As We May Now Think: A Note on Vannevar Bush’s Scaffolding Claim

¶ 1 Leave a comment on paragraph 1 0 When we invoke Vannevar Bush’s now canonical “As We May Think“—first published in The Atlantic Monthly in July 1945—we tend to do so for one of two reasons. 1 On the one hand, we do so because of the article’s influence and invention: not only does it lay the conceptual foundations for hypertext and a number of other digital media, the article performs this work in vivid fashion, sketching a never-built, microfilm-based “speculative machine^“2 that promises to radically extend human capacities for thinking and remembering: the memex.3 On the other hand, we invoke the article because of what it anticipates: Bush, an engineer and inventor turned wartime science administrator, outlines conditions we now know well, such as the problem of informational hyper-production at the societal level, or the promise of democratized access to knowledge. He also offers compelling responses to those conditions, including an idea that technologies for research and retrieval should emulate human thought, especially its apparent associative dimensions. In these brief remarks, I do not contest these important reasons for citing Bush’s text, that is whether we may do so in classrooms, publications, or social media feeds.4 Instead, I do something a bit unusual: I unpack a long-overlooked claim in “As We May Think,” and, on this basis, argue for third and fourth reasons for invoking the text going forward. I also offer a specific thesis around the creative, epistemological, and pedagogical potentials of online archives, and I conclude by asserting the value for computer-aided thinking and practice in vocabularies of emergence, provisionality, and co-construction.

¶ 2 Leave a comment on paragraph 2 0 The claim in question goes like this: when the limitless sharing of documents, data, and media becomes possible and pervasive, then, Bush writes, “the inheritance from the master becomes not only his additions to the world’s record, but for his disciples the entire scaffolding by which they were erected.”5 These few words are as far as the author takes the claim. Nowhere does Bush clarify what the “scaffolding” of (apparently male, apparently disciple-beloved) “masters” looks like, nor by what means such scaffolding would come into being or circulate.6 Thus, we are left to extrapolate.7 Doing so requires imagination. We have to find a way to situate the scaffolding claim within what I noted above: the article’s conception of that seminal machine, the memex. A personal console about the size of a desk, outfitted with a pair of juxtaposed, tablet-like screens, the memex would afford its users access to vast quantities of the “human record.” (By this term, Bush appears to mean, in effect, anything published, but also a heterogeneous mix of other kinds of documents, images, and data.) The memex would also enable—and this is crucial—the production, interlinking, and navigation of custom-produced “trails.” From the perspective of the machine, these trails would constitute ordered lists, those lists made up of the locations of indexed content. From the perspective of the user, trails would constitute interactive assemblages, those assemblages bodying forth as progressions of visual and verbal media, one selected item appearing after the other, some of them offering opportunities to travel “side” or “skip” trails. To use Bush’s normative, martial example, a trail on how “outranged Europeans failed to adopt the Turkish bow” might move from a graphic of bow types to an article on the chemistry of elasticity to an encyclopedic account of the Crusades to a newly inserted “page of longhand analysis” and onward.8

¶ 3 Leave a comment on paragraph 3 0 For Bush, technologies like the memex would have transformative consequences. Among these are the reliable preservation of objects, annotations, and associations; the emergence of encyclopedias rife with interwoven articles on endless topics; and professionals having access to “the associated opinions and decisions of [their] whole experience, and of the experience of friends and authorities.”9 One further consequence is the emergence of a “new profession of trail blazers.”10 Not unlike present-day data curators, librarians, or Wikipedia contributors, these trailblazers would “find delight in the task of”—or at least perform the task of—“establishing useful trails through the enormous mass of the common record.”11 It is directly after this latter prognostication that the scaffolding claim appears: should technologies like the memex become pervasive, we would stand to inherit the “entire scaffolding by which” researchers’ contributions were “erected.” Thus, Bush proceeds to add before shifting topics, “science may implement the ways in which man produces, stores, and consults the record of the [human] race.”12

¶ 4 Leave a comment on paragraph 4 0 So, bearing in mind the full scope of his controlled techno-hallucination,13 what can we argue Bush means to achieve through this momentary, metaphor-dependent claim? In the first instance, I see two primary kinds of readings available; which kind we favor would appear to depend on the degree to which we invest in the specificity of the metaphor itself. Along one track, content to ignore the particularity of poles and planks, we might suggest that by “scaffolding” Bush means the various forms of evidence—data, images, tables—that have gone into supporting the production of successful research projects. Under this interpretation, memex-like machines would distribute archives of relevant materials. Crucially and provocatively, those archives would become as accessible as the publications their materials had helped realize. Alternatively, along a second track, one that embraces the specificity of physical scaffolding, the implications of Bush’s claim become more elusive, and also more alluring. In terms of the actual stuff of inherited scaffolding, that might not only consist in evidentiary archives. Rather, if physical scaffolding refers to arrangements as much as materials, the inheritance could also be understood to include the intangible relationships between points of evidence (and illustration and inspiration) in that archive, these relationships presumably preserved through various means, such as documentation, linking, or mapping.14 We could even venture to suggest that events of transmitting scaffolding would not only take place at the retirement or untimely death of “masters.” Rather, the production of scaffolding could also become an active concern: researchers could opt to assemble and distribute the scaffolding of projects in process or even yet to come. More radically, those so-called trailblazers could seek to disseminate the effects of scaffolding—e.g., providing resources that propel new thinking—through the publication of especially generative trails. Thus, trailblazers would commit themselves to enabling the construction of unseen, unknown buildings; that is, they would specialize in the domain of epistemological potential, its production, dissemination, and safekeeping.

¶ 5 Leave a comment on paragraph 5 0 Such extrapolation could detain us further, and indeed one could comb Bush’s papers for context around the inclusion of the scaffolding claim, as well as his decision to use this metaphor over others, such as root systems, or none at all. But the question I am asking does not require a story around intention. Rather, I am asking how, in taking the scaffolding claim seriously, our reading of this canonical text might be both destabilized and enriched, and how we might in turn find new reasons for invoking—i.e., teaching, citing, adapting—this text going forward. In order to pursue this line of inquiry, a lens switch is necessary. While so far we have interpreted the scaffolding claim for what it adds to Bush’s vision, what we need to do instead is analyze the scaffolding claim for what it indicates about Bush’s vision. That is, we need to discern what such a claim signifies, suggests, or reflects about the arguments, stakes, and commitments of “As We May Think.” We can then build with and alongside the scaffolding claim.

¶ 6 Leave a comment on paragraph 6 0 One way of proceeding requires us to double down on the scaffolding metaphor. That means finding orientation in the qualities and operations of actual physical scaffolding. I think of things like providing support; enabling co-construction; carrying a provisional status; being subject to recombination; and requiring spatial problem solving.15 Indeed, it is my contention that if we let awareness of such features suffuse our encounters with this text, we will eventually find that two crucial dimensions come into sharper relief, and that thus we can establish two further reasons for invoking the text going forward. The first reason is this: not just in its scaffolding claim, but in its idiom and invention overall, “As We May Think” makes compelling and conceivable a commitment to knowledge practices that are not strictly invested in finished products. Instead, those practices cohere around projects of critical and epistemological capacity building. In other words, they value the facilitation of unpredictable intellectual processes: the triumphs and false starts of tested associations, the potentials and pitfalls in heuristic rearrangements. Orit Halpern points us toward exactly these commitments when she writes that, for Bush, as would later be the case for Norbert Weiner via cybernetics, the screen “was not a representation of an outside reality”; rather, it was a “dynamic space to encourage the production of new associations, and further interactions—between people and between people and machines.”16

¶ 7 Leave a comment on paragraph 7 0 The second new reason for invoking “As We May Think” is more elusive. Consider that in Bush’s imaginary future, trailblazers would not draw up sterile lists of linked codes; rather, they would fashion actual, mutable sequences of concrete materials projected on tangible screens; they would have to construct compelling, practicable ways to help others navigate ideas, images, and information. In other words, to borrow from hypertext pioneer Theodore Nelson’s “As We Will Think” (1972), memex users would negotiate “new forms of interwoven documents” or “new documentary objects.”17 The key point is this: the term “scaffolding” helps to implicitly reestablish that Bush’s vision in this text is not only oriented toward capacity building; it is also fundamentally spatial and generative. It is animated throughout by a concern for what he would later call the “art of trail architecture.”18 Indeed, although we may not acknowledge as much, it is arguable that, when we occupy and explore digital contexts, we consistently, or even constantly, confront just this problem and possibility of interpreting, navigating, and ourselves producing documentary and visual-spatial meaning.19

¶ 8 Leave a comment on paragraph 8 0 I count the above as an all-too-rapid path toward articulating these two reasons for returning to “As We May Think,” what we could label capacity building and spatio-documentary communication. I move just as rapidly to ground this effort: I seek to demonstrate that there are contexts for which Bush’s text, newly engaged along these two lines, offers enabling if not exhaustive sources of guidance, assistance, or inspiration, whether for thoughts we may now think or for actions we may now perform. Thus I conclude with an example: the Japan Disasters Archive (JDA). Led by the Reischauer Institute of Japanese Studies at Harvard University, the JDA documents the earthquake, tsunami, and nuclear meltdown of March 2011. Defying its name, this online archive does not primarily preserve materials. Instead, its core missions consist in large-scale, ongoing aggregation and participatory curation. The web interface provides a consolidated search of varyingly sourced materials, including tweets, images, documents, videos, and websites. Meanwhile, the archive also invites users to build and share connections between those materials: contributors can convene aggregations of references into micro-publications called “collections.” These collections are assemblages of links, thumbnails, metadata, a title, and a description. They can serve a variety of functions. They can advance resources compiled during a research project; they can provide waypoints for disaster-related phenomena (such as the impact of the “language of flowers” in Fukushima); and, in a performative fashion, they can, by virtue of their ongoing accumulation, encourage future acts of archival participation.20

¶ 9 Leave a comment on paragraph 9 0 That the JDA instantiates the memex’s powers of aggregation, that collections estimate memex trails in their form and function, that collection-builders work in the mode of the trailblazer—these things are readily discernible. What requires elucidation is the project’s specific connection with Bushian scaffolding. Forging trails to other scaffolding claims will aid us in this task. There are several cases; two seem most relevant.21 For Jerome Bruner, a luminous figure in twentieth-century psychology, “scaffolding” pertains to the work of education. In particular, it denotes the employment of situation-specific physical and abstract supports in pedagogical contexts, supports as varied as tools, activities, curricula, comments, and questions. Configurations of scaffolding are continuously rearranged and replaced as learners cultivate skills, insights, and questions; importantly, such scaffolding is provided not only by teachers but also by parents and peers, even the physical environment itself.22 In a different vein, in a recent article, organizational theorist Wanda Orlikowski calls attention to “the provisional scaffolding that is erected by the temporary and situated engagement of technology in knowledgeable activity.”23 Orlikowski bases her account on observations of business activities like in-person management meetings and virtual, multi-platform collaborations, but her theorization of scaffolding is inherently expansive. Through Orlikowski’s performative and materialist lens, it becomes both possible and important to recognize the following: “for the duration of a particular human practice, actors draw on various artifacts, spaces, and infrastructures to conduct their activities. Once those activities have ended, actors stop engaging with the specific materiality engaged in those activities, and their scaffold is consequently dismantled.”24

¶ 10 Leave a comment on paragraph 10 0 By way of Bush, Bruner, and Orlikowksi’s scaffolding claims, we can essay the following thought about and with the JDA: online archives—or indeed any evolving aggregations of data, documents, and audio/visual media—can meaningfully scaffold world-expanding, if also risky and uncertain, activities of communal inquiry.25 That communal inquiry can be conceived as large scale and ongoing, as in the distributed participatory use of the JDA; that inquiry can also be conceived as small scale and temporary, as in a specific example I have in mind. The example derives from a seminar I co-taught with the JDA in Fall 2013.26 In order to prepare our students for a session on mass media and disaster, we asked them to engage in a trailblazing practice: they would populate a new JDA collection with relevant material discovered both within the archive and out on the web. (One student, for instance, shared a website collating bloggers’ satires of Japanese television news.) Crucially, the resulting collection became the basis for exploratory conversation during the session itself. Discernment of common patterns in small groups preceded a whole-class conversation around broader lessons (all of this informed by the assigned readings). As I now reflect on this exercise, it strikes me that our heuristic aggregation of digital objects provided substance, shape, and support for our “knowledgeable activity.” Those objects (and their perceived relationships) supported propositions, questions, and analogies; their display via the projector allowed them to undergo ready inspection and conceptual recombination. In the end, the apparent benefits were several: evidence co-existed with extrapolation; peers and teachers enacted lively collaborative learning; and ideas emerged that wouldn’t have otherwise.

¶ 11 Leave a comment on paragraph 11 0 Maybe there was a larger lesson in this exercise, one that has been implicit throughout this note: the ongoing encounter between archives, databases, and technology need not strictly orbit around finished products or final conclusions.27 Instead, emergent relationships and provisional associations, good enough for now, can matter as much as complete and coherent ideas, meant to persist indefinitely. Indeed, more frequently than we may think, what we should seek in scholarship, teaching, and librarianship is less the advancement of the established and more the facilitation of the uncertain. In other words, we can devote our energy toward expanding our capacities to question and connect as much as toward expanding our abilities to declare and distribute. I am not saying that “As We May Think” definitively claims these things, or that Bush’s legacy should stand unquestioned. Rather I am suggesting that the article, like the enduring image of the memex, remains distinctly good to think with. Indeed, more than that, “As We May Think” might be reclaimed for an alternative and still-unfolding history of critical computational praxis, one grounded not in sprawling and atomized acquisitiveness, but in expansive and collaborative capacity building.28

0. ¶ 12 Leave a comment on paragraph 12 0
1. An abridged and illustrated version of “As We May Think” appeared in Life in September 1945. By then, U.S. warplanes had dropped nuclear weapons on Hiroshima and Nagasaki—two horrific events that Bush’s wartime administration of the U.S. Office of Scientific Research and Development had helped bring into being. The article in question, like Bush’s subsequent writing, includes calls for a post-war science oriented toward knowledge production over weapons production. Here I refer to the reprinted, unabridged version of “As We May Think” printed in From Memex to Hypertext: Vannevar Bush and the Mind’s Machine, eds. James M. Nyce and Paul Kahn (San Diego: Academic Press, Inc., 1991), 85–110. [↩]
2. Ibid., 45. [↩]
3. “Memex” is a portmanteau of “memory” and “extension.” As Nyce and Kahn write, “Memex got its name because it was to support and extend the powers of human memory and association” (“A Machine for the Mind,” in Nyce and Kahn, From Memex to Hypertext, 57). [↩]
4. I would, however, caution against too settled an account of what is, in fact, a peculiar historical document with a complex reception history, several earlier iterations, and a handful of follow-up writings. For these reasons, I recommend From Memex to Hypertext for anyone interested in better understanding the background, development, and influence of Bush’s thought. [↩]
5. Bush, “As We May Think,” 105. [↩]
6. As far as I can tell, no one has sought to interpret this claim either. In those instances I have encountered where the sentence does appear, it has been accompanied by the preceding one on “trail blazing,” and no commentary is offered on the notion of scaffolding itself. [↩]
7. Later reflecting on “As We May Think,” Bush observes, “I extrapolated freely as I wrote” (“Memex II,” in Nyce and Kahn, From Memex to Hypertext, 165–84, 166). [↩]
8. Bush, “As We May Think,” 104. [↩]
9. Ibid., 105. [↩]
10. Ibid. [↩]
11. Ibid. [↩]
12. Ibid. Bush finds trails compelling partly because they appear to externalize and augment mental associative processes. Notably, in a sequel to “As We May Think,” Bush imagines a second machine, Memex II, that would feature machine learning. Memex II would predict trails of potential interest based on data around users’ transits. See Bush, “Memex II.” [↩]
13. The idea that visual perception is a “controlled hallucination” is attributed to the computer scientist Max Clowes, as originally quoted by David Waltz. [↩]
14. Extending our extrapolation still further, inherited scaffolding could include documentation of the material objects and contexts supporting inquiry as well, things like the layout of the physical laboratory or special software programs or other tangible tools or environments for inquiry. [↩]
15. I derive the notion of breaking down the scaffolding metaphor into component features from Wanda Orlikowski, who proposes these features: temporary, flexible, portable, diverse, heterogeneous, emergent, stable, dangerous, generative, constitutive. See Wanda J. Orlikowski, “Material Knowing: The Scaffolding of Human Knowledgeability,” European Journal of Information Systems 15, no. 5: 461–62. [↩]
16. Orit Halpern, Beautiful Data: A History of Vision and Reason Since 1945 (Durham, NC: Duke University Press, 2015), 73. [↩]
17. Theodore H. Nelson, “As We Will Think,” Online 72 Conference Proceedings 1 (Uxbridge, 1972), 439–54; reprinted in Nyce and Kahn, From Memex to Hypertext, 245–60. The two quotations derive from pages 245 and 247. Nelson, seemingly referencing Bush’s idea of “disciples” inheriting scaffolding from masters, frames his own essay as one of “counter-discipleship” (246). Because, Nelson says, Bush’s article is generally misinterpreted as having to do with “information retrieval” (245), Nelson hopes to “remind readers of what Bush did and did not say, and point out what is not yet recognized: that much of what he predicted is possible now; the memex is here; the ‘trails’ he spoke of—suitably generalized, and now called hypertexts—may, and should, become the principal publishing form of the future” (246). [↩]
18. Bush, “Memex II,” 178, my italics. [↩]
19. Indeed, though we largely seem to have forgotten, some early experiments in hypertext sought to seize on the spatial linking of documents as the basis for novel ways of producing and embedding knowledge and perspective. For one instance, one approach involved linked “cards,” as with a system called NoteCards developed by Randall Trigg, Frank Halasz, and Thomas Moran at Xerox PARC in 1984. See Randall H. Trigg, “From Trailblazing to Guided Tours: The Legacy of Vannevar Bush’s Vision of Hypertext Use,” in Nyce and Kahn, From Memex to Hypertext, 353–67. [↩]
20. I should note that, at least based in my own observations, the JDA, although successful in many dimensions, has not generated quite the intensity and breadth of participation anticipated by those involved in its beginnings. (As of early September 2016, the JDA does host a high number of collections at 323; however, most of these have been assembled by project affiliates.) Of course, the JDA might well eventually generate higher levels of participation, and there are many ways to measure the success of such a project, but it seems fair to suggest that barriers will persist, including the problem of credit. Indeed, such questions of what counts as true scholarly contribution matter significantly to recent debates in the digital humanities and digital scholarship more broadly. Does contributing to a digital archive count as a fulfillment of one’s professional duties, whether one is an academic, curator, librarian, or otherwise? Anticipating such issues over two decades ago, Norman Meyrowitz suggests the following: “We must encourage trailblazers and allow the scaffolding and rhetoric of hypertext to evolve by rewarding such work professionally.” It is worth noting that this is the only direct reference to Bush’s scaffolding claim I have found; Meyrowitz does not elaborate on the term. See Norman Meyrowitz, “Hypertext—Does it Reduce Cholesterol, Too?” in Nyce and Kahn, From Memex to Hypertext, 314. [↩]
21. One notable instance is in the field of distributed cognition. As cited by Orlikowski, Andy Clark writes, “We may call an action ‘scaffolded’ to the extent that it relies on some kind of external support. Such support could come from the use of tools, or the knowledge and skills of others; that is to say, scaffolding (as I shall use the term) denotes a broad class of physical, cognitive and social augmentations—augmentations which allow us to achieve some goal which would otherwise be beyond us.” See Andy Clark, “Magic Words: How Language Augments Human Computation,” in Language and Thought: Interdisciplinary Themes, eds. Peter Carruthers and Jill Boucher (Cambridge, UK: Cambridge University Press, 1998), 163. [↩]
22. Bruner’s concept has proven highly influential in educational discourse. Here is one instance in which he employs the term “scaffold”: “How best to conceive of a subcommunity that specializes in learning among its members? One obvious answer would be that place where, among other things, learners help each other learn, each according to her abilities. And this, of course, need not exclude the presence of somebody serving the role of teacher. It simply implies that the teacher does not play that role as a monopoly, that learners ‘scaffold’ each other as well. The antithesis is the ‘transmission’ model first described, often further exaggerated by an emphasis on transmitting ‘subject matter.’ But in most matters of achieving mastery, we also want learners to gain good judgment, to become self-reliant, to work well with each other. And such competencies do not flourish under a one-way ‘transmission’ regimen. Indeed, the very institutionalization of schooling may get in the way of creating a subcommunity of learners who bootstrap each other.” Jerome Bruner, The Culture of Education (Cambridge, MA: Harvard University Press, 1996), 21. [↩]
23. Orlikowski, “Material Knowing,” 462. [↩]
24. Ibid. [↩]
25. For an insightful discussion of web-based “thematic research collections” as a kind of web-based genre, see Carole Palmer, “Thematic Research Collections,” in A Companion to Digital Humanities, eds. Susan Schreibman, Ray Siemens, and John Unsworth (Malden, MA: Blackwell Publishing, 2004), 348–65. [↩]
26. As a graduate student, I co-taught this seminar with Theodore Bestor (Anthropology) and Andrew Gordon (History). It was called “Japan’s 2011 Disasters and Their Aftermath: A Workshop on Digital Research.” Seed funding was provided by the Harvard Initiative for Learning and Teaching. [↩]
27. I adapt this notion of “encounter” from Johanna Drucker, who frames an ongoing encounter between “humanities work and digital tools.” See “Humanistic Theory and Digital Scholarship,” in Matthew K. Gold, ed., Debates in the Digital Humanities (Minneapolis: University of Minnesota Press, 2012), 88. [↩]
28. For an incisive account of our contemporary, capital-driven media landscape—and possible alternatives thereby—see Jonathan Crary, 24/7: Late Capitalism and the Ends of Sleep (London: Verso, 2013). [↩]