nternet Working Papers?We are moving into uncharted waters. In the Internet age, the meaning of "publication" is likely to be redefined. The literal meaning of publication is disseminating information to the public. Dictionaries do not yet indicate whether this refers solely to publication in hard copies or also to on-line publication. In the past, journals and working papers have coexisted harmoniously. How electronic scholarly publication will evolve in the Internet age is highly uncertain, and murky at best. The electronic and paper versions may coexist in the future, at least for the next two decades.
Most journals will become available electronically in the coming years. Publishers estimate that for extra 15 percent of the cost of hard copies, you can get both paper and electronic copies of a journal. Hard copies may still be available, but they will become more expensive because of limited print runs and are likely to be discontinued eventually. They may even become collectibles in the next century.
Look at JEL, for example. Its CD-ROM versions are widely used and EconLit is available in most libraries. Some people may simply discard hard copies of general journals —and this might be the fate of many paper journals—and rely on their electronic versions, which are infinitely more useful because of the search capability. Within the next decade, most of the top journals, as well as newly created journals, are likely to become available only in electronic editions.
The purpose of this note is to clarify an interim arrangement, which is likely to prevail during the transition period when both hard copies and electronic versions of journals are available. Once the profession defines acceptable publication practices for the Internet age, the publication process will also adjust to the newly defined professional norm.
Yes. You may put working papers, abstracts and all, on the Internet. Working papers on the Internet will be considered for publication in RIE and other Blackwell journals.
Yes. However, many important journals use the double-blind reviews. The current format of working papers does not facilitate the double-blind refereeing process. A working paper should be split into two parts for electronic submission: (a) cover page which contains all the contact information about the authors, and (b) the main body which does not reveal the identity of the authors. If you integrate both parts, the paper cannot be sent to the referees electronically without revealing the identity of authors.
Yes. In other words, a working paper published on the Internet will not be treated as a "publication." Some sites which provide Internet working papers might be willing to add a note, pointing to the journal which accepted the paper for publication or providing the exact citation of the published paper. A working paper, electronic or not, should be allowed to rest once a revised version of the paper is published.
No. Upon acceptance, the authors are asked to transfer the paper's copyright to the publisher. For instance, the copyright assignment form of Blackwell journals (Economic Journal, Economica, Econometrica, Review of Economic Studies, RIE, RDE, etc.) states:
"that this article is the author(s)' original work, has not been previously published elsewhere in its final form either in printed or electronic format, and is not under consideration for publication elsewhere."Therefore, the final version should not be made available on the Internet, in home pages or on FTP sites to permit downloading by the public.
No. The intent of a working paper is to get feedback from colleagues, and it is common knowledge that after getting inputs from editors and referees, a working paper will be modified before it is accepted for publication. You can make hard copies and distribute them personally, but the publisher does not want you to make an electronic version available on the Internet allowing indiscriminate dissemination of the paper.
Yes, you may e-mail the final version personally, at least for now. In principle, all articles can be obtained free from the authors via e-mail.1 However, do not post it on your home pages or FTP sites, which permit impersonal and indiscriminate distribution. In contrast, you may personally e-mail, fax, or mail the final version to your colleagues and fellow researchers.2
Please note that this is not a universally accepted arrangement, and our profession has yet to evolve to an equilibrium state, in which the competitive publishers recover the production costs and we authors maximize academic freedom. If zero profits are not guaranteed, any concessions may be rescinded through legal battles.
During the transition period, different publishers may have different rules governing electronic transmission. For instance, the copyright transfer agreement of Elsevier Science (North-Holland) explicitly recognizes that authors retain the following right:
The right to photocopy or make single electronic copies of the article for their own personal use, including for their own classroom use, or for the personal use of colleagues, provided the copies are not offered for sale and are not distributed in a systematic way outside of their employing institutions (e.g., via e-mail list or public file server). Posting of the article on a secure network (not accessible to the public) within the author's institution is permitted.
Publishers oppose impersonal, wide dissemination of the finished paper to the public, because it would significantly reduce, if not eliminate, the demand for journals.
Suppose two identical copies of the final version become available, one on your home page or FTP sites (free), and the other on a site, licensed by the publisher, which is available only to a subscriber (purchased). Both copies can be electronically transmitted to readers. The journal publisher advertises the titles of papers and names of authors on the Internet.
Jack is a researcher and needs a certain paper in that journal. What will he do? He finds the name of the author, Jill, locates her home page and downloads the paper. This way, the journal fails to sell a single copy. Being a competitive firm yet unable to cover production costs, the publisher shuts down. Jill, as well as Jack, now has no place to publish papers. There goes their chance for tenure and promotion. (Jack and Jill might try publishing their papers on the Internet. If all papers are published on the Internet on their own, not through the journal publishers, how will their research be evaluated? This sort of academic chaos is a possible equilibrium scenario.) This is why journal publishers are demanding that the final version of an accepted or published paper not be made available on the Internet.
There are hundreds of journal publishers; they appear to be competitive and they are small businesses by any criteria, employing fewer than a hundred employees. Some publishers get by with a handful of employees. They seem to be more concerned with survival than with trying to monopolize the market.
Yes, eventually, though both formats will be available temporarily to ease the transition. This prediction is based on the conjecture that the Internet will be ultimately preferred by both consumers and producers.
Electronic versions are instantaneously and more widely available to the public than the hard copies. If the intent of "publication" is dissemination, then impersonal Internet publication is far more effective than personal distribution of the paper version. And for this reason alone, most journals will become available on the Internet, and sooner or later, the journals not available on the Internet are likely to disappear.
From the consumers' standpoint, Internet journals have an absolute advantage in delivering information; an Internet paper arrives immediately, whereas several weeks or months intervene between the subscription decision and actual delivery of the journal. Internet papers permit an automated search, whereas human eyes must do the scrolling over the paper version. Internet papers can directly link to the papers cited, providing immediate access to other papers in the area. Unlike leatherbound books, paper journals do not even have display value and they take up precious office space.
From the producers' viewpoint, production of Internet journals is far less costly than paper journals. And as for the shipping cost ..., what shipping cost? There once was a medium called 8-track tapes. Paper journals would face the same fate in the next century.
Currently, several publishers (Academic Press, Blackwell Publishers) make academic journals available on the Internet, using Portable Document Format (PDF) or Postscript (PS) files, which can be viewed by separate programs such as Adobe Acrobat or Ghostscript. Although these viewers allow the reader to search, the texts are in black and white, and they do not permit direct editing. Any editing must be done in the original word processor files and the revised texts must be reconverted using a PDF converter or a Postscript printer.
These document viewers have two major problems. First, these programs or files do not permit hypertext links to other reference materials. Second, they cannot display color images. Thus, at best PDF or PS viewers may be a temporary publishing format, and only a partial solution. If Acrobat or similar programs evolve a pair of wings — hypertext link and color graphics — PDF may become the preferred standard for viewing academic journals in the Internet age.
However, in the next century, or perhaps even sooner, Web browsers will be capable of displaying all the mathematical symbols used in academic papers. Compared to PDF or PS viewers, Web browsers have three absolute advantages.
First, Hypertext Markup Language (HTML) 3 permits all the math symbols and Greek letters. Currently, Netscape Communicator permits Greek symbols. The next generation of HTML might be capable of printing all the math symbols. Currently they support super- and subscripts. You can get a preview of future Internet articles in this exposition. For instance, you can express national income as:
and, expected value as
HTML can display many equations. See more examples.
To print Greek letters, you need to change the font face. That is,
insert any alphabet between the symbol tags,
<FONT FACE=SYMBOL></FONT>. For instance, the following
alphabet
abcdefghijklmnopqrstuvwxyz
ABCDEFGHIJKLMNOPQRSTUVWXYZ
will be converted into
abcdefghijklmnopqrstuvwxyz
ABCDEFGHIJKLMNOPQRSTUVWXYZ
Second, HTML provides a direct link to the articles or sources being cited. For instance, if you click John Heise or MacTutor History in Question 12, you will be linked instantaneously to their home pages. Since most academic journals will be available on the Internet, this would probably become the preferred style of electronic references.3
Third, Web browsers are already capable of displaying color images, whereas PDF or PS viewers are not. So Internet papers (yes, including academic papers) are likely to use colorful images, though they are not now acceptable in paper journals.
But, again, it remains to be seen whether our profession will stick to the old tradition or move forward. One thing is certain: a century from now the Internet journals may not use the current PDF.
In a sense, publishing Internet working papers may be likened to smoking; it provides instantaneous satisfaction, even though it may be harmful to the user. Each person must weigh the benefits and costs. People are free to do whatever they please with their working papers.
Internet working papers are a quickest means of becoming well-known; Even if a working paper is not yet published, you can achieve immediate fame from it. However, such electronic publications may not be treated as publications for tenure and promotion decisions.
Say that a working paper on the Internet contains a publishable idea. It can be downloaded by anybody, who can replace some symbols, change the title, rewrite the introduction, modify the story a little, and submit it to a journal as a totally different paper in one afternoon. With some luck, that paper may even be accepted sooner than the original.
A more sophisticated person would even cite the working paper, thereby legitimizing the reconstituted paper. Of course, this has been possible all along, but the downloading capability via Internet makes it very tempting and effortless to plagiarize the ideas of other researchers. The problem is: In this case, the plagiarist may achieve publication and the original author may not. The final product can hardly be called plagiarism, albeit the intent may have been present. (It is also interesting to note that Pythagoras spent 13 years in Babylon.)
A person with a conscience would never do such a heinous thing. However, if no one ever attempts to do it, we must be living in Paradise; a world of enlightened people. (And I hope such an age of light will arrive here on earth some day soon.)
Multiple submission of articles with the same substance is not viewed as an acceptable practice by the economics profession. If they are two different papers in substance, there is no problem. But if the substance is the same, a problem may arise. Words are not important; it is ideas that count. Cosmetic changes are irrelevant.
Cuneiform tablet, Plimpton 322, Columbia University.
An old Babylonian tablet (1900 - 1600 BC), shown above, contains the so-called Pythagorean Theorem, except that it predates Pythagoras by a millennium or more. According to Neugebauer and Sachs (1946), the tablet lists in the two middle columns the numbers that satisfy the so-called Pythagorean Theorem. Specifically, from the left, the first column indexes the contents of the table (1, 2, 3, …), the second and the third the hypotenuse c and the leg a of a right triangle all in sexagesimal numbers. The fourth column shows (c/b)2 where b is the basis of the triangle. For instance, the 11th row shows 75, 45, and 1.5625 = (75/60)2. A translation of another Babylonian tablet preserved in the British museum states (John Heise):
4 is the length and 5 the diagonal. What is the breadth? Its size is not known. 4 times 4 is 16. 5 times 5 is 25. You take 16 from 25 and there remains 9. What times what shall I take in order to get 9? 3 times 3 is 9. 3 is the breadth.If the Internet had been available in those days, the unknown Babylonian mathematician would get credit for establishing the Theorem, and Pythagoras (6th century BC, image from MacTutor History) would not be able to publish his theorem. He could still write a paper providing a proof — later Euclid found a more elegant proof which is now used in textbooks — or stating his discovery of irrational numbers, a corollary of the Theorem for which he sacrificed a hundred bulls to the gods. He could not state the same theorem in different words though; its substance is the same as that of the unknown Babylonian mathematician.
If both "papers" were by the same author, s/he would be able to publish a paper for the Babylonian result, but a second paper may not be publishable if it claims originality of the theorem. A paper claiming the originality of the proof or a corollary, however, would be publishable.
In short, the two versions should be different in substance. In principle, an author shouldn't be publishing the same result twice in different words.
The principal mission of an academic journal is to publish new or original ideas, not duplications or elaborations of existing materials (Review articles that have pedagogic value are an exception.) If you are not sure about the inherent difference between two papers, try Joachim Zietz's test (1996)4: Would you be embarrassed if both papers were published in the same issue of a journal? If not, go ahead!
2. Authors can set up their e-mail program so that they can automatically respond to inquiries and send the papers via e-mail to all requesters. Publishers must draw a line somewhere but do not know where yet.
3. However, the old style of displaying the referenced articles will remain for the time being.
4. From personal correspondence.
MacTutor History of Mathematics Archive, http://www-groups.dcs.st- and.ac.uk/~history), 1996.
