Government Office for London Circular 1/2008 which provided advice and guidance on the arrangements for strategic planning in London. ¹⁴

The skinnier NPPF replaced over one thousand pages of national policy guidance dispersed across 44 documents, into a consolidated slim statement of 57 pages. In response, Johnson’s SPG document, to be read alongside the London Plan itself, filled the information gap left by the streamlining to provide guidance and context on how the plan and its policies will be implemented. Indirectly, it served as a helpful wayfinding tool around the London Plan’s own avalanche of documents, offering a roadmap for how to navigate the planning system’s rearranged paths. Essentially, the London Planning Statement is a document about the plan documents—the different types and their formal status. As a document of documents, it hints at the London Plan’s complexity.

In a move intended to better facilitate document management, the London Planning Statement guide describes the Mayor’s intention “to use a web-based approach, making it easier to keep documents up to date, allowing him to update particular sections without having to withdraw and re-write whole documents (these would effectively become “folders” of linked documents, any one of which could be updated without affecting the others).” The new system marked an attempt at making better use of the web’s network capabilities. In practice, however, the use of links has been complicated and very confusing. The problem is that there is no clear structure or discernible archiving system that make it easy to track where these files go, at least not easily discernible to readers and policy browsers. Links often expire after a certain period; documents disperse and disappear all the time without notice. They sometimes even go ‘unpublished’, only leaving behind an error note that taunts the document was here but now has gone elsewhere. Like Andrew Murphie’s concept of ghosted publics—those that are simultaneously in and not in the public sphere—they become ghosted documents, haunting with their near but intangible presence.

The GLA’s Records Management Policy was meant to manage such a diasporic document landscape. It contains a Historical Archiving Policy on the archiving of records that are of historical value, which details the procedure for transferring records for permanent preservation to London Metropolitan Archives (LMA), the archive repository for many London-wide organisations. Part of an agreement signed in 2006, the records of the GLA held by the LMA include key strategy documents relating to core GLA business and documents of historical interest (e.g. The London Plan); however, according to an LMA archivist, “the majority of records are uncatalogued” and “many of the records that are uncatalogued (mainly minutes and related papers) are actually available on the GLA website” as PDF documents. ¹⁵

The advice given by the LMA archivist is that the GLA website is still the best bet to find a GLA document. The website is, effectively, a living archive. Albeit an incomplete one. This “ephemeral nature” of documents compromises Johnson’s promise of a web-based approach and exposes contradictions of portable text (PDF) and hypertext (HTML). The mobility of the portable document file gives it a flexibility advantage—any staff can relocate it anywhere as needed and anyone with a computer can access the PDF regardless of operating system and without specialised software, as has been previously discussed—but the by-product of this portability and constant mobility is its ephemerality.

Transitory and fleeting. The documents’ online shelf life is typically a really short lifespan—documents tend to be moved as soon as consultations are over, or go randomly missing for a time, then pop up again. Borrowing digital theorist Mark Sample’s term, these documents are like ‘fugitive text’, “exist[ing] only as a trace, a lingering presence that confirms the absence of a presence.” ¹⁶ They are there but not there; there somewhere. The hint (trace) of their existence is often found in the footnotes of other documents that reference their once-textual availability.

In Sample’s literary reflection on The Archive or the Trace: Cultural Permanence and the Fugitive Text, there is a philosophical embrace of impermanence (“must everything be permanent?”) and a rumination on the fine art of disappearance. The GLA, however, can’t afford such slipperiness of their documents. It is impractical, and unmanageable, for both document producers and document searchers. Creative strategies are needed to trace the trace. In my research, I have had to rely on the Internet Archive’s Wayback Machine ¹⁷ and other webcrawling services like that of the Internet Memory Foundation The Internet Memory Foundation was a European non-profit foundation that preserved web content until 2018 and collected in a digital library. Webcrawling is the automated and systematic process by which computer scripts or programs index data on web pages. to find broken links and missing documents; and have had to use Zotero, a free reference management software, like Sample did, and as he puts it, to “capture any online document that poses even the slightest flight risk.” These archives have picked up GLA documents that fell through the crack, ones that have not made it to the LMA yet and may be floating somewhere still, or ones that are not archive-worthy because they are working copies or ‘live’ documents currently in use.

Because the GLA team has a set timeframe for the availability of documents and records of the London Plan, the result is a hit and miss of stumbling on the desired content at the right time. For instance, after 2018 and the close of the draft consultation and the Examination in Public processes, the evidence base to support the draft London Plan could no longer be found. The webpage instead displays an error message, “Sorry – you cannot access this content. This page may now be unpublished.” Similarly, the online consultation portal—the original interactive platform set up for digitally inputting feedback into the draft—has all but disappeared. The web address still exists but the webpage is empty. What’s left behind are PDF documents, an image of the submitted texts. Albeit there may be good reasons for this, to spare confusion of versions and distinguish between draft and final, the portal’s complete disappearance, without signposting to its once existence, nonetheless arguably creates more confusion when citations and references lead invariably to deadlinks.

Close to the date of the plan’s final publication, an email from Lisa Fairmaner, Head of the London Plan, to those who had participated in the plan’s formation, informed that, “once the new London Plan is formally published and any legal matters have been addressed, we will be deleting all records on the London Plan database.” ¹⁸ This note about the deletion of records advises that, “we also will not notify you of any future changes to the London Plan once it is formally published.” The onus is on participants to keep paying attention. Like the consultation process, readers and respondents must be proactive to keep abreast of document development and to stay informed of future London Plan updates.

In the book, Lost and Living (in) Archives, Annet Dekker writes that “the digital has allowed many new voices that act and assert agency,” and with “the open structure of the web, ordinary people can be heard and have the ability to influence existing content by adding their own.” ¹⁹ But, due to “the transient quality of digital archives,” which she argues “are not designed for long-term storage and memory, but for reproduction, for endless circulation between different levels, people, networks and locations,” there are also questions to ask of the living archive with “implications for memories” and “consequences of missing documents” to consider. This is the dilemma of the London Plan’s digital documentation, vulnerable to being erased, forgotten, and neglected, subject to random dispersal and constant disappearance.

In discussing the disruptive dynamism of digital archives, Dekker argues, “technology changes the content of what is archived”. Part of the disappearing act of London Plan documents relates back to the use of and the GLA’s reliance on PDF for mass communication and wide access. For all of PDF’s valuable functionality in ease of document sharing, it is an inherently disruptive technology, inserting itself into other systems. PDF’s convenience and perceived any/everywhere-ness complicate how knowledge is produced and shared, and importantly, accessed, since something shared does not automatically mean it’s been seen, let alone read. Because of the sheer volume of information PDF can make available, it consequently buries information, muddying genuine attempts at transparency and accountability pursued by the GLA and the Mayor in service of the corporate memory. As much as the portable document format facilitates, it also disrupts their record-keeping remit. Via PDF, the London Plan is made publicly available, but PDF also renders it publicly invisible.

On the history of the PDF, in an interview with Trevor Owens, the first Head of Digital Content Management for Library Services at the Library of Congress, Gitelman has said that, “today’s networked environment has helped promote the myth of total information—everything available to everyone.” ²⁰  At a click, we seemingly are able to access content from anywhere at any time. This was part of the promise of the PDF, in the age of personal computing powered by a network connection, instantaneous availability of information by the shortening of production and distribution. It is a myth, however, because what is available is not necessarily or perhaps always what is discoverable. The ubiquity of the PDF has made information widely available, yes; at the same time, it has also, consequently, made it easy to hide in plain sight. The interviewer makes reference to a 2014 Guardian article, ‘Is the PDF Hurting Democracy?’, which notes more than 30% of the World Bank’s PDF reports weren’t downloaded once in a five-year period, suggesting that, ²¹

In response, Gitelman doesn’t think it means democracy is “on the skids,” but she does recognise the problem. Gitleman points out, like grey literature i.e. technical literature which includes manuals, reports, white papers and which have a short shelf-life because they are so soon obsolete, PDFs are similarly the kinds of documents that are a “challenge to locate, much less preserve”. She makes the distinction that while documents like the World Bank PDFs may be findable, they are not necessarily mineable.

This is especially true in the case of GLA documents, which can be found—discoverable through chance, luck and/or persistence—but cannot be systemically mined for easy extraction. The GLA is aware of this paradox. A 2020 GLA Assistant Director’s decision memo (Decision ADD2460) on PDF accessibility states that, despite the tens of thousands of PDFs available online, “the majority are rarely viewed. In 2019 pages on london.gov.uk were viewed around 28 million times. The total number of views for our top PDFs combined, however, is less than one percent of the total pages viewed.” ²² Mathematically, this equates to 280,000 total PDF views, or in the 2019 calendar year, roughly 767 views per day. In a city of 8 million people, only .0096% of the population is looking at a PDF on an average day. That’s assuming that one view equals one person viewing, whereas in reality one person is likely to be viewing multiple documents at a time, in which case the total number of Londoners actually accessing PDFs is significantly less. One obvious reason for this may be that PDFs are hard to find. The burden is on the document searcher to, well, search for it. To sift through tranches of documents and follow helplessly (and hopelessly) down rabbit holes of hyperlinks. While the GLA makes their content viewable with PDF, it is not certain to be seen.

It is an accessibility problem that the GLA spends a lot of money to solve. In Decision ADD2460, expenditure of up to £25,000 was approved in the 2020-21 financial year for external services to make the GLA’s published PDF documents more accessible, following on expenditure of £50,000 in the 2019-20 financial year to audit and test accessibility of the GLA digital estate, in order to comply with the 2018 Regulations, a UK statutory requirement for public sector bodies on websites and mobile applications accessibility. ADD2460 (August 2020) follows on the findings from a February 2020 decision, ADD2424 Accessibility auditing and specialist advice, to enter into a contract with an accessibility agency to provide expert advice and accessibility testing for the london.gov.uk digital estate. An agency was commissioned to help update PDFs. Yet, even with external help, the memo acknowledges, this is a tremendous undertaking for the GLA that represents “a ‘disproportionate burden’ considering the resources currently available to the GLA and how infrequently the documents are accessed.” A number of the most important PDFs, as it notes, are very large and complex and to make them more accessible requires specific skills, tools and expertise to understand the guidelines and to convert these PDFs to meet the regulations. While most of their PDFs are not complex and can be easily converted in-house with appropriate training and guidance, as the memo further elaborates, in general it is not feasible to update all of the PDFs for accessibility nor is it a legal requirement if the PDFs were published before 23 September 2018.

For efficiency then, and to comply with the 2018 regulations, they will only update documents published on or after 23 September 2018 and that meet one or more of the following criteria: downloaded more than 1000 times in the last year; Mayoral strategies; governance documents which are required to interact with GLA (i.e. to make a complaint or to consult); or documents related to service provision where there is no alternative version. To improve accessibility in general they will include documents published before 23 September where they are key strategy documents within the current administration. What is unclear from the memo is what happens to documents that do not meet the criteria, PDFs that were produced before September 23 but are not of the consequential categories identified. Do they remain as they are, unchanged and not up to standard of accessibility, or will they be removed entirely? The state of PDFs—the state of digital documents—is therefore in flux at the moment on the london.gov.uk website. What stays online, what gets shelved, remains uncertain.

There’s a double burden at play here, one on officers to be digital archivists and make documents accessible, and the other, on site visitors to search for them. The irony is, in trying to comply with a regulation that aims to make documents more accessible, possibly more than a few documents will become inaccessible and even harder to find. Considering the complexity of the London Plan and its intensive document production, one significant consequence is the loss of work product, steps taken, decisions made, discussions recorded, and writings, notes, respondent comments and reports produced along the way. Transparency gets collapsed into only the draft document and final publication made available, leaving the in-betweens to chance, luck, and persistence to see the light of day again. It reinforces a clear linearity—a beginning and an end—at the expense of the meaningful ‘messiness’ that happens in the middle. While clarity is one goal of communication, which is often sought by governments to convey complex issues to the people they serve, this linearity sanitises the process, erases the complexity of the debate.

As Chapter 6 had touched upon about the clean version of the plan, neatness isn’t always what’s best. The final London Plan 2021 is free of markups and editorial changes, free of all signs of the intellectual, physical and emotional labour contributing to its production and shaping of London’s future. Without access to the documents that came before it, memory fades. While it’s not practical nor desirable to collect and retain every single piece of text, it is significant that in a democratic process centred on ‘having a say’ the subtexts and paratexts—the sayings themselves—stay hidden. Buried in servers, unseen and unread.

Further complicating this dispersed document landscape was the GLA’s endeavour to update the GLA website, a large and complex site that in its 2021 form was originally built in 2014-15 to meet accessibility guidelines at the time and which had become outdated. The existing site used open-source technology (Drupal 7) that became obsolete and no longer supported after November 2021. The GLA’s planned wholesale redevelopment of london.gov.uk included many pages, digital products and services in an effort to increase accessibility and meet current user and organisational demands, as part of £2 million GLA Digital & Technology Estate Rebuild. ²³ Digital estate includes services such as Talk London (a community site with high engagement), Team London (a site for matching volunteers with opportunities), Datastore (London’s free and open data-sharing portal) and the GLA Intranet. Digital estate is the GLA’s term referring to london.gov.uk and dozens of separate websites, applications and services, brought together under one banner. Around half a million people visit the digital estate every month. With the drop of support for Drupal, a content management system, it became urgent to modernise to sustain the site’s functionality and usefulness to users and equally internal operations. The last website redesign, the 2014 edition, took 18 months to complete. By the same timeline expected with this 2021 revamp, much of web development technology will have progressed, making the newly designed london.gov.uk already lagging behind.

From background info provided by the Mayor about rebuilding the GLA website and modernising the whole of the GLA’s digital estate, the websites were revealed to be a key engagement tool for various organisations and provider of services and information: “the digital estate is core to communicating, serving and getting Londoners involved in the work of the Mayor and the London Assembly.” ²⁴ Outlined in the Mayoral Decision and echoed from GLA Digital and Technology Strategy is the GLA’s ambition “to be an exemplar digital organisation, adopting and maintaining best practices for technology, digital and data. The GLA seeks to be a modern, user-centred and data-driven organisation.” It was also noted that, “the diversity of the GLA’s work means the website needs to communicate a wealth of information across a wide variety of themes. It is both an information source for Londoners, and a data repository.” Such a framing sees Londoners as information-seekers.

Migration of existing content is one of the major tasks in the overhaul. Here then is another opportunity for documents to be lost, for PDFs to be missed in the reshuffle. As a member of the Talk London community forum, I was invited to help test the new City Hall website. The test took me through a series of questions, each a variation of, if I was looking for X information on the website, where would I click expecting to find it, provided with options to choose my path from homepage to where I think the page with the desired content would or should be located. The web developers presumably were testing whether they had put things in the right place where people would intuitively expect them to be. The attempt to understand user journeys indicates that the redesign addresses a major challenge, site navigation, the ease or difficulty of finding things. This tracks with statistics that show, in 2019, of all views on the site, 58% of them were within the ‘What we do’ section.

Where do I go? How do I find X? These basic questions frustrate most site visitors to large institutional websites, especially those that are information-heavy, so the GLA is not alone in needing to ensure smooth content flow from one page to another. But in the instance of document hunting undertaken by participants in a consultation, the problem is that it partly assumes that visitors know what X is in order to even look for it. It assumes tacit knowledge of, or some familiarity with, an underpinning logic to their document infrastructure. The GLA stores nearly 80GB of website files. Officers themselves have a hard time finding what they are looking for. ²⁵

Even when website visitors know what to look for, finding documents is a particular challenge of engaging with the London Plan. £140,000 was spent by the GLA in 2017 on developing the London Plan Consultation Database (LPCD) to facilitate, record and analyse online responses to the consultation draft London Plan. ^{26 27} The move to an automated system to capture and analyse the responses was necessary because previous London Plan consultations used Excel spreadsheets and Access databases which were resource intensive, requiring responses to be entered manually and involving a substantial amount of officer time for analysing and coding responses. Developed by CTI Digital, a web and app development and digital marketing agency, the new system debuted at the end of 2017 coinciding with the draft London Plan publication, enabling respondents to browse the London Plan by sections and to make comments online in addition to submit responses by letter or email. With a search function within the database for internal users, LPCD drastically reduced officer time required to parse the information.

In comparison, for example, Mayor Livingstone’s office had received over 600 responses to the draft Replacement Plan raising over 12,000 specific comments or suggestions and a further 600 responses to questionnaires which were very widely distributed throughout London as part of the summary version of the plan. ²⁸ The 650+ detailed responses received were summarised and manually entered into a Draft London Plan consultation response database to assist in the task of searching and grouping comments, prepared for the GLA’s own purposes. The original responses were physically available for public inspection at the London Plan EIP office at City Hall. Reports extracted from the consultation response database were made more widely available as a PDF resource to assist in considering responses to the plan. The documents and responses were listed on the GLA’s website. ²⁹

Use of the 2017 system for the submission, analysis and publication of the comments was deemed a success with interest expressed in re-using it for further consultation exercises. But based on feedback, several structural issues persisted, namely the public search function as well as the visibility of comments. Users who had submitted comments could not easily view their comment even though it was stored on the system. Enhancements have since been made to address this key accessibility issue, but it represents another instance in which information is viewable but not mineable. Following the conclusion of the draft consultation and the EIP processes, the LPCD is no longer available for public viewing. It has gone offline. All online comments submitted through the system were converted into PDFs and remade available as links to the responses on the main London Plan landing page, once more collapsing the depth of the data into information silos, removed from their original contexts and severed from their relationships to and associations with other content.

In parallel, a new publications content type has been developed—a type of webpage that can be used instead of a PDF—to enable content supporting the next London Plan. consult.london.gov.uk is the Greater London Authority Engagement Portal launched in 2020 where Londoners can, through the dedicated website, give their feedback and complete surveys on the work the GLA is developing on behalf of the Mayor of London. Here they can also find documents and other material from earlier public engagement events, as well as participate in current consultations. An additional £40,000 was earmarked in February 2021 to further develop website functionality required to support the publication of the London Plan and London Plan Guidance and to ensure it is accessible to all Londoners. ³⁰

One of the expected outcomes from this was, “new webpages in a clear and easily navigable structure, with user-friendly formatting and engaging content.” In its current form, however, consult.london.gov.uk replicates some of the limitations already discussed. ‘User-friendly formatting and engaging content’ turns out to be animated graphics, while ‘clear and easily navigable structure’ translates into content organised in boxes and the use of bold fonts. Outside of these formal design tweaks, however, little has fundamentally changed. There are bits of useful information provided in HTML text, but to learn more, participants are still prompted to follow links, a few leads to other HTML webpages while the majority lead to downloadable PDFs. Taking the digital estate as metaphor, the grounds are vast, while its document trails and paths are sinuous.

While it has its advantages of being dynamic and more interactive, the transition to HTML away from PDF presents additional complications to an already challenging task of finding information. How will the two content types, the old and the new, interact and talk to each other? How will online content be supported off-line? The advantage of PDF and why it has been such a persistently favoured format as noted previously is its wide accessibility independent of having an internet connection. Will there be a duplication of documents to take account of different viewing and media consumption habits? What type of hybrid document then will come out of this moving forward? In the four years since the draft London Plan was published in November 2017, several million pounds were spent to answer the question, where is X? Or rather, where is the PDF located?

The move away from PDF to online-based content is in keeping with similar initiatives on other government websites and with wider public sector calls for greater access to information. Neil Williams, former Head of GOV.UK, the Government Digital Service, made an argument in 2018 for ‘why GOV.UK content should be published in HTML and not PDF’, based on an understanding that “GOV.UK exists to make government services and information as easy as possible to find and use,” similar to the core of the mission of london.gov.uk, which is a subdomain of GOV.uk. ³¹ Williams is now the Chief Digital Officer for Croydon Council. In 2017, the GLA also appointed its first Chief Digital Officer, Theo Blackwell, to lead London-wide digital transformation, data and smart city initiatives at City Hall. The following year he helped launch Smarter London Together, a roadmap that “sets out how we plan to transform London into the smartest city in the world.” ³²

Aims to make London smart aren’t new. Boris Johnson had his own Smart London Plan, which sat within the overarching framework of the Mayor’s Vision 2020. ³³ Going further back, the ‘smart city’ agenda has been around since Livingstone’s time, he too had ideas about the potential of “embracing information technology as a tool of government […] to provide more efficient services, reduce waste and empower citizens to make more informed decisions about the resources they consume and to take a more pro-active role in governing their locality”. ³⁴ The pattern across governments and mayoralties seems to be a technocratic approach to government communication, faith placed in putting the right systems in place to do the job of opening up government.

Nesta, self-billed as the UK’s innovation agency for social good, defines ‘digital democracy’ simply as “the practice of democracy using digital tools and technologies.” ³⁵ They do acknowledge, however, that democracy itself is not easy to define and that there are numerous definitions of digital democracy. “For some it refers to the use of digital tools to provide information and promote transparency, for others it describes the ways in which information and communications technologies (ICTs) can broaden and deepen participation, while others talk of promoting empowerment by enabling citizens to make decisions directly through online tools.” ³⁶ The UK Parliament views digital democracy through the lens of participation. Because “[t]he jargon and practices of the House can be alienating and the sheer weight of information about politics, now available, can act as a wall, keeping the citizen out of the mysterious world of Westminster,” it sees digital democracy as an opportunity to “encourage greater participation in politics from the widest possible range of people.” ³⁷

Implicit in this line of thinking, to which the GLA subscribes and that which drives the digital documentation of the London Plan, is that given the right tools, more people can participate. What this thinking obscures is that, sometimes, the tools themselves create the walls thwarting the access they purport to offer. PDFs add to as well as reveal the complexities and tensions in the idea of a digital democracy. The PDF is a perfect example of media opening up the opportunity to participate while simultaneously creating the barrier to do it. That is, if the PDF document can be found in the first place.

The reason for the particular difficulties in locating PDFs associated with the London Plan is that they are, by design, part of the ephemeral record, meaning they have no continuing administrative value to the GLA after the draft consultation ends and the final plan gets published. As suggested by the category titles of the EIP Library’s organisation, these documents were produced and made available for the specific purpose of public examination: “examination documents submitted by/to the panel,” “additional documents accepted by the Panel for the Examination,” “written statements in response to the panel’s matters paper,” and “wording changes suggested or read out loud by participants during hearing sessions.” ³⁸ The public status of documents, as in when they came online and went offline, was tied to the draft review process. Because “publics have activity and duration,” according to Michael Warner, from this time-based call for input we can say the producers and respondents of the London Plan belong to a durational public, gathered by and around a unique moment—around specific documents. ³⁹ Beyond this moment, these documents do not exist, at least not in the way they did during the consultation.

Outside of this temporal context, their public-ness is uncertain. Some documents, if fortunate enough to gain a mention in the final plan, become footnotes. Other documents, once having leverage, lose any power they may have had during the debate to sway arguments, instead relegated to history’s dustbin of the once-considered but politely-rejected, hidden away as a link part of a list of links. They will spend months/years languishing in a time-capture of the EIP Library, divorced from the finished product, until there is another London Plan to restart the documenting process anew. The documents, thus like the public(s) that access them, are also durational; they are not enduring, rather, existing temporarily, idling until the next iteration of the London Plan, their visibility reliant on whether/when they may be called upon again.

Such uncertainty, ephemerality and dispersal of documents pose a challenge to a conventional definition of publication, most commonly understood as being a singular, epistemic object of recognisable form and value, a book for example, attributable to an identifiable authorship and pinned to a specific site and context. The London Plan does not fit neatly into this framing; its form, value, site and context are durational, not static. A more meaningful way to think about its publication lies in philosopher Gilles Deleuze and psychoanalyst Félix Guattari’s definition of a book: ⁴⁰

Deleuze and Guattari sees assemblages as similar to machines, the book in this case is neither a whole nor a part, it is relational, its components not stable and fixed. They can be displaced and replaced, a type of coding that can be decoded. For them, “there are no individual statements, only statement-producing machinic assemblages.” ⁴¹ The coming together of the London Plan can be considered thus along the same lines of flight—there are no individual PDFs, only PDF-producing machinic assemblages. The London Plan is an assembled document of multiple documents. Or as Rebecca Ross puts it, “an assemblage of multiple documents.” Exchange with Rebecca Ross. Email. 13 September 2022. As a series of hyperlinks of associations with innumerable authors, it is an unattributable multiplicity.

This reframing of the London Plan’s publication is crucial to a better reframing of the meaning of ‘public,’ circling us back to earlier discussions in Chapters 1-3 about what/which public(s) is meant by public participation and examination in public. It is not the singular, universal public sphere conceptualised as the ‘all’ imagined by policymakers and participants alike. Rather, it is akin to sociologist Elaine Campbell’s re-imagination of a/the public sphere as an assemblage in this Deleuzian-inspired context, conceiving it “as a space of connectivity brought into being through a contingent and heterogeneous assemblage of discursive, visual and performative practices.” In her alternate notion, a/the public sphere is an assemblage, “an emergent and ephemeral space” assembled by connections. The public(s) of the London Plan is such a machined assembly, emerging from the ephemeral space created by its documentation and linked by the flurry of discursive, visual and performative activities. But, as with all machines, it’s vulnerable to breakdown.

Given this hyperactivity, with so much document movement, keeping up with the London Plan is taxing. Rather than a traditional reading of a book from cover to cover, materially engaging with the London Plan involves dealing with a variety of non-sequential text, akin to searching a database, trawling documents, chasing after deadlinks and becoming intimately familiar with the find tool. ⊕ A legacy webpage from Mayor Johnson’s time on london.gov.uk in which media content is left published but not publicly accessible. Only private. A memory of what once was. Readers read by keywords of interest. For example, in interviews, a community campaigner talked about looking for anything with the word ‘social’ in the draft plan while a housing activist was on the lookout for content related to ‘affordable housing’ so they can tailor their responses. ⁴² The same housing activist created her own spreadsheet of documents as a way to manage her way through them. ⁴³ It’s routine even to have a division of (reading) labour to share in the workload. Because of the plan’s complexity, collaborative reading becomes common practice. Conservative Assembly Member Andrew Boff, who was also the then chair of the Planning Committee with oversight of the London Plan 2021, necessarily had help from a scrutiny team to read the plan for/with him. ⁴⁴ He had not read every page, only skimming. Instead, a non-politically aligned researcher worked with each Assembly Member to review policy areas of interest, read on their behalf, and report back to them. Similarly, the LGTBQ+ response to the draft plan emerged from a group reading session—with the help of many hands—the act of reading together rather than alone to broaden understanding.

Webcrawling, wordsearching, keyword-filtering, and co-reading are documentary practices needed to navigate the London Plan’s hypertext. Readers engage in a combination of two types of reading modes literary critic Katherine Hayles discusses in the essay How We Read: Close, Hyper, Machine: 1) hyper reading, communication theorist James Sosnoski’s concept of “reader-directed, screen-based, computer-assisted reading,” and 2) machine reading, or “human-assisted computer reading, that is, computer algorithms used to analyze patterns in large textual corpora where size makes human reading of the entirety impossible.” ⁴⁵ Rather than a close reading, in which the primary text (e.g. the draft London Plan) is studied and read in its entirety, hyper reading is humans reading, guided by computers, while machine reading is inversely computers reading, guided by humans. The large textual corpora of the London Plan demand a hybrid of hyper-and-machine reading. But as Hayles points out, the consequences of reading on the web in this way, studies have shown, is that “hyperlinks tend to degrade comprehension,” “as the number of links increased, comprehension declined.” ⁴⁶ Understanding and interpretation are left up to what can be scraped by algorithms or textually parsed by bleary eyes.

As the archive expands beyond the limits of human attention, Kartrina Sluis, a scholar in politics and computation culture, finds the way in which users annotate and share mnemonic media to be a significant problem and raises questions about how social and cultural memory is exploited and ‘read’ by both human and non-human actors. She challenges digitisation’s promise “to unshackle our documents from the limits of space and time in favour of universal accessibility,” and instead argues, the digital archive isn’t limitless cyberspace. ⁴⁷ Rather than data liberation and dematerialisation, as database fever supplants philosopher Jacques Derrida’s archive fever, Sluis observes, the archival web is constrained by the technologies of memory, “linked to the industrial processing of information and the performativity of software” which have “material structures” and “contingent and specific economies.” ⁴⁸ Users of database-driven websites like the GLA’s rely on the search box to retrieve information, on the search algorithm to moderate the connections between seemingly infinite hyperlinks. Memory is therefore collectively reconstructed and recontextualised based on what the search results can yield.

As a communications designer and not a planner, unfamiliar with planning in general and planning in London in particular, interacting with the London Plan, its language and technicalities, was initially particularly challenging for me. I had to use my graphic skillset to look for ways to read the plan without actually reading it. Grappling with the texts involved using Adobe Acrobat’s advance search tool for PDF word filtering and Nvivo, the qualitative data analysis software, for word frequency, as well as writing my own HTML code to find link associations between documents. Since it’s impossible to read everything, a practice of selective reading was necessary, relying on machine help to process the texts. As plan readers, I and others became ‘co-extensive with technology,’ borrowing Hayles’s concept in Writing Machines. ⁴⁹

Siuán NiDhochartaigh, an artist who works with archives, has also identified such technological co-extension as a necessary condition of engaging in the world today, especially in light of COVID-19 that shifted attention and reconfigured visibility to an online space, exposing previously concealed sites of production. In her view, “new technologies, such as data harvesting, means that our attention is a form of production. Body and screen has become the extended technology of attention for socialising, labour and creative practices.” ⁵⁰ We must pay attention with and through our bodies.

In the case of engaging with the London Plan, between 2017-2021, I paid attention to its links of association by clicking and scrolling, opening and closing application windows, but most of all, by sitting. Following the London Plan around has involved various acts of sitting: sitting (in public meetings) and listening (to others speak); sitting (in front of my computer) and watching (online video feeds); sitting (at a table) and writing/typing (my response to the draft plan); and sitting (in front of someone, in-person or by video) and talking (about our experiences). Throughout the course of paying attention to the public processes of the London Plan, I have had to attenuate the way I sit, where I can sit, and with whom, taking note of those who take seats next to me.

Reading the plan didn’t just entail reading, it involved distributed forms of sitting and attention-giving. I have had to be hyper aware of texts on screen, their movements across digital environments (email, social media, official government or institutional websites, unofficial blogs), and their place off-screen, alert also to the space print copies of documents take up in my physical surroundings. As Sluis points out, the digital isn’t immaterial. For me, it has weight, sitting heavy on my back as I carried the plan and my laptop from public meeting to meeting. The digital extends to the fatigue I and others felt to sustain our focus on moving pixels. It extends to the worry during an EIP hearing of running out of battery on my laptop where my documents were located, needing to sit close by to an electrical outlet to keep it charged but at the risk of being out of hearing range from the speakers because of the room’s poor acoustics. The plan, taking up both mental and somatic space, ostensibly, makes material, contingent and specific demands on the body.

It demands we pay attention in embodied ways that go beyond simply staring at a screen. With the London Plan, participants must calculate where to distribute their attention so as not to miss an update or a document. They must balance a diversity of reading habits, go through the motions of version tracking, engage in a form of digital document stalking, in order to make what amounts to minor, non-fundamental changes to the draft. Catherine D’Ignazio and Lauren Klein, in Data Feminism, underscores the need to “show your work” and “make labour visible” so that alternative views/data/texts can be seen, recognised and valued. ⁵¹ The hyperlink is one way for the GLA to show the work behind the London Plan and make the labour of its production visible. Hypertext represents an opportunity to widen the plan’s perspectives and to add new associations that weave more threads in the archival web, yet it’s also an inadvertent way to hide all that work and labour—and divide attention. The EIP Library draws attention to the activities of the London Plan’s development while it simultaneously splits attention into unseen knowledge silos.

Consultations in planning are often framed as an easy set of steps to follow in order to participate: visit the website, download the document(s) to retrieve information as needed, review and then upload your comments to have your say. The London Plan is set up in this way, all paths lead back to the EIP Library on the GLA website, inferring that simply clicking, reading, and commenting would achieve the desired outcome of meaningful participation. It sounds like something that can be done breezily in an afternoon: click, read, comment. However, it’s far from a linear process. There are hundreds of steps in between, dozens of micro negotiations with different media, strains on personal time, and frequent oscillation between hyper and machine reading. It is also hyper in another sense, demanding hyper vigilance of individual and collective attention spanning more than a few hours.

Complicating the London Plan’s writing and reading is the density of documents that plan producers and respondents alike must grapple with. A tremendous number of lists and links must be followed to come to grips with the plan’s content. Information needs to be searched, parsed and managed. Because of the nature of the portable document format and the structure of hypertext language, this is sometimes easy, but at other times difficult because the paths are not always clear and the documents are constantly on the move. The dependency on PDF and reliance on the GLA’s website working as a live digital archive have implications about the publicness of the public consultation process considering that PDF documents are rarely viewed if they can be found in the first place. Text is fugitive; documents ephemeral. Not easy to pin down, sometimes, it slips out of grasp.

A deep dive into hyperlinks and hypertext reveals a rigid dichotomy between the perceived singularity of the London Plan and the digital context of its document multiplicity. The plan is the fulcrum of many conversations during the draft review about the shape the city should take, yet, outside of this process, it only ever appears as one publication—a clean slate de-contextualised from the public discourses animating its production. But the London Plan, as an assembled document of documents—visualised in the final chapter in my design proposal for a London Plan Public Library—exists in complex relationship with its media and takes intense effort to consume.