When I talk about what I do, the index finger of my right hand automatically enacts a scrolling motion. The finger bends into a hook; it straightens; it repeats, a mid-air gesticulation. Each time the finger achieves its most acute position there is a twinge in the knuckle which repeats with the action. A feeling of numbness that crackles like static.
For six months now I have performed Quality Assurance for the journals archive of a large publishing organization. Content is being digitized, and with PDF images of journal pages comes a certain proportion of hand-typed content. The publisher’s supply company enters article titles, author names, abstracts, and footnotes manually, keying-in the information in code. The code produces a searchable page which the customer can view before buying content. When applied correctly, the code can be used to attach hyperlinks to citations in footnotes. These hyperlinks offer access to electronic versions of cited books or journal articles. They also offer access to other citations. The resulting product is a formidable archive, a reference network in which, for instance, publications cited in 1865 can be cross-referenced, generating an online page of related information. The code enables an outward motion. The code enables connection. Entered by hand, there is a possibility that both the code and the data content can be incrorect.
para2 correct incrorect as incorrect
Entered by hand, there is a possibility that both the code and the data can be incorrect. At times, keyed-in data is transcribed incorrectly. It can be misunderstood by the copier and coded wrongly. A journal citation might be captured as a book; text explaining the citation might be captured as part of an author’s given-names; special characters might not be properly inserted. The aim of the digitization process is to overcome the material and typographical frailties of the printed hard copy, preserving and encoding key information accurately in accordance with copyright laws. Yet despite being converted to a digital format, the printed copy retains a certain resilience. In a digitization process which works to convert hundreds of different journals to one online format, typography becomes an obstacle. Footnotes stacked in columns, two columns to a page, disconcert the hand used to typing long-line notes: lines which should not, merge. Shifting methods of tracking and kerning disorientate the copyist’s eye
para3 remove additional spaces between kerning and disorientate
Shifting methods of tracking and kerning disorientate the copyist’s eye. In the oldest journals ink is often faded and difficult to read, and the descending ∫ gives rise to pages of erroneous material. In some typefaces, to italicize is to make h seem b, e seem c.
para3 capture h, b, e and c in <italic>
to italicize is to make h seem b, e seem c. Copyright laws mean that original mistakes, too, must be preserved.
I work to erase the mistakes which betray the residues of the printed text. I detect these slips of hand and feed back corrections to be made. Once corrections are made and the code is functional the journals can be published, or made Live. The end product circulates: digitized, it takes on a new accessibility and becomes part of a wider conversation. Digitized, it acquires new value. I work with texts that are not Live.
para4 inverted commas (‘ ‘) around Live are missing, please insert in both cases
I work with texts that are not ‘Live’. Separated from their printed, material form, these texts do not yet operate according to our code, or fit the screen on which they will be viewed. They are curiously static, as though unable to acquire significance until digitization is complete. I facilitate their transition by proofreading the copied data, by making sure it fits, and that it will display correctly. It is about display. I do
para5 insert paragraph break between display. and I
I do not work with what the text is saying, I work with its surface. An arrangement behind a screen. The arrangement must be repeated for continuity’s sake. I repeat the arrangement.
The index finger of the right hand performs a scrolling motion; the eyes move over the text. We are asked not to go deeply into the content and what it is saying; we must know how to stop ourselves from reading. Knowledge of the English language means that I am able to detect mistakes almost automatically in a process half of grasping context and half of seeing shapes. I have developed the capacity to tell when a <given-name> or <surname> has been mis-spelled, and when a name is improbable. In Classics and Science journals, I read Latin and Greek by sight, following outlines and making up sounds. I cannot correct the text manually, I can only feed back. I have no access to the copied data and cannot change it. When corrections are made, I re-scan. At times corrections are missed. I repeat my feedback. I repeat my feedback until the copy is accurate and displays correctly. I write commands in short, spare phrases which include items of code. In the beginning I used the word please.
para6 please insert missing comma (,) between beginning and I
In the beginning, I used the word please. Please was soon eliminated, as was anything else that made the feedback needlessly long, such as the names of journals, which are shortened to acronyms. I check that the coded hyperlinks are operational. Occasionally, I check so many hyperlinks that the search engine assumes I am an automaton and suspends my connection.
It is about display and looking right. We are asked to complete a HSE workstation test to teach us how to use our VDU
para7 correct HSE and VDU as Health and Safety Executive and Visual Display Unit, as per PDF
We are asked to complete a Health and Safety Executive workstation test to teach us how to use our Visual Display Unit. I learn to arrange myself so that, in addition to a visual display of the words I scan, I carry out a number of self-regulating checks. Are feet flat on the floor without too much pressure from the seat on the backs of the legs? I cannot reach, and request a footrest. I catch myself crossing my legs. Place feet firmly on footrest without exerting too much pressure. Admonish self for craning neck for closer scan. Are forearms horizontal and eyes at roughly the same height as the VDU? I lift my wrists from the surface of the desk with wrist rests, typing, cushioned, to avoid unnecessary damage through contact. I measure my distance from retina to VDU with my arm’s length. I work correctly on the surface, but I am prone to fidgeting. Losing patience with repeated mistakes, I admit that sometimes I fall into reading, and writing things down for future reference. And I admit that there is a twinge in the index finger, tensed muscles in the neck and an ache in the shoulder and blade of the shoulder which makes connection with the small of the back. It is an outward motion that relays down the left leg and into the foot. The sensation crackles like static. My right foot twitches into a tap on the days I can’t sit properly. I shift to keep it still, re-align my seat again, scroll faster, or switch to Page Down. The HSE VDU test asks me if the air is comfortable.
When the hard copy of a text is converted to PDF, the final image is comprised of layers of information. On pages with text, an OCR (Optical Character Recognition) text layer is embedded in the PDF, and hidden so that the page retains its original look, but becomes searchable. On pages with pictures – plates bearing graphs, photographs of archaeological sites, old scientific drawings of light as it hits soap bubbles – there may be tonal layers in addition to the OCR layer. Occasionally the scanning process picks up visual noise not immediately evident in the original, black smudges which cling to letters. In some instances the OCR and original text image may not align correctly during the scanning process. The original image’s digital duplication is visible when keying quickly through the article: the edges of the page flick ragged black and white, a strange peripheral archaeology. Sometimes plates with pictures are missing a tonal layer, and appear faint and diminished on the screen, as if being seen from a distance. In these cases, the digitized image manifests multiple layers of hands and work, which must be eliminated before the image can reach publishable perfection.
I leave no trace on these images, hovering over them.
Birch wood splinters unpredictably when you cut against the grain. The solution is to cut deeper into the surface, gouging a new line which boldly transforms the splintering into a clean-edged furrow. It is also worth remembering that the pressure of the roller on the wood also forces down any splinters or snags, and illuminates areas which could be cut away further. In fact, once the woodcut is finished, the surface can be sanded without danger to the image. Sanding the surface of materials such as birch wood and linoleum is often necessary to remove tough outer layers which affect the materials’ absorption of ink. Sanding can be done to different degrees, and while some artists work to gain smooth blocks of colour, some prefer to incorporate the natural texture of the material into the image.
The delicate grain of the birch reminds me of static. Initially, I thought I had sanded enough to eliminate the fine strands of the grain, but, when printed, I was glad that the mistake had left its mark. I admire the residual evidence of the original matter, and of the crafting hand. A smudge left by cloth wiping away excess ink. Paper almost engraved, cut through by the pressure of too much card, wood-board and printing press. Planes of white which show where wood has been removed, and the smudge of red from where I slipped and split the surface of my hand. These mementos complement muscle memory.
In the application of ink, care must be taken to ensure that the ink transfers smoothly to the roller. Its hardened skin must be picked away from the top of the ink tin, and the ink accessed must have the consistency of molasses. Use a palette knife to smooth the ink onto a table top and pick out remaining pieces of skin; once the line of ink is smooth, begin rolling. The ink should be rolled repeatedly until it loses its orange-peel texture. If you are making a blend of inks, you must move carefully between two lines of colour, merging them. It helps if both inks have the same consistency. The connection of the two colours should produce a glowing haze where they meet. The sound of roller on ink on table is called a printer’s kiss. Ink must be wiped from surfaces with white spirit.
My final image was made under pressure of time. Digging rows of fields through wood became a repeated hacking, an unearthing. Turning the birch methodically, and pressing against the bench hook for better purchase, the shavings and dust clung to the wool of my jumper. Woodcut tool was pushed forward by lower palm of right hand, pushed down by index finger, held steady by thumb. Pushing prolonged and at pace, the crackles of static increased to a numbness and shoulder locked tight. A burst of red in my palm was blood returning, feeding the surface. Connection of telephone wires against the white of sky emerged from splinters and the black of ink rolled fast, spitting onto hands which fidget around the edges of proofing sheets, wary of leaving fingerprints. It is about display but it is also about the connecting of ink with paper. I used the largest and slowest press, spinning a wheel the size of myself, my whole body moving with its repeating rotations, and the numbness in the hands unleashed and building. Three prints pressed beneath wood and press and packing, almost cut through with the weight. Phone lines connect, black wires, black furrows of the field.
I wiped the ink from the tabletop with white spirit. It took about ten minutes to remove it all. Outside, as I rubbed ink from my neck and nose with crackling hands, the red bloom of a nosebleed.
To comment on an article in The Junket, please write to email@example.com; all comments will be considered for publication on the letters page of the subsequent issue.