Proposed amendment of Articles 8, 9, 10, 21 and 78 of the International Code of Zoological Nomenclature to expand and reﬁne methods of publication

For 250 years, taxonomists have relied on having access to physical copies of published works in order to verify information about taxa. The difficulty of tracking down publications can be a substantial impediment to taxonomic work, which has led some researchers to advocate modifying the codes of nomenclature, for plants and bacteria as well as for animals, to allow electronic publication of scientific names (Knapp et al., 2007; Wheeler & Krell, 2007). Electronic publishing and access to online information can clearly accelerate progress in taxonomy (Wheeler, 2008). Even when the fourth edition of the International Code of Zoological Nomenclature (ICZN, 1999) was published, it was clear that electronic publication could not be ignored (Ride, 1999, 2003). In all fields of science, however, long-term access to digital information has been and remains a significant concern (Anonymous, 2001; Howe et al., 2008; Szalay, 2008). Botanists have rejected proposals for electronic publication because of concerns for archiving and security from tampering (Zander, 2004; McNeill, 2006). Although some works printed on paper are rare, the number of instances in which no copies are known to survive is vanishingly small, so it certainly has been possible to reliably archive paper copies. Will we be able to say the same for electronic works 100 years from now? Can we have both rapid and effectively permanent access to publications and the information they contain? The International Commission on Zoological Nomenclature (ICZN) considered these issues at a special session in Paris, August 23–25, 2008, preceding the XXth International Congress of Zoology (ICZ). The Commissioners voted to start the process required to amend the International Code of Zoological Nomenclature to allow electronic publication of new scientific names and other nomenclatural acts. A summary of the draft amendment was presented at the ICZ on August 27, during the Systema Naturae 250 symposium. The draft amendment was then refined via email discussion among Commissioners during the following month. In Paris, Commissioners voted separately in favour of three principles relating to publication. None of these passed unanimously, but all had at least a two-thirds majority among the twelve voting.

For 250 years, taxonomists have relied on having access to physical copies of published works in order to verify information about taxa. The difficulty of tracking down publications can be a substantial impediment to taxonomic work, which has led some researchers to advocate modifying the codes of nomenclature, for plants and bacteria as well as for animals, to allow electronic publication of scientific names (Knapp et al., 2007;Wheeler & Krell, 2007).
Electronic publishing and access to online information can clearly accelerate progress in taxonomy (Wheeler, 2008). Even when the fourth edition of the International Code of Zoological Nomenclature (ICZN, 1999) was published, it was clear that electronic publication could not be ignored (Ride, 1999(Ride, , 2003. In all fields of science, however, long-term access to digital information has been and remains a significant concern (Anonymous, 2001;Howe et al., 2008;Szalay, 2008). Botanists have rejected proposals for electronic publication because of concerns for archiving and security from tampering (Zander, 2004;McNeill, 2006). Although some works printed on paper are rare, the number of instances in which no copies are known to survive is vanishingly small, so it certainly has been possible to reliably archive paper copies. Will we be able to say the same for electronic works 100 years from now? Can we have both rapid and effectively permanent access to publications and the information they contain?
The International Commission on Zoological Nomenclature (ICZN) considered these issues at a special session in Paris, August 23-25, 2008, preced-ing the XXth International Congress of Zoology (ICZ).
The Commissioners voted to start the process required to amend the International Code of Zoological Nomenclature to allow electronic publication of new scientific names and other nomenclatural acts. A summary of the draft amendment was presented at the ICZ on August 27, during the Systema Naturae 250 symposium. The draft amendment was then refined via email discussion among Commissioners during the following month.
In Paris, Commissioners voted separately in favour of three principles relating to publication. None of these passed unanimously, but all had at least a two-thirds majority among the twelve voting.
• Electronic-only publications should be allowed, if mechanisms can be found that give reasonable assurance of the long-term accessibility of the information they contain. • Some method of registration should be part of the mechanism of allowing electronic publication of names and nomenclatural acts. • Physical works that are not paper-based (e.g. CD-ROMs, DVDs) should be disallowed.
One development favourable to the acceptance of electronic publication is the emergence of mature platforms for archiving electronic documents, such as Portico, which offers a permanent archive for electronic journals, and LOCKSS (Lots of Copies Keep Stuff Safe), an international, communal initiative based at Stanford University Libraries (Maniatis et al., 2005;Fenton, 2006). These systems comply with the ISO standard for an Open Archive Information System (OAIS) and seek to assure the long-term retrievability of archived electronic publications, even when the publisher no longer supports distribution. The ICZN does not advocate these over other such archives, but sees them as examples of archiving technology that may prove acceptable to the zoological community.
The International Commission on Zoological Nomenclature therefore proposes the following amendment of the fourth edition of the International Code of Zoological Nomenclature (hereafter 'the Code'), to allow electronic publishing coupled with archiving of works and registration of new scientific names. In addition to electronic publication, the proposal perforce addresses methods of publication in general. It is followed by an explanation of the procedure for amendment and adoption, and discussion of the reasoning behind the proposal. Comments from the community of zoologists are sought, whether to oppose, support, or suggest improvements and alternatives for the various parts of this proposal.

PROPOSED AMENDMENT
In the proposed amendment, normal font represents existing text of the Code. Text in italics is new text; titles that otherwise would be in italics are shown in small uppercase letters. Text in square brackets [ ] describes the changes proposed in a given section. The amendment affects Articles 8, 9, 10, 21, and 78, and the Glossary.

Article 8. What constitutes published work.
A work is to be regarded as published for the purposes of zoological nomenclature if it complies with the requirements of this Article and is not excluded by the provisions of Article 9.
[Article 8.1 is revised to add electronic publication separately from methods that produce physical copies.] 8.1. Criteria to be met. A work must satisfy the following criteria: 8. The Commission will revise the amendment in light of comments received from zoologists (Constitution Article 16.1.4). Particular areas for discussion are highlighted below, but the Commission seeks comments on all aspects of the proposed amendment. Some comments will be published in BZN or posted on the ICZN website (www.iczn.org). Revised text of the amendment will also be posted there, and in the ICZNwiki (iczn.ansp.org).
Formal comments should be sent to Dr Ellinor Michel, Executive Secretary of the ICZN (iczn@nhm. ac.uk) or to Dr Gary Rosenberg, chair of the ICZN Editorial Committee (rosenberg@ansp.org). Zoologists may also contact Commissioners directly for informal discussions (Appendix 1; http://www.iczn.org/ Commissioners.htm).
Assuming that sufficient consensus is reached in the zoological community, the revised amendment will be presented at the International Union of Biological Sciences (IUBS) meeting in Cape Town, South Africa, in November, 2009 for provisional ratification under Constitution Article 16.1.5.1. Final adoption would then be by a vote of the Commissioners.

DISCUSSION
Although this proposal affects several Articles of the Code, it is presented as a single amendment. If presented as multiple amendments, each would be subject to separate vote, but the intention is to present an integrated package addressing issues of methods of publication.
In formulating the amendment, Commissioners have been concerned with issues of verifiability and encouragement of best practices. With printed publications it generally has been possible to verify from the work itself whether the names and nomenclatural acts it contains are available (leaving aside the problem of doctoral theses). With the advent of electronic publication, this distinction has been blurred. It is not unusual today to receive a PDF of an article without knowing if the journal in which it appeared was printed on paper or issued only electronically. This leaves the user unable to judge availability.
Registration of names is one means of addressing this issue, although it introduces dependency on an outside agency, as does the requirement for archiving of works. To obviate the need to check both the Official Register and an archive to determine whether a name is available, the proposed amendment requires that the location of the future archive for a work be stated when the name is registered, and that the registration number be given in the original publication. This should allow the user to determine from the work itself that a name fulfils these requirements. Further considerations on archiving, registration, and other aspects of the proposed amendment are presented below in order of the number of the proposed article.

FORMATS FOR ELECTRONIC PUBLICATION (PROPOSED 8.1.3.2)
It is difficult to specify which formats can be accepted for electronic publication. E-journals that have a fixed and final published form clearly are appropriate outlets for publishing nomenclatural works, whereas continuously modifiable forms such as online databases and interactive keys are not. In specifying 'fixed content and format', the intention is that the same text and illustrations should appear with the same pagination and proportions whenever a copy of the work is viewed, recognizing that different software will not render a page identically on different computers. Also, external links are excluded under 'fixed content'; linked material such as 3-D tomographic images or videos of behaviour would not be considered part of a publication. Information necessary to confer availability must be in the work itself.

WORKS PRINTED ON PAPER (PROPOSED 8.4.1)
The third edition of the Code excluded electrostatic and xerographic methods from conventional printing, in Article 9(3). The fourth edition does not mention these methods, and despite providing cut-off dates for the acceptability of methods other than conventional printing in Articles 8.4 and 8.5, does not define or restrict conventional printing. The effect is to make works produced by electrostatic and xerographic printing acceptable retroactively. In practice, zoologists have accepted as published anything that looked like a journal article or monograph, whether produced by letterpress, inkjet printer, laser printer, or xerography. They have not accepted printing on thermal paper or output from dot-matrix printers, presumably because they could judge that they did not fulfil the requirements of Article 8.1.1.
With the proliferation of printing techniques, it is impractical to ask the user of a work to judge how it was produced. The proposed amendment therefore has a single, simplified article governing printing of works after 2009, accepting all forms of ink or toner on paper. The exceptions are hectographing and mimeographing, which it is clear the Code intended to exclude after 1985, and which are readily identified as reproductions of typewritten works. This exclusion is transferred to Article 9.2 in the proposed amendment, allowing the current Article 8.4 to be dropped.
The Commission is not aware of paper-based works that zoologists have judged not to be published because of electrostatic, xerographic, or other unconventional methods of production. Therefore, the proposed revision drops the current Article 8.5. If there are such examples, we would like to hear of them. This will help in assessing the advisability of this proposed simplification of the Code.

CD-ROM AND DVD (PROPOSED 8.4.2)
The fourth edition of the Code does not specifically refer to CD-ROM or DVD in its legislative text, but the Introduction (p. xxvii) refers to 'read-only laser disks', so clearly the intent was to include them as methods of production that do not employ printing on paper (current Article 8.6). As relatively few works have been published on CD-ROM or DVD, and the long-term durability and readability of those media are controversial, the proposed amendment disallows them after 2009. It was in principle possible for a work on CD-ROM to have been issued before 2000 under Article 8.5 (then Article 8d), as a method other than conventional printing. The Commission, however, is not aware of any that contain the statements required under Articles 8.5.2 and 8.5.3. If such exist, they should be brought to the attention of the Commission.
ARCHIVING (PROPOSED 8.5) The library community shares with the taxonomic community concern for long-term accessibility of published works, and has played a leading role in the development of archives such as Portico and LOCKSS mentioned in the Introduction. Both of these are curated, dark archives; 'dark' in that a work will be released by the archive only if the publisher no longer supports distribution; 'curated' in that the archive takes responsibility for migrating the content to new formats as needed to address changes in technology. This means, however, that the original format (e.g. PDF) may not be preserved even though the content is preserved.
Zoologists may be surprised to know that a number of taxonomic journals are already in such archives, for example, The Auk, Copeia, Journal of Paleontology, Journal of Molluscan Studies, Zoologica Scripta, and Zoological Journal of the Linnean Society. For Portico and LOCKSS, the titles held can be determined from their websites ‡, but as yet, it is not possible to tell what range of years or specific issues are held unless one has audit-level access. The Commission intends to learn more about these and other archives to determine whether their mode of operation is, or can be modified to be, compatible with the needs of zoologists.
When it is found that the requirement for archiving was not fulfilled, the proposed amendment (8.5.3.3) requires that the case be referred to the Commission. While there is concern that this might lead to an increased number of cases, it seems necessary to formally decide whether a work was published. Otherwise, some names and acts might be rejected, while others in the same work might still be used by those unaware of the problem. Also, some names in a work might be better known than others, so it may be necessary to conserve some and reject others. Formal action by the Commission is the best way to achieve consistent results in such situations.

RECOMMENDATIONS FOR ARTICLE 8
A suggested minimum number of 25 copies in a printed edition is added in proposed Recommendation 8B to address several concerns. One is that 'numerous' in the current and previous editions of the Code is undefined. Article 8.6 of the current Code has been taken by some workers as establishing 5 copies as the minimum number, but that Article applies only to works not printed on paper. Some authors of electronic works currently use print publication of a small number of copies to satisfy requirements of availability, but it has been controversial whether such works should be considered 'obtainable' (Article 8.1.2) if sent only to a few colleagues. A similar route might be taken by authors finding that their electronic publications failed to meet requirements for archiving or registration. Also, Article 21.8 allows date of publication to be advanced by distribution of preprints, so this recommendation provides guidance on a minimum number to distribute. ‡For titles committed to Portico and LOCKSS, see http://www. portico.org/about/committed_titles_alpha.html and http://www. lockss.org/locks/Publishers_and_Titles_List. A related issue is print-on-demand. Print-ondemand in principle could be used to produce a traditional print-run. One could order 'numerous' copies of a work (or 25 under the proposed revision), so the method fulfils the requirement of producing 'simultaneously obtainable' copies (Article 8.1.3). Yet print-ondemand can also be used to produce a single unique copy of a work, which would not be published under the Code. Because there could be multiple versions of a work, as the author finds and corrects errors, printon-demand does not 'assure' identical copies. This is not a new problem: there are works known to have been produced in multiple silent editions, with changes between them that affect nomenclature -but each of those editions was in multiple copies. The ease and inexpensiveness of production without need for physical storage make such problems much more likely with print-on-demand. Furthermore, often the user of a work will not know that it was produced by print-on-demand and so subject to these concerns.
Concerns about the ease of modifying works produced by print-on-demand apply with even more force to works distributed electronically. If the text distributed by a publisher were to be silently modified in aspects that affect zoological nomenclature, it would be difficult establish what the original text was, and on what date the modifications occurred, especially since there likely would not be date-stamped copies in libraries. Proposed Recommendation 8D therefore urges publishers not to allow changes to works after publication, particularly for electronic and print-ondemand works, and to handle any needed corrections through errata notices or other separate publications. The requirement for archiving within one year also limits the time during which silent changes might be made to electronic publications.

REGISTRATION OF NAMES (PROPOSED 10.9)
The Commission intends ZooBank (www.zoobank.org) to be the Official Register of Zoological Nomenclature. While ZooBank has the potential to benefit taxonomy by making nomenclatural information readily accessible (Polaszek et al., 2005), the proposed amendment focuses on registration as a mechanism to make it readily verifiable that some of the criteria of availability have been fulfilled.
In particular, the user of an electronic work can determine that the requirements for archiving and registration have been fulfilled when registration numbers are given in the work. This also avoids a 'limbo' period during the one year allowed between publication and archiving (proposed article 8.5.3.3), when it would not be possible to tell if a work was published under the Code. The presence of the registration number allows the user to accept that the archiving requirement has been fulfilled, but the user can also directly examine the archive if desired. Should a work prove not to have been archived, a decision by the Commission would be needed, as explained above.
Although some authors have advocated that registration be mandatory for all names (Wheeler & Krell, 2007), the proposed amendment requires registration only for names published electronically. Those who wish to publish on paper may continue to do so, without the need for archiving or registration. The proposal does not require registration for other nomenclatural acts, such as lectotype and neotype designations and first revisers' choices, as it is likely that this could cause problems, with authors neglecting to register many such acts. We hope that registration of names and acts will become a community standard, enforced by reviewers and editors, as has submitting genetic sequences to GenBank and publication of GenBank numbers.
WORKS ISSUED ON PAPER AND ELECTRONICALLY (PROPOSED 21.9) For the purposes of the Code, 'electronic publication' and 'electronic-only' publication are effectively synonymous. An electronic publication is not the same work as the printed edition of that publication, and each edition has its own criteria of availability. This does not mean that names and other nomenclatural acts are at risk of being published twice. The same situation already exists with repaginated reprints, or second printings that still state 'new species'. In practice, when the relevant content is the same, most taxonomists deem that only one name was introduced or that only one nomenclatural act has occurred. This needs to be made explicit in the fifth edition of the Code, but as this is a broader concern than electronic publishing in itself, it is not treated here. In cases where there are differences between versions (e.g., different spellings of a name, different specimens designated as the holotype), the versions are treated as different works. Divergence between electronic and online versions will surely be common (e.g., black-and-white images in print, color online), but such differences will probably only rarely affect availability of names and nomenclatural acts.