| Computers & Paper. Printers > General > Need for Print |  | Navigation Icons Guide
|
In a world filled with computer screens why do we have printers at all?
Today's idea of a paperless office has a long history - it can be traced back to Vannevar Bush's 1945 idea for the "memex" and the first commercial computers like LEO.
Paperless working might seem to be a practical objective. Holding information in computer compatible forms is a rapidly growing trend - catalogues and reference material on CD and sales brochures and support manuals on the web for instance. Once the disks and network connections are in place the costs of storing and distributing documents are trivial. If the document can be read on the screen there is no apparent cost at all.
Today's hard disks and networks designed for audio/visual material can find and shift enormous amounts of text. The disk in a commonplace desktop computer can store more text than all the books in a city library (although people rarely use computers this way). Broadband Internet connections can deliver a screenfull of information in a fraction of a second and a bookfull in under a minute. So why do people still use paper?
Inertia is clearly one reason. For instance:
| Most people are more used to books than computers. | |
| Books, magazines and newspapers are copyright material and paper media ensure payment for the author and publisher, disk and network delivery do not. | |
| Business and government transactions have to be paper by default and can only take electronic form where people agree. |
People haven't yet developed the political, economic or technical willpower to change things.
There is more to it than this because even where something originates in a computer and can be read on the screen people still prefer to print it out.
Paper is popular when people want to comprehend something. Computer users can often be observed peering at something on the screen for a second or so then hitting the "print" button. It costs money to print, but it seems to cost more effort to read on screen. Rising computer use might be expected to cut paper consumption but it seems to have dramatically increased it - perhaps doubling the total used. (Rising prosperity might also be a factor)
The rising tide of paper is galling to office managers who tend to have an anti-clutter and anti-paper attitude. Paper certainly poses problems - it is fairly expensive to print and then very much more expensive to file, so maintaining a computer and paperwork can easily mean duplication of effort. There is a feeling that computers were meant to eliminate paper, not increase it.
Several explanations of what is happening are likely:
Paper might be residual - custom and practice rolling on but dying out. Ubiquitous computers are relatively recent. Paper remains the default standard for exchanging information like business forms. Computers only have a few de-facto standards like Word "DOC", Excel "XLS" and Adobe "PDF". Newer XML standards still seem to be struggling to create order. Computerisation is a big cultural change and these things take time.
People might be clinging to old ways. The people at the top in many organisations have been there longer than the desktop computers and they might prefer the old ways. Lots of computer projects are stymied by lack of management understanding - which the managers blame on nerds spouting jargon like "XML".
Computer presentation of information on screens may be acceptable for clerical tasks but may not be adequate for "knowledge workers" who tend to use much more information losely organised in heaps around them. Today's screen metaphors of desktop and folders are nowhere near matching the 3D instant access to grasp and read piles of material that paper gives.
Computers aren't working as they should. Computers may have become easier to use but they are more difficult to program. Package systems generally don't fit what an organisation wants to do - but can be expensive to customise. Most organisations have several incompatible systems - again XML is supposed to be a bridge. Systems often don't reach the far corners of an operation - the hospital ward, warehouse and salesman in the field - or if they do current hardware isn't very suitable. Tasks are often part paper and part computer with bridging technologies like bar-codes and OCR forms.
Paper may be easier to work with - at least for some jobs. Tasks that require people to get an overview and detail at the same time like maps, satellite images or engineering diagrams for instance. This sort of work needs a high resolution display and no screen comes near matching paper. With a paper catalogue people can often flick straight to the right page. Screen data can be changed instantaneously - but scroll, zoom or page-forward actions are often visually disconcerting. Team work by several people may be easier with paper - the marginal notes, crossings out and underlining on paper may be far more helpful than coloured comments on a screen or a string of e-mails.
Comprehension may be reduced when dealing with screens. Screen design involves lots of assumptions about how human sight and eading work. Test evidence suggests that people's comprehension and retention of information shown on screen is significantly lower than for material on paper. Young people often like screeens and think they are "the future" but reading comprehension tests still show screens working less well than paper. What screens get wrong is uncertain but innadequate resolution seems a possibility.
Computers don't work well with paper. Printers turn computer data into printed pages readily enough but scanners are rather poor at taking information back in. Optical Character Recognition (OCR) is often better than 95% accurate handling typescript - but accuracy drops away with handwriting - and fails altogether with quite readable marginal scrawls. Information is in paper form will often rely on someone finding the computer file and re-keying it - so computer records are often out of date.
The aim in this section is to outline and quantify some of the problems.
Punch Cards
The UK census was like all others in that it had to be manual, there was no alternative. The US Census Bureau had a rapidly growing population and a duty to enumerate the population so as to determine each state's number of Federal Representatives in Congress. The 1880 census had taken nearly 1,500 people 7 years to calculate and had found a 30% growth in population - so figures were clearly obsolete before they were available. Since the population was still growing rapidly the results of the 1890 census threatened to take intollerably long.
Herman Hollerith at MIT has the idea that holes puched in cards like those used to identify class and age on railway tickets might be electrically ead, counted and sorted. John Shaw Billings of the Census Bureau urged him to build some tabulating machines and with this equipment the 1890 census was completed in two and a half years and a total population of 62,622,250 announced in just six weeks - to widepread disbelief. Hollerith founded a company to make tabulators which merged with others to become IBM.
Tabulating machines proved popular with other census bureaus and with ailroads and insurance companies. The British Tabulating Machine Company manufactured and sold Hollerith -based machines under license in Britain and the Empire (excluding Canada) but there were modifications to cope with non-decimal currency. BTM merged with Powers-Samas, another puch card company in 1959 to become ICT - after further merges it became ICL.
Norwegian engineer Fredrik Rosing Bull developed his own line of punch-card apparatus which founded Bull- first as a Swiss and later a French-based company. The US Census Bureau felt Hollerith's charges were too high and in 1907 hired James Powers to make them compatible machines avoiding the Hollerith patents. Powers founded his own company in 1911 and it sold punches, a sorter that was easier to use than Hollerith's and a printing tabulator developed by W.W. Lasker. Powers later merged with Remington Typewriter and Rand Cardex to form Remington -Rand. 
Punch-card appatus became one foundation of the computer industry but others were adding machines eg Burroughs. Remington-Rand and Cash Registers - eg NCR.
It doesn't take much aquaintance with paper to develop a dislike for paperwork.
| Tedium is a problem. A lot of tasks in government, business and finance are inhumanly repetitious so it is obvious a machine should do them. Payroll and account balances were early candidates for automation. Half of the staff in the "Prudential" office of 1880 were transcription clerks simply making copperplate copies of documents. Nearly a century later the Xerox machine replaced this job. |
| Legibility is definitely an issue. Handwritten paper records tend to be erratically written and there is an unfortunate tendency for people who write most to write illegibly. For a century, typewriters were used to give legibility to inter-office communication - at the cost that everything was delayed. |
| Scale can be a problem with paper records. Any business that keeps records finds it needs cabinets and shelves full of paper. Paper organisation can be idiosynchratic to eflect the nature of the job or the personality of the clerks - and it often is. Different organisations paper systems can be completely incompatible. |
| Loss of records is a definite problem. Most paper system lose some records - they are misfiled or accidentally destroyed. Some estimates suggest that 8% of paper records go missing. |
| Comprehension is another problem. Paper filing systems of any complexity tend to have all sorts of funny little practices - writing files in use in a ledger then ticking them off when returned, indexes that have to be updated and so forth. Procedures might be written in a quality control manual - another piece of paperwork. The juniors in a business often learn by a kind of osmosis - but it isn't a wonder files go missing. |
| Cooperation with paper files can be both good and bad. Good because if someone ammends a file everyone else can see it- even just a note in a margin. Bad because once a file is in use everyone else waits until it's returned. |
| Costs are a definite issue. Paper sheets cost very little but with susbstantial amounts of printing, filing cabinets and floorspace the cost of a few thousand files become noticeable. Labour is the real problem. It can be difficult to put a figure on the number of people involved but anything up to half of the population are primarily involved in administration. Farming, fishing, factory, construction, hotels, catering, and health are not primarily intended to be "paperwork" tasks. Civil service, managers, clerks and accountants primarily are. Teachers, scientists and engineers tend to have a lot of paper but only some of it is administration. Nearly everyone complains about paperwork. |
Back in the 1960s an executive of one bank estimated that, without computers, it would shortly take the population of the country just to run the check clearing system. Work expands to fail the time available. Administrative procedures do the same.
Computers were invented in 1945 and quickly inspired ideas of business automation. In Britain "Lyons Electronic Organisation" (LEO) was established before 1950 aiming to automate the bakery and café chain. Similar ideas in the US meant IBM and Univac's early computers proved unexpectedly popular.
Computers are often seen as an altenative to paper. A popular management attitude has been that getting rid of all those untidy paper files
| Tedious repetitive tasks can be computerised |
Information tends to start life in electronic form: it is typed, or aquired through instruments ranging in complexity from microswitches to web-cameras. One pinnacle of data production is CERN's new particle accelerator will produce a gigabyte of data per second - not something to commit to paper (for comparrison its about a whole UK phone directory per second).
Newspaper stories, books, technical reports and invoices are all typed into a computer - they could be instantly distributed on the Internaet - but they are printed on paper. Businesses find that people print e-mail, memos and manuals that they could have read on the screen.
The puzzle is why printed paper remains the preferred medium for so much information?
There is very little question that paper is still preferred. Newsagents shelves are full of magazines, not CDs or ebooks (CDs are stuck to the front of magazines). In bookshops the computer section is often one of the biggest - some have a CD with their contents. Printing paper consumption in the "advanced" economies roughly doubled as the microcomputer became popular. Being stuck without a printer can be very frustrating.
Superficially it seems obvious that people should be able to work from screens and that there could be a big advantage if they did. Not only might all those forests go to better uses, but information could be less costly and more up to date if it wasn't for printing and distribution. There would be all sorts of side benefits as well. Computer literacy would rise. Errors and omissions in computer records would be more likely to be corrected. At the moment there are millions of budding authors who can't get published - their material could be as readily available as any other. Blind and partially sighted people could get braille, audio and big-print versions easily. Organisations could eliminate the costs of print cartridges and the effort of printer support. Information would be available and could be corrected at its point of use - on the hostpital ward, in the warehouse, on the policemans beat. Perhaps most importantly the web and similar techniques can democratise information, making things readily available and findable in a way paper can't match.
Attempts to go paperless generally haven't worked very well. A high profile failure was Chiat-Day, a US advertising agency that attempted paperless working in the early 1990s. Staff wound up working from the boots of their cars and one woman used a childs trolley to move her papers around. Ultimately a rival took the firm over and ended the experiment.
HCI, a private hospital in Clydebank tried a similar experiment, with a paperless medical record system. Unfortunately the hospital failed - perhaps more through bad commercial judgement than because of the systems. Paperless working is a reasonable objective in medicine, because if things are first handwritten then later transcribed the patients condition changes faster than the computer record - leaving it with roles in statistics and accountancy but not much else.
//
People intuitively feel uncomfortable when they have to read anything detailed from a screen. Research has uncovered something like a 20% eduction in comprehension and retention for material read from a screen. Unfortunately - perhaps becuse people have read the reports on screen - the problem has not been much noticed.
Obvious technical problems with electronc information were solved more than a decade ago.
Computers have become ubiquitous - and if they aren't yet as cheap as TVs they probably soon will be. Almost every desk and more than half of households in Western countries have a computer. Most of the households that don't have a computer have TV and video - they just don't want a computer because they don't think it offers them much of interest. Computers work as digital TV and recorder so the distinction is dissapearing and a combination of features and scale economies will take over.
Desks aren't always comfortable or convenient places to read so portability is important.
Portable screens are readily available. Notebook PCs, pocket sized PDAs and the ebook all make LCD screens easy to use anywhere. Notebook PCs have been succesful enough although the tablet form factor best suited for easy reading has never really caught on. The PDA market was distinctly slow for several years - arguably its role as an organiser was partly usurped by mobile phones. Keyboards are a problem with small portable devices. Ebooks just don't seem to have caught on - unless you count the acrobat "pdf" sales brochures and manuals that people normally print out. Perhaps people don't like the small screens - or the price of material/
Battery power is a definite limitation. Colour LCD screens need backlighting and that needs a fairly large battery - but most PDAs and notebooks will last for two to four hours. Actually most people in the UK and US are within range of a powerpoint most of the time so echarging isn't too problematic. (Power is a problem in many countries)
Network connection issues aren't entirely solved. A billion people worldwide have Internet Access. WiFi and cellphone connections are patchy and / or rather expensive but they aren't really essential in a eader. Arguably people know roughly what they might read - to within a few dozen books. Several hundred books fit easily in a gigabyte pen drive, so it is easy to carry a huge amount of information without needing a network.
With the obvious technical limitations to reading on screen apparently solved why do people still need print?
People seem to intuitively feel that there is something not quite right about reading from a screen. What they don't commonly do is to quantify how little information a screen can show and how disruptive this limit can be to human comprehension. The current generation of screens are fine for low quality pictures and records - name, address and booking date stuff. Most screens are rather poor for anything more complicated - perhaps even for text as long as a one page letter. Using screens for more complex material could hobble human recollection and understanding.
The problem with screens is quite well documented in educational and psychological literature.
Screen resolution seems to be the core problem. Most people ealise that screen displays are made from "pixels". A pixel or picture element is a dot used to make up an image. 
It takes a lot of pixels to reliably fool the human eye into seeing an image. There just aren't enough pixels on a screen to support ordinary human reading and they are in the wrong place. As a way to deal with this screen layout is unnatractive.
Counting the pixels gives a simple way to look at the problem.
A4 paper sheets work out at about 8 million pixels ( 2,400 x 3,300 pixels - allowing for margins and printing at a relatively low 300 dpi. Resolutions are usually given in dots per inch; 300 dpi is 12 dots per mm)
The typical desktop computer screen resolution is 1.3 million pixels or less ( 1280 x 1024 pixels. Most notebooks have less space with 1024 x 768 being typical. Mobile phone displays and PDAs are truly cramped, with 320 x 240 a maximum.)
A single sheet of paper shows at least 6 times as much information as a screen and the paired pages of an open magazine show 12 times as much. Furthermore paper has very good contrast, even in sunlight. Most people like sunlight but backlit LCD screens become almost unreadable.
Difference in resolution has two immediate effects:
| more information contextual information can be seen on the printed page - and on those nearby in the heaps people commonly make. |
| individual characters are much better formed - which may be less obvious but more important |
A conventional model of reading is that it starts at the top left and works linearly through to bottom right. If this were strictly what people do they would be comfortable reading scrolled messages on small displays like pagers; actually they are not. A reader's eyes dart about, efreshing their memory and putting names and numbers in context. Reading a word at a time without contextual information is rather difficult. (Try making this web-page so small that you can only see one word at a time and then scrolling). Behaviour changes with the document of course. Novels are usually ead from beginning to end. Textbooks, catalogues and reports are flicked through and skimmed - very different behaviour.
People seem to find that something like an A4 double-spread of paper gives them just about enough contextual information. If people were happy with less then A5 or A6 paper would be more common - A4 is actually a bit too big to be convenient, which is why paperback novells are smaller. The layout and look of conventional printing suggests what people will need if screens are to work.
Print preview at a push can show an approximation of two pages side by side on fairly common 1280 x 1024 SXGA screen. It does this by compromising in another and potentially corrosive way - reducing the number of pixels in a character. Reducing the pixels per character leaves text perfectly comprehensible for anyone who thinks about it. Unfortunately when people are to read and understand text the aim is that they shouldn't have to think about the material at all. Text that is too small and badly formed will leave people struggling with pattern recognition and unable to handle the meaning. Proof reading documents gives some evidence of what happens. Type a document and check through it on the screen - there are usually various typos and mistakes. Print the document out and read it again - most people find mistakes which they never spotted on the screen. Things which make people uncomfortable get in the way of understanding. If people are struggling with the mere act of reading on a screen comprehension goes down and mistakes rise.
What Else?
Better Screens.
Bigger, better screens should solve the problem. The market might seem to be moving that way. Screen manufacturers are slowly working their way to the 2048 x 1536 QXGA standard but for most users that is a long way off. Arguably it is nowhere near good enough.
No easy fix is going to make screens suddenly much better although a few quick-fixes are possible. Some of those are outlined below.
Bigger screens do not necessarily address the problem - all that happens on scaling up to a 42 inch TV is that the pixel-grid of an ordinary TV gets bigger and coarser - ordinary TV resolution is about 800 x 600. There is no more information on a large screen.
HDTV may help by pushing screen resolutions a bit higher. The HDTV screen is defined as either 720 or 1080 lines at 16:9 aspect ration so it works out as 1280 x 720 (poorer than today's computer monitors) or 1920 x 1080 (at 2 million pixels arguably a bit better). Screen manufacturers will have scale economies for HDTV. There are a couple of problems:
| LCD TV manufacturers will probably aim for big screens - so they may not be much good for people working with computers. |
| Wide screens made popular by 16:9 atio TV don't help much - they are the wrong shape for text. |
Resolution seems to be the key problem and current screens don't come near matching what is needed. Worse, manufacturers don't seem to have much incentive to aim higher. The current market focus is on switching from CRT to LCD monitors which often means a reduction rather than an increase in resolution.
Some of the CRT screens on the market are capable of higher esolutions up to 1600 x 1200. This is not a dramatic improvement (about a fifth more data). Users are shifting to LCD screens which genuinely do have much better contrast and sharpness so they are easier to read. This doesn't fix the problem - recent studies of comprehension and retention that used LCD screens still show the effect. Unfortunately LCD screens tend to be look good at the cost of resolution and have often taken a step back to the old 1024 x 768 standard.
Bigger and higher resolution LCD screens seem to be difficult to make so although they are available they are expensive. QXGA 2048 x 1563 screens are very expensive and tend to be used with Sun workstations and Apple Macs.
No screens currently come anywhere near giving the visual experience of the printed page. Multiheaded PCs with two or more monitors are a possibility. This gets more information onto the desktop. Multiheaded computers do at least allow a user to have several documents open at the same time but its far from being a comfortable reading experience.
Paper's Other Talents.
Poor screens are a problem but there are others. Paper messages like bills and invoices usually have some legal standing that may not apply so easily to electronic messages. Paper is a saleable item with some inherent protection of copyright. Paper tends to last, but computer media generally prove ephemeral. People like bright, brief things on screens but if they have to concentrate then paper seems better.
Put It On Paper.
Reading ordinary text from a screen seems to reduce people's ability to comprehend and retain information. Computer printing raises some questions.
Putting information on paper is swift and easy but it tends to be a bit of an afterthought. Paper is outside the workflow
A huge range of print technologies is available. Mass production tends to be cheaper in print as in most things, so a million copies of a catalog or newspaper from a web-offset machine are low cost and a thousand copies of a book or manual cost rather more. Until the 1960s when photocopiers became common short print runs were costly and difficult. Laser printers are computer driven photocopiers. Recently the cost of printers has fallen so low that just about everyone can have one. Printing a ten page report with a few photographs and illustrations generally takes a minute or so on any recent printer so time wasting isn't usually a problem, cost might be.
Costs of operating small printers can cumulatively be quite high. Average business use is that someone prints 30 pages of text and a few pictures at a cost of about £1 per day. An organisation with a thousand staff can easily use a third of a million pounds worth of print consumables in a year - quite apart from any big billing runs or publicity it might do.
Individual printers have the advantage that the people concerned can take responsibility for what they are doing - cartridge use can be monitored. Little printers tend to have little cartridges, so they cost more to operate.
Bigger workgroup printers usually offer much lower cost per page. This is reinforced by features like duplex print, collation and stapling. Staff can, however, treat the office printer as a perk and school essays, holiday snaps and pickle jar labels might all be seen as fair use. There are ways of monitoring use - (spooling files etc) and restricting what can print (permissions etc) but they impose an administrative burden.
Print subsytems on computer don't necessarily do sensible things. It is reasonable to print a one page memo on the locla printer. A 200 page book is more difficult, it takes longer and the cost on a small printer is likely to be high. Ideally the print susbystem should detect the nature of the job and suggest a big central printer. //// the binding will be right.
Information on paper also becomes a sort of orphan. People are often seen correcting paper copies but all too often the mistake their proofreading found isn't corrected in the computer. Should paper copies be filed - it seems wasteful to do anything else but the costs of filing can be enormous? If paper is recycled will you have to pay to have it collected? Paper also aises security issues and can need shredding.
There is a definite suggestion that printing is environmentally wasteful. The UK's annual paper consumption is sometimes said to consume a forest the size of Wales - although the actual sources of pulp are about 60% ecycling supplemented by imported pulp from Scandinavia and North America.
Costs arising from not printing may be less detectable but worse.
e-books