It has taken digital a lot longer than many had thought to provide a serious challenge to print, but it seems to me that we are now in a new moment in which digital texts enable screen-reading (if it is not an anachronism to still call it that) as a sustained practice. Here, I am thinking particularly of the way in which screen technologies, including the high-resolution “retina” displays common on iPhones, Kindle E Ink, etc., combined with much more sensitive typesetting design practices in relation to text, are producing long-form texts that are pleasurable to read on a screen-based medium and as e-books. This has happened most noticeably in magazine articles and longer newspaper features, but is beginning to drift over into well-designed reading apps that we find on our mobile devices, such as Pocket and the Reader function in Safari.
With this change, serious questions are being asked about our writing practices—especially in terms of the assumptions and affordances that are coded into software word-processors like Microsoft Word, which assumes and sometimes enforces a print mentality. Word wants you to print the documents you write, and this prescriptive behavior by the software encourages us to “check” our documents on a “real” paper form before committing to it—even if the final form is a PDF. The reason is that even the PDF is designed for printing, as anyone who has tried to read a PDF document on a digital screen will attest. But when the reading practices of screen media are sufficient, then many of the assumptions of screen writing can be jettisoned, especially the practice of writing for paper.
There is little doubt that writing and reading the screen is different from print (Berry 2012; Gold 2012). These differences are not just technical; they also involve forms of social practice, such as reading in public, passing around documents, sharing ideas, and so forth. They also include the kinds of social signaling that digital documents have been very poor at incorporating into their structures, such as the cover, the publisher, the author’s name, and the book’s unique design. Nonetheless, at the present phase of digital texts, it is in the typesetting and typography, combined with the social reading practices that take place, such as social sharing, marking, copying/pasting, and commenting, that make digital a viable way of creating and consuming textual works. In some ways, the social signaling of the cover artwork, etc. has been subsumed into social media such as Facebook and Twitter, but I think that it is only a matter of time before this is incorporated into mobile devices, since advanced screen technologies, especially an E Ink back cover, can be built for pennies.
To return to the texts themselves, the question of writing, of putting pen to paper, is on the cusp of radical change. The long thirty-year period of stable writing software created by the virtual monopoly that Microsoft gained over desktop computers is drawing to a close. From its initial introduction in 1983 on the Xenix system as Multi-Tool Word and renamed that year to the familiar Microsoft Word that we all know (and often hate) today, print has been the lodestar of word processor design.
As the next stage of digital text emerges, many of the textual apparatuses of print are migrating to the digital platform. As they do so, the advantages of new search and discovery practices make books extremely visible and usable again, through tools like Google Books (Dunleavy 2012). There is still a lot of experimentation in this space, and some problems still remain: for example, there is currently not a viable alternative to the “chunking” process of reading that print has taught us through pages and page numbering, nor is there a means of book marking that is as intuitive as the changing weight of the book as it moves through our hands, or the visual clues afforded through the page volume changing from unread to read as we turn the pages. However, this has been mitigated by turning away from the very long-form book- or monograph-length texts of around 80,000 words, to the moderate long-form, represented by the 15-40,000 word text which I want to call the minigraph.
By minigraph I am seeking to distinguish a specific length of text that is able to move beyond the limitations of the 6-8,000 word article, but avoids the chunking problem of reading lengthy digital texts. In other words, in its current stage of implementation, I think that digital long-form texts are most comfortable to read when they stay within this golden ratio of 15-40,000 words, broken into five or six chapters. The lack of chunking is still a problem without helpful “page” numbers, and I don’t think that paragraph numbering has provided a usable solution to this. However, the shortness of the text means that it is readable within a reasonable period of time, creating a de facto chunking at the level of the minigraph chapter (2,000 – 5,000 words). Indeed, the introduction of an algorithmic paging system that is device-independent would also be helpful, for example through a notion of “planes” which are analogous to pages but calculated in real-time.1 This would help sidestep the problem of fatigue in digital reading, apparent even in our retina/e-ink screen practices, but also creates works that are long enough to be satisfying to read and offer interesting discussion, digression and scholarly apparatus. Other publishers have already been experimenting with the form, such as Palgrave with its Pivot series, a new e-book format: “at 30,000 to 50,000 words, it’s longer than a journal article but shorter than a traditional monograph. The Palgrave Pivot, said Hazel Newton, head of digital publishing, ‘fills the space in the middle’” (Cassuto 2013). Indeed, Stanford University Press has also started “to release new material in the form of midlength e-books. ‘Stanford Briefs’ will run 20,000 to 40,000 words in length.” Cassuto calls Stanford’s format the “mini-monograph.”
How should one write a minigraph? It’s likely that Microsoft Word will algorithmically prescribe paper norms, which in academia tend to either 7,000-word articles or 70,000-word monographs. Here, I think Dieter (2013) is right to make links with the writing practices of Book Sprints as a connecting thread to new forms of publishing (Hyde 2013). The Book Sprint is a “genre of the ‘flash’ book, written under a short timeframe, to emerge as a contributor to debates, ideas and practices in contemporary culture…interventions that go well beyond a well-written blog-post or tweet, and give some substantive weight to a discussion or issue…within a range of 20-40,000 words” (Berry and Dieter 2012). This rapid and collaborative means of writing tends toward the creation of texts of an “appropriate” size for the digital medium. Book Sprints usually involve 4-8 writers, facilitated by another non-writing member. The output of each writer throughout the sprint conveniently maps onto the structure of minigraph chapters discussed earlier. For Dieter, the Book Sprint is conducive to new writing practices, and by extension new reading practices for network cultures, and therefore “formations that break from subjugation or blockages in pre-existing media and organizational workflows” (Dieter 2013). In this I think he is broadly correct; however, Book Sprints also produce texts that are conducive to reading and writing in a digital medium, especially in terms of word count.
Nick Montfort (2013) has suggested a new predominantly digital form of writing that enables different forms of scholarly communication, the technical report, which he argues “is as fast as a speeding blog, as detailed and structured as a journal article, and able to be tweeted, discussed, assessed, and used as much as any official publication can be. It is issued entirely without peer review.” Montfort, however, connects the technical report to the “grey literature” that is not usually considered part of scholarly publishing as such. Experiments like the “pamphlets” issued by the Stanford Literary Lab, and which Montford argues are technical reports in all but name, are between 10-15,000 words in length: slightly longer than a journal article and a little shorter than a minigraph.
However, a key difference is that neither the Book Sprint nor the technical report are peer-reviewed, although they might be “peer-to-peer reviewed” (see Cebula 2010; Fitzpatrick 2011). Rather, they are rapid production, sharing, and collaborative forms geared toward social media and intervention or technical documentation. In contrast, the minigraph would share with the other main scholarly outputs—the journal article and the monograph—the need to be peer-reviewed and produced at a high level of textual quality. This is where the minigraph points to new emergent affordances of the digital that enable the kinds of scholarly activity, such as presenting finished work, carefully annotated and referenced, through these nascent digital textual technologies. If these intuitions are right about the current state of digital technologies and their affordances for the writing and reading of scholarly work, then the minigraph might be the right structure and form for digital scholarship to augment the current ecosystem of the article, review, monograph, and so forth.
In some ways the minigraph is a much less radical suggestion than the multi-modal, all-singing, all-dancing digital object that many have been calling for. However, the minigraph, as conceptualized here, is still potentially deeply computational in form. We might describe the minigraph as a code-object. In this sense, the minigraph is able to contain programmable objects itself, in addition to its textual load, opening up many possibilities for interactive dimensions, like those suggested by the Computable Document Format (CDF) created by Wolfram.
The minigraph as described here does not, of course, exist as such, although its form is detectable in the documents produced by the Quip app, the dexy format, as “literate documentation,” or the Booktype software. It is manifestly not meant to be in the form of Google Docs/Drive, which is essentially traditional word-processing software in the cloud, and which ironically still revolves around a print metaphor. The minigraph is a technical imaginary for what digital scholarly writing might be. It remains to be coded into concrete software and manifested in the practices of scholarly writers and readers. Nonetheless, as a form of long-form text amenable to the mobile practices of readers today, the 15-40,000 word minigraph text could provide a key expressive scholarly form for the digital age.
 Minigraph chunks would be at 250-350 word intervals, roughly pages, and chapters of 2-5,000 words. There is no reason why the term “page” could not be used for these chunks, but perhaps “plane” is more appropriate in terms of chunks representing vertical “cuts” in the text at an appropriate frequency. So “plane 5” would be analogous to page 5, but mathematically calculable to approximately (300 x plane number) to give start word, and ((300 x plane number+1)-1) to give the end word of a particular plane. This would make the page both algorithmically calculable and therefore device-independent, but also suitable for scholarly referencing and usable user-friendly numbering throughout the text. As the planes are represented on screen by a digital, the numbering system would be comprehensible to users of printed texts, and would offer a simple transition from paper page-based numbering to algorithmic numbering. If the document was printed, the planes could be automatically reformatted to the page size, and hence further make the link between page and plane straightforward for the reader (who might never even realize the algorithmic source of the numbering system for plane chunks in a minigraph). Indeed, one might place the “plane resolution” within the minigraph text itself, in this case “300”, enabling different plane chunks to be used within different texts, and hence changing the way in which a plane is calculated on a book-by-book basis—very similar to page numbering. One might even have different plane resolutions within chapters in a book, enabling different chunks in different chapters or regions.