E-Book Technology: Waiting for the "False Pretender"
Raymond Kurzweil proposed seven stages in the "life cycle of a technology." Libraries can use the technology life cycle concept to determine when to invest in newer technologies. Kurzweil gave no criteria for determining what stage a technology had achieved in the life cycle. This article will present a set of criteria to evaluate new technologies within the framework of the life cycle, demonstrating their applicability using Kurzweil's example of audio technology and then applying the criteria to current e-book technology.
Technology is one of the foundations of today's library. Our catalogs have become databases while the venerable cards are shipped to the recycler or used as scratch paper. We process and deliver interlibrary loans via the Web along with an increasing number of journals and indexes. This transition forces libraries to make decisions about a plethora of new technologies on a daily basis. As an electronic engineer until 1993, I am especially interested in the relationship of technology to my new profession.
A model of the life cycle of a technology was proposed by Raymond Kurzweil in 1992. This model can help libraries determine whether a technology is appropriate to adopt. In this article, I will amplify Kurzweil's technology life cycle model and apply this improved model to determine the current state of e-books within Kurzweil's framework.
Raymond Kurzweil, noted inventor and futurist, wrote "The Futurecast" column for Library Journal in the early 1990s. A three-part series in this column titled "The Future of Libraries" appeared from January through March of 1992. The first part of this series, "The Technology of the Book," listed seven stages in the life cycle of a technology:
5. False Pretenders
In the precursor stage, ideas about a new technology exist, but have not been implemented (e.g., DaVinci's helicopter). Invention gives the ideas concrete form. Development hones the technology into a practical form (e.g., automobile technology at the turn of the last century). Finally the technology reaches maturity and is practical and useful. Maturity can last for years, decades, or centuries depending on how well the technology meets the need for which it was invented.
As the mature technology ages, newer technologies arise to challenge it. If the newer technologies have some superior features but are not yet comparable in all facets, they may become false pretenders. Kurzweil defined this stage: "Here an upstart threatens to eclipse the older technology. Its enthusiasts prematurely predict victory. While providing some distinct benefits, the newer technology is found on reflection to be missing some key element of functionality or quality" (1992a). The false pretenders may coexist with the mature technology, but they will not supplant the mature technology.
A newer technology can supplant the mature technology only when most or all of its features are comparable, and it has some improved feature to compensate the user for the trouble and expense of switching technologies. When a newer technology supplants a mature technology, the older mature technology enters obsolescence, in which it coexists with the newer technology for approximately 5 percent to 10 percent of its mature lifespan. Finally a supplanted technology reaches antiquity and ceases to be produced or used.
Kurzweil's stages provide a useful framework for evaluating current and proposed technology. Investment in the precursor, invention, and development stages of a technology is not worthwhile because of their instability and lack of features. A false pretender, on the other hand, may have a long lifespan and be worth some investment. For example, audio cassette technology is a false pretender that has remained popular for thirty years. The majority of library investments, however, will be in mature technologies with proven quality and functionality.
Kurzweil did not provide criteria for determining which stage a particular technology had reached. After examining his examples in "The Future of Libraries" articles, I developed a set of eight criteria for determining how mature and newer technologies compare with each other.
With respect to the mature technology, the newer technology must have comparable:
3. initial cost
4. continuing cost
5. ease of use
In addition, the new technology must:
7. be standardized
8. have extra features
These criteria provide a nonarbitrary method for evaluating a technology. Kurzweil used audio technology as an example of his stages. The mature technology was the LP record, the false pretender was the audio cassette, and the newer technology that forced the LP into obsolescence was the audio CD. To show how these technologies compare using the eight criteria, see table 1, which shows these three technologies and also two other contemporary, competing technologies, eight-track and reel-to-reel tape. Comparing the three tape formats using the eight criteria demonstrates why cassettes became the false pretender.
Table 1. Comparison of Tape Formats and CD to LP Records Using the Eight Criteria
LP Reel-to-Reel Eight-Track Cassette CD Quality Very good audio Slightly lower Much lower Worst Better Durablity Fairly sturdy, lasts for years with gentle handling, deteriorates gradually Better Much less Better Much Better Initial Cost $50 or more Much greater Comperable Comperable Comperable Continuing Cost $10-$20 per album Comperable Comperable Comperable Comperable Ease of Use Very easy Less Better Better Comperable Features Random track selection, changer No No No Comperable Standardized Yes Yes Yes Yes Yes Extra Features Recordability, portibilty Recordability, portibilty Recordability, portibilty Recordability, portibilty
Quality for audio media can be determined from the frequency range and signal-to-noise ratio. For the three tape media, these are governed by the speed of the tape and the width of the audio track. Faster tape speed produces greater frequency range and lower noise. A wider audio track produces less noise. Reel-to-reel moved up to four times faster than eight-track and eight times faster than cassette; its audio track was also much wider, therefore it had the highest quality of all competing tape media. Eight-track came in a distant second in quality with cassette at the bottom. Later, the introduction of high bias tape and Dolby® noise reduction improved sound quality for all three tape technologies, but they never matched LP.
The unusual drive mechanism of eight-track tapes caused them to stretch, bind, jam, and break. Both cassettes and reel-to-reel are somewhat more durable than LPs since they are less fragile and not as subject to damage from heat. LPs and all tape media deteriorate gradually with use and wear.
The initial cost for audio is the player. Reel-to-reel players were generally professional-quality machines that cost much more than LP, eight-track, or cassette players. Also, stand-alone reel-to-reel players never became popular, so the player had to be hooked to an expensive stereo unit.
Cassettes and eight-tracks were comparable in media price to LPs. Reel-to-reel tapes were higher priced and little prerecorded media was available.
Ease of Use
Cassettes and eight-tracks were comparable or easier to use than LPs while reel-to-reel required tedious threading of the tape mechanism. Reel-to-reel tapes also required careful handling since the tape was not secured on the reel. When dropped, the tape would unravel and require careful untangling and hand rewinding.
Playing a specific selection on an LP was a simple matter of placing the needle in the space between songs. No tape media had the LP's random access feature. Also, record changers could play a sequence of albums without intervention. Changers were not possible with reel-to-reel and were not economical with cassettes or eight-tracks. Otherwise, the feature sets were comparable.
All tapes were standardized, but reel-to-reel tapes could be recorded at three different speeds while most players could only handle two of the three speeds without hardware modification.
All tape media had the added feature of recordability, although this was seldom used with eight-tracks. Cassettes and eight-tracks also added portability. Cassettes excelled at this. They could easily fit in a pocket and the players were soon miniaturized to pocket-sized. Eight-track tapes were significantly larger than cassette (about the size of a paperback book) and the players were boom box size.
Adding to their portability, cassettes and eight-tracks could be used in cars and other high vibration environments. It was portability, along with cost, size, and durability, that made cassette the false pretender. Yet cassette never supplanted LP because it did not have comparable sound quality and lacked the random track selection of mature LP technology.
Audio CD, on the other hand, matches the sound quality, cost, and features of LP with improved ease of use and durability. CD portability is almost as good as cassette and it is becoming increasingly easy to record them. CD quality and features made it the new mature audio technology, rendering LP obsolete.
Kurzweil stated in 1992 that electronic books were at the false pretender stage. To test this, table 2 shows the results of using the eight criteria for year 2000 technology e-books.
Table 2. Comparison of E-Books to Books Using the Eight Criteria
Print Book E-Books Quality Very easy to read Poor Durablity Five hundred years with proper handling Much less for hardware, media may quickly become obsolete Initial Cost None $200 or more Continuing Cost Low Comperable Ease of Use Very easy to use Less Features Quickly browse and scan Slow or impossible Standardized Yes No Extra Features Searchable text, hyperlinks, higher data desity, rapid updating
For both print books and e-books quality is determined by the display. The print book's display is far superior to either the Cathode Ray Tube (CRT 1) or Liquid Crystal Display (LCD 2) used for e-books. Display quality can be measured in terms of display resolution (in units of dots per inch - dpi) and contrast. Most books are printed at a resolution of 1,200 dpi. Commercially available video displays (whether CRT or LCD) have a maximum resolution of 100 dpi. The 1,200 dpi resolution of print is not twelve times the resolution video displays, but 144 times since a display is a two-dimensional surface and print is twelve times the resolution of a video display both horizontally and vertically (figure 1).
To illustrate this in more practical terms, figure 2 shows a magnified character in 10-point Times Roman font at print and video resolutions including the "enhanced" resolution of Microsoft ClearType used in some e-book readers.
Figure 1. Comparison of Dot Size for Print and Video
Figure 2. Comparison of Ten-Point Fonts in Print, MS ClearType and Video
Kurzweil recognized display technology problems in his article but underestimated their magnitude and the time it would take to make improvements. He predicted that video display resolution would catch up to print in three years (Kurzweil 1993), but display resolution remains stalled at 100 dpi. As Jacobo Valdés (developer of Clearview, a competitor of ClearType) put it, "Screen resolution hasn't improved much over the past decade. It is still abysmal. Until it increases another 50 percent or so in each direction - which means about twice as many pixels per square inch - things will not get substantially better" (Schreiner 2000). Or as Bill Gates put it, "The computer screen is a terrible limitation versus reading the newspaper" (Gates 1996).
Only in the last year has IBM announced a 200 dpi LCD display that has yet to go into production or have a price announcement (IBM 2000), and this display will still be only 1/36 the resolution of print. Entirely new electronic display technologies such as electronic ink, which is being developed by MIT and Xerox, are still a minimum of five years from being produced commercially ("Outlook," CQ Researcher 2000).
In addition to resolution problems, contrast is greater for print than for CRT displays. Contrast is the ratio of maximum brightness to maximum darkness. For CRTs it can be as high as 100:1 (Matkovic 1997) and for LCDs 300:1; however, this is only a theoretical maximum. Any display device which emits light (CRT, backlit LCD, projectors, etc.) loses contrast with increased ambient light (hence our darkened electronic classrooms and movie theaters). For LCD, the contrast also decreases if the display is at an angle to the reader. Practical figures for the contrast of CRTs in normal light is 20:1 to 40:1 (Goldwasser 2000) and for LCD it is approximately 100:1. Ink on paper has a contrast ratio of around 120:1 or three to six times greater than CRT displays and, since paper reflects light, the contrast does not deteriorate with ambient light. Most library patrons still use CRT displays and will continue to use them to read the e-books our libraries purchase. John Dvorak (1998), a columnist for PC Magazine, said that for e-books to be practical, "The display needs to be at least 300 dots per inch with a contrast ratio of 40:1 to 50:1 and it must be readable in both the brightest sunlight and the most poorly lit office. Good luck."
The physical problems caused by computer use have been classified by the American Optometric Association as "Computer Vision Syndrome" or CVS. Symptoms of CVS include eyestrain, blurred vision, headaches, back and neck aches, dry eyes, distorted color vision, temporary myopia, double vision, after images, and increased sensitivity to light (Anshel 1999; Von Stroh 1993). A Harris poll called "computer related eyestrain" the number one office-related health complaint (Seymour 1995). Nearly 10 million people annually seek eye exams because of problems related to computer use, forty times the number of people afflicted with carpal tunnel syndrome and other repetitive stress injuries (Von Stroh 1993). Between 50 percent and 75 percent of PC users complain of eye problems associated with computer use (Thomson 1998; Menezes 1999). The National Institute of Health and Safety said that 88 percent of the 66 million people in the United States who work at computers for more than three hours a day suffer from eyestrain (Seymour 1995). All types of computer monitors (color or monochrome) produce these symptoms (Salibelo 1995). Investigations at UC Berkeley indicate that the effects of CVS decrease productivity from 4 percent to 19 percent. The treatment cost for CVS approaches $2 billion annually (Chambers 1999).
Reading on computer displays is also inherently slower and less accurate. Proofreading experiments showed that reading speed is between 10 percent and 30 percent slower on a CRT and accuracy is 10 percent to 20 percent less (Ziefle 1998; Wright 1983). In one experiment, Gould et al. (1987a) showed that an extremely bad copy of poor quality print deemed unacceptable by test participants could still be read as fast and as accurately as a CRT, even by experienced computer users.
This combination of physical stress and lower reading speed and accuracy is why the "paperless society" envisioned in the 1970s and 1980s never happened. Whenever readers encounter more than a few paragraphs of text, they print them out as an unconscious protection mechanism. Bill Gates acknowledged this in a speech at Harvard in 1996 where he said, "when you get a large document it's very typical to print it out on your local printer and then read it on paper. Many people do this because anything more than about four or five screensful is just easier to read that way." Walt Crawford (1998) estimated that, "people will print out anything longer than five hundred words or so." Kurzweil (1992b) also recognized this: "Until the computer display truly rivals the qualities of paper, computers will increase the use of paper rather than replace it." E-books transfer the cost of printing from publishers to patrons without reducing library expenses and with increased environmental damage. Paul Curlander, chair man of printer maker Lexmark, predicted that office consumption of paper worldwide would increase from today's three trillion pages to eight trillion in 2010 (Tanner 2000).
New products such as Microsoft ClearType promise to increase the resolution of color displays and smooth the jagged edges of fonts by working at the level of the individual color dots that form the pixels on a computer screen rather than the pixels themselves. However, the maximum resolution improvement from ClearType is only three times the current resolution, which is still just 1/48 the resolution of print. As seen in figure 2, ClearType text is not, as stated by Microsoft, "as smooth as the words on a piece of paper" (Schreiner 2000). While ClearType may, in fact, render text somewhat more readable, further research should be done to determine the technique's overall effectiveness. One study of a similar technology found no difference in reading speed or comprehension with or without font smoothing (Gorin 1987). 3 This study, done as an MIT senior thesis project, was limited in scope and used a small sample but merits follow-up. In addition, ClearType will not correct the contrast problems of CRTs, and most stand-alone e-books use monochrome displays that cannot benefit from the technology.
The print book is a model of durability. It can be dropped from great heights, exposed to sand and food, and even fully immersed in water for brief periods without losing its information content. As a truly severe test, give a print book and an e-book reader to a six year old and see which one survives longer.
Even if the pages of a print book are ripped out, the information content still remains. Libraries recognize this and rebind books to return them to full usefulness. Print books degrade gradually; the content remains useful even when the pages are yellowed and the binding is worn. Electronic devices, by contrast, tend to fail catastrophically; a single transistor in one chip can turn the entire device and all contents into scrap.
From five-thousand-year-old papyrus scrolls to five-hundred-year-old Gutenberg Bibles, paper has demonstrated durability. Print on nonacid paper can last for five hundred years. No digital storage media are stable for longer than one hundred years (Van Bogart 1997a) and I have found none guaranteed longer than twenty-five years. This is less than the lifespan of acid paper which most libraries prefer not to collect.
Even if a digital medium and its data survive one hundred years, the hardware and software needed to read it will no longer be available. How many computers can read a 5 1/4" inch disk, the standard only a decade ago? The time before magnetic or optical media becomes obsolete is estimated by Van Bogart (1997c) of National Media Laboratory (NML) to be "ten or twenty years (or less)." As Kodak (1998) says on its Web site, after bragging about the projected one-hundred-year lifespan of its CDs:
The principal fact of life for all digital storage media is the rapid obsolescence of hardware and software. Users of CD technology should be reassured by the long physical life of CD discs, but they must not lose sight of the need to maintain a viable path for migration of data to new hardware and software platforms. Digital storage media impose a strict discipline that human-readable records do not: their rapid evolution creates a continual progression of technology that cannot safely be ignored for too long.
A more serious durability problem is the nature of the Internet, which is used as the primary means of distribution for most e-book systems. Internet companies and sites are notoriously short lived, as recent "dot-com" shakeouts have shown. The Kodak site still exists; however, the information on media durability quoted above was removed from NML's site after it was used in a U.S. News & World Report article (Tangley 1998), which Van Bogart said misrepresented NML's findings (Van Bogart 1998). The site itself is no longer open to the public. The only verifiable copies of the Van Bogart data are the print copies from the 1996 conference where he presented the data. A book in the hand is worth a database of books on a shut-down server.
No hardware is required to read a print book. All e-books require an expensive reader, whether a $1,000 computer or a $200 SoftBookTM Reader. There is a social aspect to this initial cost: the poor cannot afford computers or e-book readers. Unless we wish to develop an elitist collection, libraries must lend the expensive e-book readers. If they are lost or damaged, the library will also have to assume the replacement expense. It would be unacceptable in most libraries (especially public libraries) either to charge a deposit or to hold poor patrons liable for massive damage expenses. This makes e-book readers a continuing expense for the library.
Electronic books are still as expensive or more expensive than their print counterparts (Wildstrom 2000). One of the paradoxes of any new technology is that to become inexpensive, a technology must be ubiquitous, but it will not become ubiquitous until it is inexpensive. To overcome this, most new technologies must be produced at a loss and marketed on a par or at a discount compared to their mature competition until they gain ubiquity. That e-book vendors are not doing this may say something about the lack of confidence they have in their product.
Some continuing costs are unique to each media. Print books have shelving and space costs for libraries. If delivered via the Web, e-books have continuing subscription costs and costs associated with the computers, servers, and networks used to access them. If e-books are stand-alone, then the books must either be periodically repurchased or transferred to new media as they age or their format becomes obsolete. Since these costs are long-term and e-books have only been around for a short time, it is not yet possible to compare them.
Ease of Use
While technologically savvy people find e-books easy to use, nothing matches the simplicity of a print book - just open it and read. There is no learning curve involved. In addition, print books can be annotated easily with a pencil or highlighter and, while such annotations are the bane of libraries, they usually do not damage the information content. Dedicated e-book readers have some moderately complicated mechanism for annotation, while computer-based e-books have either no mechanism or one which requires significant practice to master and greater time to perform.
E-books have features comparable to print with the exception of skimming, browsing, and sharing content. A print book can be rapidly flipped through to find a certain text or illustration while the inherent slowness of computer displays, especially LCD displays, along with the inherent difficulty in reading displays makes this task impossible for e-books.
Most dedicated e-book readers tie the purchased book to a specific reader while e-books accessed over the Internet are usually tied to a specific set of IP addresses. This means the book cannot be loaned or given to another person without including the reader. "Paper seems ideally suited for sharing. . . . For the most part, paper provides an easy and inexpensive solution that is unlikely to be bettered by reading appliances" (Schilit 1999).
Most e-books use proprietary formats that cannot be read on different machines. This may change in the near future with the advent of the Open eBook standard (www.openebook.org); however, this standard could make current e-book readers obsolete.
E-books have four added features: text searching, hyperlinking, greater data density, and rapid updating. Text searching provides the ability to find specific sections in an e-book; however, this has limited usefulness since one must search on the exact word or phrase used by the author. By contrast, human-created indexes, often omitted from e-books, index concepts rather than words and cross-reference commonly used alternate terms. Hyperlinking can make electronic indexes very friendly and allow rapid switching between related sections of the text. This feature is most useful in highly cross-referenced texts and not as useful in linear texts such as novels and longer, descriptive works. Increased data density means that many e-books can be stored in one reader, allowing a person, for example, to carry a small reference library in a limited space. Rapid updating can be accomplished via the Internet. This can keep reference works much more up-to-date than their print counterparts.
E-books fail six of the eight criteria. They are not comparable to print books. As Wildstrom (2000) put it, "They're too pricey, hard to read, and offer limited titles." The print book, as the New York Times stated in a 1994 editorial "is close to perfect: cheap, durable, portable, and complete unto itself." As Harold Bloom, noted educator and literary critic put it, "Imagine that for the last five hundred years we had nothing but e-books, and then there was some great technological advance that brought us the printed and bound book. We would all be ecstatic. We would be celebrating after the long horror of the e-book" (Colker 2000, 10).
E-books are still very much in Kurzweil's development stage and not yet advanced enough to be a false pretender. The sales figures show this too; only twenty thousand to fifty thousand dedicated e-book readers have been sold (Holmes 2000; Menn 2000), far fewer than the first day sales of the latest Harry Potter book or the contents of a small branch library. The recent success of Stephen King's e-book, Riding the Bullet, does not counter this, since it is estimated that only 1 percent of those who downloaded the book actually read it (Menn 2000).
In spite of this, e-books may have a limited place in library collections where their special features outweigh their flaws. If searchability, linking, and currency are highly important and text is in short, discrete segments, e-books may be a useful solution. Such categories of books include:
- technical manuals
- handbooks (e.g., PDR, Merck Index, CRC Handbook of Chemistry and Physics)
As Gass put it, "Gazetteers, encyclopedias, and dictionaries are scholarly tools, but they are consulted rather than read" (emphasis mine) (Gass 2000). E-books do not work well with long segments of linear text such as novels, scholarly research works, most nonfiction, and text books. Even journals with lengthy articles do not lend themselves to electronic reading. These categories still comprise the majority of works in a library's collection.
Some librarians feel "they need to get on the e-book bandwagon now or risk being marginalized" (Ott 2000). Hage (2000) stated, "People want their library to be hip. They want their library to be willing to experiment when something new comes out," and pointed to library collections of VHS and Beta videocassettes as well as eight-track and audio cassettes. These arguments are predicated on the notion that libraries have always been on the cutting edge of technology, but this is not the case. Libraries may have collections of Beta tapes, but how many collected U-Matic videocassettes? We may have collected eight-tracks, but where are the collections of four-track cassettes or quadraphonic records? Libraries may have PCs in storage dating back to the IBM original, but how many of our libraries bought Altairs or IMSAIs? If the reader does not recognize U-Matic, quadraphonic, and Altair, this only emphasizes my point; these technologies were too preliminary and transient to be taken seriously for our collections. Libraries have seldom adopted the earliest, developmental stages of a technology. We are not true "early adopters" but "moderately early adopters." We wait until technologies sort themselves out and reach some level of standardization. This is all that I am suggesting for e-books. John V. Lombardi made a similar point during his speech to the 2000 ALA Annual Conference: "Being first to invent large-scale digital library projects is for those with money to lose, tolerant customers, and tenure. If it will take ten years to deliver value, let someone else invent it." Being the first to invest in a new technology is always expensive and generally unrewarding. The earliest libraries with online catalogs found themselves stuck with a huge investment in mainframe computers and custom software that provided limited, unreliable service for their patrons. Yet they could not afford to scrap these expensive systems and upgrade when smaller, cheaper, more reliable systems with greater functionality became available. Even choosing the most sophisticated new technology is not a guarantee of success as anyone with a Betamax video recorder or the OS/2 computer operating system will attest. Waiting for the market to settle lowers expenses, increases reliability, and reduces the chances that you will be left with rapidly outmoded technology.
Display technology is slowly improving and a standard format for e-books is in the works, but this still does not ensure the final adoption of e-book technology as the successor or even false pretender to the print book. Walt Crawford warns, "perhaps 80 percent of the time, the new technologies simply disappear or fade into specialized use" (1998). Other technologies now on the market such as print-on-demand will compete head-to-head with e-books. Print-on-demand stations can download a work from a vendor and provide a properly bound, paper book with all its inherent advantages in as little as five minutes. When the initial cost of this technology drops (as the cost of computer printers has dramatically done), it may be possible for libraries to provide patrons with print copies of any work and change the paradigm for collection development from "just in case" to "just in time."
Kurzweil's stages reassure us that we have no reason to hurry. Even if all of the technological problems with e-books were solved tomorrow and print books reached the stage of obsolescence, they would still be produced and remain useful for 5 percent to 10 percent of their mature lifespan. For print books, which have been produced for over five hundred years, that means we would have twenty-five to fifty years (one or two generations) to transition to the new technology.
If we begin more than limited collection of e-books, we risk alienating patrons who quickly will weary of the eyestrain caused by current e-books. We also risk wasting money on hardware and software that rapidly will become obsolete. As progress is made on e-book development, the eight criteria outlined here can be used to judge whether the technology is finally ready for widespread acceptance.
3. ClearType differs significantly from the type of anti-aliasing used in Gorin's experiment; however, Gorin's results suggest that the actual effects of the ClearType technology may not be as great as the stated, three-times effective increase in resolution.
Anshel, Jeffrey. 1999. The CVS Epidemic: How you can make computers easier on the eyes. Review of Optometry Online. Accessed Aug. 23, 2000, www.revoptom.com/issue/ro05f4.htm.
Atencio, Rosemarie. 1996. Eyestrain: The number one complaint of computer users. Computers in Libraries 16, no. 8: 40.
Chambers, Anne. 1999. Computer vision syndrome: Relief is in sight. Occupational Hazards 61, no. 10: 179.
Colker, David. 2000. Getting a Good Read on Electronic Books. Los Angeles Times, 2 November, T:1.
Crawford, Walt. 1998. Paper persists. Online 22, no. 1: 42.
Dvorak, John C. 1998. The next really bad idea. PC Magazine. Accessed Oct. 24, 2000, www.zdnet.com/filters/ printerfriendly/0,6061,2170454-2,00.html.
Gass, William H. 2000. At Issue: Will reading e-books be as satisfying as reading print books? No. CQ Researcher 10, no. 24: 561.
Gates, William. 1996. Keynote address May 29, 1996 in The Harvard Conference on the Internet and Society (Sebastopol, Calif.: O'Reilly & Associates), 27-35.
Goldwasser, Samuel M. 1994-2000. TV and monitor CRT (picture tube) information. Accessed Nov. 1, 2000, www.repairfaq.org/ sam/crtfaq.htm.
Gorin, Amy. 1987. One aspect of CRT displays that doesn't make a difference. Accessed Aug. 16, 2000, www.best.com/~ringo/ ringo-resume/thesis.html.
Gould, John D., Lizette Alfaro, Vincent Barnes, Rich Finn, Nancy Grischkowsky, and Angela Minuto. 1987a. Reading is slower from CRT displays than from paper: Attempts to isolate a single-variable explanation. Human Factors 29, no. 3: 269.
Gould, John D., Lizette Alfaro, Rich Finn, Brian Haupt, and Angela Minuto. 1987b. Reading from CRT displays can be as fast as reading from paper. Human Factors 29, no. 5: 497.
Hage, Christine L. and Stephen Sottong. 2000. Should libraries jump on the e-book bandwagon? American Libraries 31, no. 7: 61.
Harrison, Beverly L. 2000. E-books and the future of reading. IEEE Computer Graphics and Applications 20, no. 3: 32.
Hawkins, Donald T. 2000. Electronic books: A major publishing revolution. Online 24, no. 4: 14.
Holmes, Stanley. 2000. The cutting edge: Focus on technology: Making e-books easier on the eyes. Los Angeles Times, 20 Mar., C:1.
IBM, Inc. 2000. Display technology breakthrough: Researchers develop 200 pixels-per-inch display. Accessed May 17, 2000, www.research.ibm.com/news/detail/factsheet200.htm.
Kodak, Inc. 1998. Permanence, care, and handling of CDs. Accessed Feb. 12, 1999, www.kodak.com/US/en/digital/ techInfo/permanence9.shtml.
Kurzweil, Raymond. 1992a. The futurecast: The future of libraries part 1: The technology of the book. Library Journal 117, no. 1: 81.
_____. 1992b. The futurecast: The future of libraries part 2: The end of books. Library Journal 117, no. 3: 140.
_____. 1992c. The futurecast: The future of libraries part 3: The virtual library. Library Journal 117, no. 5: 63.
_____. 1993. The futurecast: The virtual book revisited. Library Journal 118, no. 3: 145.
Lombardi, John V. 2000. 20/20 vision for the future. Paper presented at ALA 2000 Annual Conference, 8-10 July, Chicago, Illinois.
Mason, Marilyn Gell. 1996. The yin and yang of knowing. In bricks, books & bytes, D ÆDALUS 125, no. 4: 161.
Matkovic, Kresimir. 1997. Tone mapping techniques and color image difference in global illumination. Institut für Computergraphik eingereicht an der Technischen Universität Wien Technisch-Naturwissenschaftliche Fakultät. Accessed Nov. 1, 2000, www.cg.tuwien.ac.at/research/theses/ matkovic/node24.html.
Menezes, Joaquim. 1999. Research points to PC use as cause of eyestrain. Computing Canada 25, no. 12: 15.
Menn, Joseph. 2000. E-book publishing: Much ado about nothing much? Los Angeles Times, 24 July, C:1
Murphy, Kim. 2000. Electronic literature: Thinking outside the book. Los Angeles Times, 24 July, E:1
Editorial. 1994. The New York Times, 21 Aug., sec. 4, p. 14.
Ott, Bill. 2000. E-books move the earth at Book Expo. American Libraries 31, 7: 30.
2000. Outlook: Turning a page. CQ Researcher 10, no. 24: 562.
Salibelo, Cosmo and Erik Nilsen. 1995. Is there a typical VDT patient? A demographic analysis. Journal of the American Optometric Association 66, no. 8: 479.
Schilit, Bill N., Morgan N. Price, Gene Golovchinsky, Kei Tanaka, and Catherine C. Marshall. 1999. As we may read: The reading appliance revolution. Computer 32, no. 1: 65.
Schreiner, Keri. 2000. E-books: It's all in the resolution. IEEE Multimedia 7, no. 2: 15.
Seymour, Jim. 1995. Your eyes come first. PC Magazine 14, no. 18: 93.
Sottong, Stephen. 1999. Don't power up that e-book yet. American Libraries 30, no. 5: 50.
Tangley, Laura. 1998. Whoops, there goes another CD-ROM. U.S. News Online Accessed Aug. 23, 2000, www.usnews.com/ usnews/issue/980216/16digi.htm. (Originally in U.S. News & World Report, Feb. 16, 1998.)
Tanner, Adam. 2000. Just say no to paper. Reuters on CMPnet. Accessed Oct. 11, 2000, www.techweb.com/wire/story/ reuters/REU20001010S0003.
Thomson, David W. 1998. Eye problems and visual display terminals - the facts and the fallacies. Ophthalmic and Physiological Optics 18, no. 2: 111.
Van Bogart, John W. C. 1997a. Archival stability of digital storage media; section 11.3: Life expectancies for magnetic tape and optical disc. National Media Laboratory. Accessed Mar. 4, 1997, www.nml.org (no longer accessible).
_____. 1997b. Disposition chart: 28¡ C (6814 F) & 40 percent RH. National Media Laboratory. Accessed Mar. 4, 1997, www. nml.org (no longer accessible).
_____. 1997c. Magnetic tape storage and handling: a guide for libraries and archives. section 4.2: Practical life expectancies. National Media Laboratory. Accessed Mar. 4, 1997, www. nml.org (no longer accessible).
_____. 1998. Update on "short life expectancy" articles. CD-Info. Accessed Aug. 23, 2000, www.cd-info.com/CDIC/ Industry/news/email-280498.html.
Von Stroh, Rolland. 1993. Computer vision syndrome. Occupational Health & Safety 62, no. 10: 62.
Wildstrom, Stephen H. 2000. A new chapter for e-books. Business Week 3674: 18.
Wright, P. and A. Lickorish. 1983. Proof-reading texts on screen and paper. Behaviour and Information Technology 2, no. 3: 227.
Ziefle, Martina. 1998. Effects of display resolution on visual performance. Human Factors 40, no. 4: 554.
Stephen Sottong ( firstname.lastname@example.org) is Engineering, Technology, and Computer Science Librarian and Leader of the Library Information Technology Team, California State University, Los Angeles.