Alongside art history and archaeology, palaeography is one of the disciplines most dependent on visual data. Palaeography is strictly speaking the study of the history of handwriting, but the term is used to refer more broadly to the study of handwritten books, including their physical make-up, a focus also known as codicology, and the wider history of the book. Students in many humanities disciplines learn basic palaeography as a research skill, one of the Grundwissenschaften or ‘foundational subjects’. A smaller number make palaeography the focus of their research: the history of styles of script, page design and physical features of manuscripts which differ from age to age. To study the script or design of books depends on seeing artefacts or, for scholars far from the objects, images of them.
Yet manuscript books are fairly numerous: to take just those from before the introduction of printing in the late 1400s – and people still wrote by hand after that, and still do now – the Bodleian Library has about 10,000, the British Library even more. We might know some of the treasures on display in galleries, but the majority are not especially beautiful; and while they all matter to some specialist, nor are the all of them of massive historical or literary significance. And to take just western European ones: through inheritance, upheavals and the antiquarian trade, they have ended up in libraries scattered far from their origins, across European border, with many won by wealthy collections in the USA and others in Japan, Canada, Australia and New Zealand. It will be a long time before the money and organization is found to digitize them all – or even to catalogue them thoroughly – and it takes a long time to travel round consulting the manuscripts germane to any one research project.
Palaeography has two problems, then: it depends on visual evidence; and its sources are scattered and voluminous. To overcome these challenges, some of the earliest scholarship in palaeography involved forging new typefaces to represent older styles of Anglo-Saxon script in the sixteenth century and attempts at engraved facsimiles by Jean Mabillon in the seventeenth century. When photography was invented in the nineteenth century, contemporaries such as Ludwig Traube recognized its transformative effects and called their age ‘the age of photography’ in the history of this discipline. Photographs offered a way of remembering manuscripts, of comparing and contrasting manuscripts and of proving one’s arguments to others. Palaeography, a site-specific sort of scholarship, became networked, interconnected, through photographs.
Early processes of digitization made it easier to make and share photographs of manuscripts, and some simple sites which simply archive full image sets survive. The picture files can be difficult to handle, poorly labelled and minimally framed with explanation. They are, nonetheless, a valuable tool for consulting important manuscripts to which access can be hard to obtain, for reasons of conservation or value. Any image is, as they say, ‘better than nothing’ – and that is too faint a praise for such early but crucial tools. Teaching and research in manuscript studies have long benefited large-scale or library-led digitization.
That said, there has been some caution about the reliance on consulting images of medieval manuscripts rather than consulting the manuscript themselves. For instance, A. S. G. Edwards in the UK’s Times Literary Supplement wrote a sceptical critique of the intellectual value and the value for money of large digitization projects such as Parker-on-the-Web. He was being polemical but he has a point. Large-scale digitization will not solve palaeographers’ problems of access, until the enormous manuscript libraries of the UK and Europe receive some correspondingly big grants. And the process of photography would arguably still have a role, in research and teaching palaeography, even after the books had been properly and accessibly digitized and networked. For scholars to study these images easily and build them into their teaching flexibly, the images would need not to be mounted on – and those already mounted would need to be removed from – the complex image-viewing platforms used in many digitization projects and made readily downloadable, following the methods used by William Noel for the Walters Museum. And libraries would need to allow open use of their images for scholarly purposes. Until the finances, technology and legal framework of the world are transformed, DIY digitization by users will have a crucial role in palaeography.
To speak for myself only: the ability to take photographs for myself has transformed my process of research since 2010: the workflow, the kinds of questions asked. But while there have been some analyses of the effect of digital images on research in general, and on networked images in particular, there is little reflection on the process of taking photographs of manuscripts for oneself. Giulio Menna has written a warm and personal account of the physical challenges and aesthetic pleasures of taking photographs of manuscripts. His images are outstanding, using focus and depth of field to direct the eye and to communicate evocatively – often with a wider audience as part of a team working with Erik Kwakkel to share palaeography with the general public.
First and most obviously, DIY digitization assists access in many ways and overcomes the geography of manuscript holdings. It allows people who can visit a library only briefly to take images from which they can study the manuscripts more fully later. It removes the need to make repeat visits to check things later. It allows people to share images of manuscripts that they have consulted which others could not travel to consult. The postgraduate students in Oxford, for example, are known to take photographs of manuscripts in the Bodleian to share with other students round the world. The curator of medieval and early modern collections in the Bodleian Library, Dr Martin Kauffmann, reports anecdotally, that some people make flying visits to the Bodleian and photograph books rather than read them now. In these respects, DIY digital photography seems little different from large digitizations such as Parker Library on the Web, or even from microfilms or Photostats, which were long used to obviate the need for travel or were shared between colleagues.
Increasing access, though, is not the something that DIY digitization best does: you still need one research trip of some duration or need at least somebody to visit the library to take pictures. In this respect, it is less convenient than digitizations mounted online by library professionals. DIY digitization is helpful for people far from special collections libraries, but is most easy for people already near them. Rather, DIY digitization is important not, or not only, because it transforms access to manuscripts, the resources of research, but also because it transforms the process of researching itself.
First, DIY digitization changes what researchers and teachers can do with the access to books that they have. While DIY photography cannot replace a visit to the library, it can supplement or extend it, so that a reader can consult a book in person and then continue to work on it in photographic form. This extension of access extends the kinds of research one can do. For instance, attention to the style of script in a manuscript is difficult without a long acquaintance with the book; script is not uniform page by page, as a typeface is, but is subject to small alternations letter by letter as a scribe works, and with changes of scribe. Likewise, transcribing a lengthy text from a manuscript might take too long on a short visit to a library far from home; or it might seem too risky without time to check the transcription. DIY photography makes it possible to continue and check such work later, after a visit to a library. This is not to say that DIY photographs could take the place of the final critical edition of texts; they do not, of course, have all the scholarly apparatus that an edition has. (Equally, nor can a professionally produced digital facsimile supplant the critical edition: the kind of reading and textual knowledge it offers is quite different – a focus on the document more than on the text.) But it could, in theory, be a crucial tool for renewed research into textual transmission and editing texts from manuscripts. To take a personal example: over several months in 2006 and 2009 I studied textual corrections in dozens of manuscripts in the Huntington Library in California; I could not afford to recheck everything twice. By contrast, if I were doing this in 2016, I could photograph the relevant materials and recheck my sources. When, in 2011 and 2012 I made a complementary study of manuscript copies for which the scribes’ direct exemplars survive, I could take photographs of one manuscript to compare to another elsewhere, as I did with related copies of Chaucer’s The Canterbury Tales in Chicago and Oxford, London and Tokyo.
In my own specialist area, the study of manuscripts in Middle English from the twelfth to the fifteenth centuries, such textual scholarship, thanks to photography, would be a slight divergence from some other recent work, which has often focused on the study of the compilation of the contents of books, on their material format or on the marginalia or ownership notes of readers – features which can be assessed more quickly than one can transcribe long passages of text. Whereas recent literary studies – whether or manuscript culture or later periods – have focused on the book as artefact or on the sociology of literature, an ability to capture and attend to the words on the page – literally – might be a shift, perhaps a backwards one, towards close attention to the text. That could seem formalist, or superficial, or an intense kind of ‘surface reading’, tracing the work of scribes more fully, even tautologously, than has hitherto been possible. These are speculations, though, for the potential of photography to transform the study of such manuscripts has not yet been realized.
Yet DIY photography could transform the researcher’s mode of interaction with manuscript sources – not only by extending the amount of access but the mode of it. Some past discussion of digital technology has considered the cognitive shifts of looking at digital images of manuscripts rather than the books themselves. The shift from looking at the real books to looking at images has been seen as a procedure which risks error and misrepresentation – true enough, unless one is cautious about trusting images. And it has been seen as a spiritual shift, almost, whether of reduction or of mystification: Walter Benjamin long ago suggested that there was a loss of ‘aura’ from the artwork in the age of ‘mechanical reproduction’, while Michael Camille conversely argued that facsimiles increase the mystique of manuscripts – referring specifically to the Très Riches Heures of the Duc de Berry, one of the most famous illuminated manuscripts from fifteenth-century France – by creating fame for the manuscript and giving libraries reason to prevent people from ever handling the real thing.
But those risks and imaginative shifts might occur whether palaeographers are looking photographs taken by others or by themselves; or the risks and shifts might be less if people look at photographs they have taken of books they have already handled. What differs between images obtained from other large digitizations already mounted online and DIY photography is the researchers’ process of making images for themselves. There is a long tradition in art criticism of photography of considering the aesthetics, ethics and politics of taking a photograph – about the viewer’s perspective and ethical stance as she selects what to photograph, how to frame it, whether to publish it. Some of those points could be fruitfully considered in relation to taking photographs as a researcher into early manuscripts, and indeed other sources. One does not need to range too far through art criticism and the theory of representation; one just needs to reflect, in quite practical terms, on what we do when we take pictures – selecting, framing, saving, compiling. For this process makes DIY digitization into a new way of seeing for palaeography.
Many commentators have noted photography’s power of selecting, framing and preserving. The essential terms of the debate emerge in one the most famous essays On Photography by the great critic Susan Sontag. Sontag notes that what we photograph reflects our prior conceptions of what is important: ‘it is never photographic evidence which can construct – more properly, identify – events; the contribution of photography always follows the naming of the event.’ Photography in archives is always shaped by our prejudgements of what is important. (This is true even of digitizations of what purport to be whole manuscripts. Do they include the binding? The fore-edge? The view of the closed book from above?) Conversely, though, Sontag notes the way that ‘photographs alter and enlarge our notions of what is worth looking at and what we have a right to observe’ and suggests that ‘To photograph is to confer importance’ on a subject, even to beautify it. By selecting something to photograph, reify our point-of-view or our interpretation of what matters – of what to highlight, of what to ignore – in ways that shape future thought. The decision to gather images of one text or another, one feature of manuscripts or another determines what one will go on to study later – especially if one then leaves the library and works at a distance from the manuscripts themselves. Sometimes, what one selects for photography will also be guided by preconceptions about what is important; at other times, what we photograph – maybe what ‘invites’ photography, as we say – might be guiding, in a way not necessarily envisaged before entering the library, what is photographed and so what is later studied. I do not think we need to pin down the circular process one way or another: what is important to note is that this method, which would seem descriptive, documentary, is in constant balance with a research question which drives it, and which at other times it drives.
For DIY digitization is not a method of archiving, of librarianship, but of scholarly attention and analysis. (In this, it differs completely, in my view, from the production of large digitization schemes.) It is not a way to consume or archive images for other people but is, like other techniques or technologies for note-taking and communicating, a new way to produce new description and interpretation. As intellectual historians have shown us, scholarly methods shape what scholars know. In this respect, studying a manuscript with a digital camera is most similar not to using a large-scale digitization online or a microfilm but is more similar to taking notes by pencil or laptop. Just like the note-taker, the photographer must select what to include or exclude, quote or summarize, zoom into or zoom out from. The management of selection and the balance between closeness or depth and distance or surface are two of the greatest intellectual challenges in research and writing in the humanities.
The process of selecting things to photograph might be compared to the process of curating an exhibition, with all the intellectual effort that entails. By selecting parts of manuscripts to photograph, one is selecting materials for reflection on a given theme or topic – a copy of a text for transcription, perhaps, or a timeline of the history of a style of script, or a corpus of examples of book bindings. That process of organizing knowledge around the researcher’s own ideas is something that consulting somebody else’s digital archive – even one made by a funded research project – does not quite offer. A collection of images which any one scholar or project takes is, in a way, useless for other scholars; it asks and answers the maker’s questions in particular, and it risks trapping others to think in their pre-established grooves. For instance, an archive of images of all the works of an author must often decide, in an age when attributions of works are often unproven, what is by that author; and other scholars might not share those attributions, or they might not have been the ones made in early centuries, when many manuscripts carry what we now think were false attributions. Similarly, a decision to photograph specimens of a certain style of script will depend on the palaeographer’s sense of the boundaries between one style and another, which might not be shared with other people. Selecting what to photograph will itself be determined by, and will later determine further, analysis.
One possible model for considering how this works is the playlist. The playlist and related forms of listing digital materials, by ticking or clicking on links that one likes on Facebook or Twitter, or by compiling a page on Tumblr have been described as a tools of ‘digital self-fashioning’ or parts of ‘identity formation’. Such digital curating is not entirely random – as the internet at large in some ways feels – because, as Jim Collins has put it, while ‘The database as file depot may be nonnarrative, [. . .] our negotiation of it is endlessly narrativized as we download files into our personal archives, at which point the files are transformed into curated articulations of personal taste – you are your playlist, not your database.’ Taking pictures, even more strongly than gathering pictures taken by others, involves building an argument, developing a set of ideas, establishing a corpus which expresses one’s sense of priorities in research – what one should look at – and range of knowledge – how far one has looked.
By the selection of a ‘playlist’ of multiple images, DIY photography can give the palaeographer a wide breadth of vision. This is allied to the ways in which DIY photography supplements one’s access to a library or archive and allows the consultation of images from more than one repository – repositories often far apart – in quick succession. As such, it helps to form a corpus of materials that can be more readily compared, analysed and remembered. One need no longer compare manuscripts weeks and hundreds of miles apart; one can photograph them on that schedule but then compare them together. In a subject such as palaeography, which analyses visual features such as styles of script, that is a crucial aid to visual memory and recognition. For instance, one by product of palaeography’s study of the history of script is the identification of particular scribes whose work recurs in more than one manuscript. DIY photography makes easier the compilation and analysis of a corpus of images from which to make and prove such identifications. One example of this was the study of the manuscripts containing the handwritten notes of John Stow, photographed by Alexandra Gillespie and her research assistants in libraries all over the UK and USA. The assembling of a corpus might also allow study of script beyond the biographical frame but along the longue durée, perhaps even, if one could work out measurable characteristics, in quantitative studies.
Yet there is always a question whether the widening of perspective, as one compares manuscripts, might also distort what one sees, say by losing detail. For instance, if I gather photographs of the same technique employed across dozens of manuscripts, to compare the techniques and rationale across many books over centuries (as I have done, for a forthcoming study, of extra leaves sewn into books), I can make sure that those techniques are comparable or alike with more accuracy than if I had compared my descriptions of them in words. But when I review all these photographs as a set, far from the manuscripts from which each came, am I not decontextualizing each example? And does not my curated set of photographs treat as one coherent phenomenon something that early readers of manuscripts normally encountered singly, rarely? Selecting and curating images create a new object for analysis – the photographic archive – that differs from the multiple objects from which the images came.
Or, conversely, can photography prevent decontextualization? For instance, a scholar who transcribes texts from manuscripts must make decisions about interpreting each mark. Nothing can be left untranscribed as uncertain, if a complete text is to be offered for analysis; and some of those uncertainties must be resolved, in the end, by educated guesswork, often fast guesswork if there is a lot of transcription to get through. Being able to take photographs of the manuscript one is transcribing, allows those uncertainties to be analysed again in context. And details of the text that might be mysterious in a typed transcription, as it were out of context, become less so when they can be seen on a photograph of the page. For instance, scribes often make mistakes at line-breaks, when the eye and pen must dart back and forth across the page. A scribal error baffling in a transcription might be more explicable when one can check a photograph and see that the mystery occurs at the end of the page’s line.
Photography can also allow us to look at context in even close detail. For perhaps what is most striking in the ways it changes what palaeographers can do is its addition of new levels of depth. Some photographic lenses can see things the eye cannot. That is not always the case; many people rely on telephones and tablets for photographing manuscripts, and the results are in some cases less clear than vision in person. But the best images even on commercially available cameras are often impressive in the depth to which they can be blown up. This promises a transformation what we can describe. For instance, in the history of script conventionally, the analysis has been of the finished letterforms on the page; the penstrokes which produced them, which would take us closer, phenomenologically, to the experience of the scribe writing – the movement of the pen or its flow or ‘ductus’ – have seemed invisible or, if suspected, difficult to illustrate, analyse and describe. But being able to photograph specimens makes it easier to zoom in and see more clearly and in detail how strokes form letters – something that earlier scholars overlooked or thought impossible to look for. (It might even prove more practicable for a scholar to take such photographs for herself than to direct a photographic studio in photographing single letterforms on the page.) For me, this means that the basis on which palaeographers work can shift from art history or biography. And there might be other ways in which DIY photography might alter the level at which we describe manuscripts.
There are still problems for such analysis though. It is not clear that the naked eye, looking at photographs or the real thing, can really observe strokes overlapping and shifts in ink colour with complete confidence. Colour perception and spatial awareness are not trustworthy. Technology is also faulty: cameras do not record colour entirely well. (Some sort of spectrographic analysis would be safer – but impractical letter by letter.) Moreover, there are theoretical problems in such analysis. When we zoom into a detail of a letter are we looking too closely? We can look at the strokes of the pen in detail and at a slowness that the scribe working, or the early readers with the naked eye – and by candlelight, and with other business to be getting on with – could not. So is the palaeographer who analyses penstrokes on a blown-up photograph seeing the manuscript in a way that its makers and earliest users could not? The depth allowed by DIY photography might be an anachronistic mode of seeing.
Yet a palaeographer – like an historian, an anthropologist, a psychologist – is in a sense somebody with the task of looking at and thinking about things more closely than people ordinarily do day-to-day: to take a bigger perspective, to look in depth, to reveal what is latent. And some of the details which cameras can see and the eye cannot are valuable for that reason. A digital camera of a commercially available sort and software included in most ‘office’ packages can together reveal things previously unseen. For instance, text erased by scribes or readers is near invisible and usually illegible; but a digital photograph ‘photoshopped’ to sharpen its contrast or dim its brightness often reveals more of the text than the eye could see. Or the paper used by scribes of the fourteenth and fifteenth centuries had, embedded into its fibres, faint indentations from the papermaker’s mould, called watermarks; such watermarks suggest where and when they paper – and so perhaps the book – was made; but these marks are very difficult to see. If one photographs the page, as though while turning it, so that ordinary daylight in the reading room shines through from the other side, then that image, ‘photoshopped’ for contrast again, will often reveal more of the watermark. It is unlikely that any scribe intended readers to see the words erased or the watermarks of the paper; but they are things that palaeographers, analysing the scribe’s work, rightly want to see – and that DIY photography is often sufficient to show.
Such images of erasures and watermarks taken by a researcher on an everyday digital camera are not usually suitable quality for publication, and libraries often explicitly deny permission to scholars to reproduce the photographs they have taken. (Daryl Green and Martin Poulter question the legality of such refusal of permissions elsewhere in this book.) But they usually are of sufficient quality for display on PowerPoint in lectures and talks. Showing one’s results to others is something else that DIY photography can do. DIY photographs, of things visible or not, can be valuable for showing colleagues unable to see the manuscript itself what is found there; they empower more people to make critical assessments of scholarly arguments.
There are, though, questions about the increasing proliferation of photographic images in academic communication. (There are practical problems too: a slide show for a conference can employ dozens of photographs, which would be too expensive and take too much space for a publication in print.) Photography has long been recognized to carry a kind of truth claim: as Susan Sontag put it, ‘Photographs furnish evidence. Something we hear about, but doubt, seems proven when we’re shown a photograph of it.’ But as consumers of photographs were are also wisely suspicious: we understand, especially in the news media, that the choice of images and their perspectives distort facts as well as reveal them. So the reliance on photographs for evidence brings problems for palaeography as for other humanities. Should analysis depend on visual evidence? Scholarship involves description, and photograph might serve that purpose, but does that then make the scholar’s descriptive vocabulary go rusty? And scholarship involves argument too, not just show-and-tell. To show a photograph of a feature of handwriting at different times does not explain its changes across time or contextualize it in its social function. Photography cannot supplant analysis in words.
And the temptation to photograph might redirect our attention away from things which we can analyse to things which we can show. There is an obvious temptation to seek out the most photogenic manuscripts. It often seems that the photographs which manuscript scholars share most on social media – where medieval manuscripts have a lively presence – are the most lovely: painted illustrations, decorated marginal drawings, often of comic or cute sorts. To judge from Twitter, blogs and some parts of Tumblr and Flickr, one could be mistaken for thinking that all MSS include lovely illuminations, often of animals, and pages damaged in unusual ways. And within academic publication, the use of photography might be warping attention. Finding something striking to photograph might well direct the scholar’s attention to the lovely but also to the ugly, the abnormal, the quirky; and having such an image or corpus of such images on file might refocus the agenda of research. By far the majority of pages of the majority of manuscripts consist of text, with minimal decoration beyond that which helps to articulate that text (such as chapter titles) Digital media alongside their many merits are often thought somehow to discourage reading long passages of text, with the temptation to click to the link on the internet, or to scroll or click through non-networked images, such as a collection of photographs. If reading is hard on screen of typography, is the reading of outdated languages and scripts on screen harder? While DIY photography might make the transcription of texts easier and more accurate (as was noted above), it might also direct attention away from text towards things which are more photogenic and more readily taken in at a glance.
Nor perhaps, should we try to catalogue everything in images till we have merely duplicated the manuscripts. Susan Sontag said that ‘the most grandiose result of the photographic enterprise is to give us the sense that we can hold the whole world in our heads’. I confess that I feel this wish to capture and hold everything some days when I start clicking in the reading room – the desire to hold things in memory and understanding transformed into a desire to photograph them. But photographing a book is not the same as reading or knowing the thing itself. Reading involves not only sight but though, understanding not only description but reflection.
But while photography cannot fully capture the experience of manuscripts themselves, it can produce a new mode of studying them. It is not a surrogate for books but it is another method for studying them. All methods of study involve diverge from the things they study, whether through historical hindsight, through selective quotation, through theoretical metalanguage, through analytical models brought to bear. We need not be fretful about that distance: it is unavoidable, and our methods of research also allow us to see new things, even as they obscure other things. But we might recognize that the DIY photography of manuscripts does not create ‘surrogates’ for them, like digitized collections online or like the microfilms which libraries offer when manuscripts are too valuable to consult; it creates new methods of seeing, selecting, organizing and thinking comparable to, but different from, other scholarly practices of description and analysis.
 For a brief introduction, see my entry ‘Palaeography’ in Sian Echard and Robert Rouse, ed., Blackwell Encyclopaedia of Medieval English Literature (New York: Wiley-Blackwell, 2017), 0000-0000.
 E.g. Peter J. Lucas, ‘A Testimonye of Verye Ancient Tyme? Some Manuscript Models for the Parkerian Anglo-Saxon Type-Designs’, in Of the Making of Books: Medieval Manuscripts, their Scribes and Readers: Essays presented to M.B. Parkes, ed. P.R. Robinson and R. Zim (Aldershot: Scolar, 1997), 147-88.
 Ludwig Traube, Zur Paläographie und handschriftenkunde, ed. Paul Lehmann (Munich: Beck, 1909), p. 57.
 An example would be Early Manuscripts at Oxford, http://image.ox.ac.uk/, created in 2001.
 A.S.G. Edwards, ‘Back to the Real?’, The Times Literary Supplement, 7 June 2013.
 For details of this project, see the information provided on The Digital Walters: http://www.thedigitalwalters.org/03_ReadMe.html.
 Giulio Menna, ‘Manuscript Photogeny – When Manuscripts Smile to the Camera’, on the blog Medieval Fragments, 13 July 2012, https://medievalfragments.wordpress.com/2012/07/13/manuscript-photogeny-when-manuscripts-smile-to-the-camera/.
 I thank Martin Kauffmann for allowing me to share his observation. He has also drawn my attention to the concerns for conservation which this different kind of handling brings.
 Though for suggestions how digital facsimiles could offer materials akin to critical editions, see Christoph Flüeler, ‘Digital Manuscripts as Critical Edition’, a talk given at the International Medieval Congress, Kalamazoo, May 2015, archived online at https://schoenberginstitute.org/2015/06/30/digital-manuscripts-as-critical-edition/.
 For related points about the use of photography, however done, to aid the verifying of editions, see Elena Pierazzo, ‘Digital Genetic Editions: The Encoding of Time in Manuscript Transcription’, in Marilyn Deegan and Kathryn Sutherland, ed., Text Editing, Print and the Digital World (Farnham: Ashgate, 2009), 169-186 (169-70).
 Daniel Wakelin, Scribal Correction and Literary Craft: English Manuscripts 1375-1510 (Cambridge: Cambridge University Press, 2014), 14.
 Wakelin, Scribal Correction, 45-63, 271-273.
 On other recent trends to look away from the text, see Heather Love, ‘Close but not Deep: Literary Ethics and the Descriptive Turn’, New Literary History, 41 (2016), 371-391 (373-374, 382).
 On such surface description, see most recently Sharon Marcus, Heather Love and Stephen Best, ‘Building a Better Description’, Representations, 135 (2016), 1-21.
 Edwards, ‘Back to the Real?’
 Michael Camille, ‘The Très Riches Heures: An Illuminated Manuscript in the Age of Mechanical Reproduction’, Critical Inquiry, 17 (1990), 72-104 (72-73, 104).
 Sontag, On Photography, p. 119.
 Sontag, On Photography, respectively p. 3, p. 28.
 Sontag, On Photography, pp. 6-7.
 Jim Collins, ‘Reading, in A Digital Archive of One’s Own’, PMLA, 128 (2013), 207-212 (209). See related points, in the same special issue of this journal, by Michael Cobb, ‘A Little Like Reading: Preference, Facebook and Overwhelmed Interpretations’, PMLA, 128 (2013), 201-6 (202), and Lisa Nakamura, ‘“Words with Friends”: Socially Networked Reading on Goodreads’, PMLA, 128 (2013), 238-43 (240). For some scepticism about ‘curation’, see Carina Chocano, ‘Pinterest, Tumblr and the Trouble with “Curation”’, New York Times Magazine, 20 July 2012.
 Alexandra Gillespie, Old Books, New Science, online at https://oldbooksnewscience.library.utoronto.ca/about. The images were mounted online using Omeka.
 Albert Derolez, The Palaeography of Gothic Manuscript Books (Cambridge: Cambridge University Press, 2003), 6-7.
 The website http://retroreveal.org allows users to upload photographs of erased text for such modifications.
 Sontag, On Photography, p.5.
 For expressions of concern about this distorting effect, with reference to historical sources from all periods, see Sarah Werner, ‘It’s history, not a viral feed’, on her blog Wynken de Worde, 26 January 2014, http://sarahwerner.net/blog/2014/01/its-history-not-a-viral-feed/, and Rebecca Onion, ‘Snapshots of History’, Slate, 5 February 2014, http://www.slate.com/articles/life/history/2014/02/_historyinpics_historicalpics_history_pics_why_the_wildly_popular_twitter.html
 For some concerns about the challenges of reading from electronic images, see Naomi Baron, Words Onscreen (Oxford: Oxford University Press, 2015), and David Mikics, Slow Reading in a Hurried Age (Cambridge, MA: Harvard University Press, 2013). That need not be the way. Contrast Eleanor Parker’s prizewinning blog on medieval topics which uses long quotations, often in the original languages: A Clerk of Oxford, http://aclerkofoxford.blogspot.co.uk/.
 Sontag, On Photography, p.3.