Final Discussion

Taxonomic Authority Files Workshop

Stan Blum:

At the beginning of the workshop I asked a number of people to act as recorders—to take notes about any consensus positions that emerged and issues that needed further discussion. We convened over lunch and came up with a list. We also talked about what sort of process we should use in this wrap-up session to facilitate the development of recommendations.
 

I'm going to go through these points (Text Box 1) to give us an opportunity for some further discussion or clarification. Then we will to take a short break and ask each of you to write down what you feel are the top five action-items or topics that need further work. After you've had a chance to reach your own conclusions and if we have enough time, we'll open the floor for discussion again and give you an opportunity to raise any issues you thought were treated too lightly. Finally, I would like to collect your list of action items. I will then organize and tally them (at home), which will then give us some measure of strength or consensus around those items.
 

The main points noted by our recorders so far are listed here.

The first two recommendations concern the taxonomic codes. Would anyone like to make a comment about either of those?
 

Gary Rosenberg:

I guess one of the main things that people think about with modernizing the code is to allow electronic publication of new taxa. I personally don't want to see that happen for a while, because as it is, computerized databases can sit outside the system. You can document what's there; if you have a misspelling, it doesn't have any standing what so ever; emendations can't occur electronically. So I sort of like having them as separate systems. Once we work out better how archiving might work for publication of new taxa, maybe ten years down the line, we'll get to the point where electronic publication will be appropriate. I think right now the system works fine and computerization gives us a parallel system for tracking biological information without introducing new things that need to be tracked.
 

Bill Eschmeyer:

I just went on the Commission of Zoological Nomenclature, and we're a little too late; the draft is already well along, the basic changes have already been agreed to, and although it started out with some liberal changes, it ended up not changing much. So electronic publication is not acceptable, and so forth. The only thing that I see that might be taken into account for computerization is that you can have the same genus in Botany and Zoology—we couldn't in 1843, but they changed that later—and if you're doing faunal inventories you're going to have to put some code in to say whether it's an animal, or a plant, or a fungi. But we're just too late for this go-around, and it's usually a ten year period or more between revisions.
 

Chris Thompson:

It's not that we're too late, it's that the community rejected it, more or less overwhelming some of the modernization things when they were proposed. They were in the draft version that was presented to the zoological community—things like eliminating, once and for all Latin requirements for agreement between epithets and genus, etc. They were put to the community and they were rejected.
 

Walter Berendsohn:

Well for Botany, there has been a commission, the IAPT commission working on this problem, and in fact it was also not endorsed, to allow electronic-only publication of taxa. I don't know if the next Congress will follow that recommendation, but probably, if I can say this, the majority of the community does not see electronic publication as an immediate need. In fact, it is not such an urgent problem because you can of course put the names up and show them, it's just a question of all this permanent storage and persistence, all the insecurities connected to electronic records, which made us not recommend it for the time being.
 

Eimear Nic Lughadha:

I believe that resolving conflicts in the codes is a pretty low priority, given that they account for a very small percentage of the problems we've addressed in the last two days. Most of the names that are ever going to be published in Botany have already been published and will continue to be governed by the code under which they were published. If we now amend the code substantially, we will be in a situation where we have an old code and a new code. We'll be applying some conditions to the new names and some to the old names. It's not going to make things any simpler for the man on the street who wants to understand what's happening, or for the people who have to apply the code, or codes. It's not the answer.
 

Stan Blum:

OK, Unless anyone has anything critical to add, I'm going to take that series of remarks as indicating that modernizing the codes was not viewed as a critical problem in this forum. I think we can move on.
 

Chris Thompson:

I think the general issue, for the outside user, is that the codes are an internal sets of rules that we use to indicate the status of names in the various authority files, or whatever you call them, but they're not going to be terribly important to the outside user. So I think that we can live with the codes that we have.
 

Stan Blum:

Lets move on to #3. There was definitely some confusion about what an authority file is, and perhaps some significant apprehension about the term. But continuing on from that need for a clear definition, there was a call for thinking about how the various components of information—taxonomic names, bibliographic records, geographic names—feed into building nomenclatural authorities, which in turn feed into taxonomic authorities, which feed into classifications, and I won't let you forget how they get applied in item-level cataloging, whether in biological collections or observation-based data sets. Any further discussion on that?
 

Larry Speers:

Maybe a comment more than discussion. I'm trying to coordinate biodiversity information systems, and of the presentations I heard, Species 2000 and ITIS are the only ones looking at integration of both zoological and botanical names. From a biodiversity point of view, it's the association between the plants and animals that users need. We've got to break down these walls when we're building authority files and make sure that they're scalable, like the Species 2000 and ITIS systems.
 

Stan Blum:

But I would add to that, that I didn't hear anything in the description of the Plant Names Project to indicate that you couldn't use that technology very easily in the zoological community.
 

Nancy Morin:

Maybe an action item could be working out a way that we could compare the standards that are being used in the different disciplines. Are the zoologists using different standard abbreviations for their periodicals or geographic names? Maybe we ought to see just how much difference there is.
 

Stan Blum:

So that's a call for working more within the framework of TDWG, or something like that, to develop better standards. That seems to be getting at #4, standards development—we need appropriate mechanisms to foster standards development processes, and to get broader participation from the community, we need greater awareness of the processes and resulting standards.
 

Karen Calhoun:

I just wanted to make the point that the International Federation of Library Associations has been dealing with the difficulties of having competing national name authority files, and the same kind problems have cropped up (like hair going up on the back of the neck) when you use the word "authority", or talk about requiring only one authorized form of a heading, only one allowable form, one classification of that name. So we have begun to stop saying "authority control files" in that kind of setting, and we say "access control files," and that helps a lot in preventing misunderstandings
 

Stan Blum:

So it's access control now—you're not going to let me access this, huh? [Laughter]
 

Karen Calhoun:

Well, yes, that would perhaps lead to another set of issues, but at least you get rid of the word "authority."
 

Frank Bisby:

In relation to item #4 and TDWG, as we've mentioned a great number of times in this meeting, one or two of the people who spoke didn't seem to realize that TDWG has been devoted for two and half years now to standards across the whole of zoology, microbiology and botany. It did start from botany, Heywood and I started it in 1985, but at the request of IUBS, it has recently broadened to cover the whole scope. It has been an organization that has had fine aspirations and often not succeeded in them, and that's perhaps due to the lack of resources, but I was curious as to how many people in the room here, who have talked about standards, either know when the meeting is this year or are planning to attend it? For instance, one of the discussions sessions this year will be on the standard that I wrote for them, or proposed to them seven years ago, for plant names in botanical databases, and because of the action in Species 2000, we've been asked to generalize that to a system that will work for plants, animals, and microbes, across all three codes. That will be one of the discussions there. Similarly, they had a system for geographical areas of the world that is used by botanical databases very widely, the Brummit system, and again there is a debate to broaden that out, to cover the oceans as well as the land, and to make it more acceptable to zoologists as well as botanists. So I think there are things there that may be of interest to this community. That's in the TDWG organization.
 

Paul Morris:

Going again back a sociological point, let me suggest that we use a phrase like "resource files" to discuss these, that lacks any of those connotations of control or authority.
 

Randy Ballew:

Well, Stan, I hate to say it, but I want to get rid of "file." [Laughter] It's kind of left over from the days when we could never underestimate the bandwidth of a station wagon full of 12 track tapes. So something like Plant Names Resource might be more appropriate.
 

Mary Mickevich:

I would like to propose that there be a committee—I hate committees, but how about two people—one from each field, who would coordinate information resources, for those who would like to have them. For instance, I would have found it very nice to go in and zap off a publishers' city file, or an author file. It would have made my life much easier.
 

Stan Blum:

Well that was one of things I was hoping would come out of this. John Riemer has prepared a one-page bibliography for some of the on-line resources, and perhaps we could expand it. Some of it was for documentation of standards, but there are readily available resources that we all need direction to, from the systematics side, and conversely we've been hearing from the library side that they would like to have similar sorts of pointers to resources within our community.
 

Mary Mickevich:

Yes, that was the next thing I was going to say. They would probably really like to have some of my author files with all the different names and references for all the people that I have. So, if we could do this it would be much easier for all of us. We could share our various information resources.
 

Adam Schiff:

Let me just volunteer. I maintain the web site for the Natural History Caucus of the Special Libraries Association, and on that site are links to a lot of these databases already. There is a set of links to reference resources. I'm not sure that I have compiled all the ones that should be on there, but I would certainly be willing to collect those and get those up on there and have a centralized site, which would be particularly of use to natural history librarians as well as systematists. We already to point to the Plants database, we point to ITIS I think, we point to some of the specific disciplinary lists, like the "Crocodile, Tuatara, Turtle" list (which is up on the web), the Mammal Species of the World.
 

Mary Mickevich:

Would you have the lists from the library community that I might want, such as a publishers' cities list, or authors lists? I'm probably not calling it the right thing, but I saw someone flash up a list of cities and countries, and counties, and publishers.
 

Stan Blum:

That might have been the Name Authority file, which is searchable on the web at the Library of Congress web site. We could point you to it.
 

Adam Schiff:

The Name Authority file contains millions of items, though, so you wouldn't want to download it. It is searchable at the Library of Congress, and in some cases, even through individual institutions' library catalogs. So we could point you in that direction, but there is no one list that you could get printed out or anything like that.
 

Laurel Jizba:

Isn't it true, that we use the USGS terms for setting up new headings, not our own. Our source data is the USGS.
 

Adam Schiff:

For geographic names that is the preferred source.
 

Laurel Jizba:

So, isn't that what she's asking?
 

Adam Schiff:

Right. Actually we use the Library of Congress Name Authority File, but when we create authorities we are instructed to use the USGS databases as the source. I can give the URL of the site now, and actually I would appreciate it, if you know of things that should be on the list but aren't, please let me know. The URL is http://www.lib.washington.edu/sla/ and that will put you in the Natural History Caucus of the Special Libraries Association, and on that page there is a link to "References Sources" in the natural sciences. Those are arranged by discipline, such as Botany and Zoology, for example. Those are reference resources for authorities as well. Basically, we're trying to point out these lists that we're talking about. And if you know of things that are missing, please let me know about them.
 

Michael Dadd:
 

I just want to add that there are also quite a lot of lists and pointers to databases on our web-site. It's one of the things that we've been collecting and evaluating for some time. The URL is http://www.york.biosis.org, and you can follow through and find a lot of resources there.
 

John Riemer:

For how much longer will this workshop continue to have a web page? Perhaps we could point from that page to these resources.
 

Stan Blum:

I will add these.

We were talking about standards development before we got off onto the resource issue. The other item that came up in relation to standards work (#5) was training and dissemination. Does anyone have a comment on that?
 

Chris Thompson:

Under item #4, we do have a standards organization. The problem with TDWG is that the zoologists continue to ignore it. In essence, the recommendation is that we should use some of the resources that we have for standards development, and the zoologists should be participating to ensure that existing standards for authors, geographical areas, etc., are adapted to our needs, rather than create new committees and things.
 

Michael Dadd:

I would agree entirely with what Chris said, I would just like to ask Frank a question for clarification of that. You [Frank] said that IUBS have expanded TDWG's remit, but how many zoologists do you get turning up at the meetings?
 

Frank Bisby:

I don't run TDWG; I'm not an officer. The last meeting was held in Taipei and the number of zoologists was small, but he total of people was small as well. So let's turn this around a bit. I've just moved jobs, and I now just happen to be at the institution that is hosting the meetings this year. (That wasn't planned.) The thing that would be most positive would be for the TDWG officers to receive an invitation to hold the next year's meeting at an institution that is dominated by zoology; for example, the Natural History Museum, or the Smithsonian, or USDA, or something. TDWG did hold a meeting at the Smithsonian, but it was run by the botanists and almost none of the zoologists attended. There have been many shortcomings in TDWG's activities; it's become very quiet, it's had difficulty in keeping going, but it's made a start. Why don't you get a move to get a zoological organization to host the next meeting. I'm sure the officers would welcome receiving that.
 

Stan Blum

OK, #6: We've heard discussion from the Library community about the role that LOC (Library of Congress) and other very large organizations have played, and systems that were centralized, probably for technological reasons way back when, but continue to operate effectively. We don't seem to have any analogs in the systematics community. So it was suggested that we need a financially and politically empowered entity to have an impact on the way we do business. Comments on that point?
 

Gregory New:

I'm wondering if both #6 and #7 aren't both a little ahead of the game. We're trying to get away from "authority file" terminology and then we're asking for a politically empowered entity and then an individual to spear-head/manage a concerted effort. Does this workshop have a permanent structure? [SB: No.] Shouldn't we be thinking about of some sort of committee to continue the activity or the thrust, to make sure that it doesn't get lost. This is something that might be referred to a committee, to look into things like this rather than make a decision or make a recommendation.
 

Barbara Bauldock:

I'd like to make this group aware that the President's Council of Advisors on Science and Technology (PCAST) just issued a report called "Teeming with Life", that looks at the National Biological Information Infrastructure, and says that's what's needed now is a second generation NBII. The PCAST report calls for funding this at $40 million a year, for five years, and until such time as that financial support can be provided to NBII-2, to quadruple the amount of money that is currently going to support what we're doing in NBII now. So I would encourage you, as you hear about the PCAST report, to look for opportunities within the framework of NBII. The report is just a recommendation. What we're trying to do now is to get initiatives in the Congressional budget to get the necessary funds appropriated, but please support these activities. Perhaps that will lead to a politically and financially empowered entity.
 

Nancy Morin:

It's really a tremendous oversight that we haven't been talking about NBII. Maybe it's because not enough people know about it, but it really has put itself forward as a facilitating, centrally-coordinating, helpful, national approach to a lot of the things we've been talking about. So, it could be that even before that funding comes through, we could use that as part of the way to get our stuff done.
 

Karen Calhoun:

I think having "entity" as a singular noun might overlook some opportunities to have partnerships with more organizations. In the discussion of what made things work in the Library world, I think we might have glossed over the role of the professional associations. The American Library Association (ALA) has been very active in the development of standards and support for all of the initiatives that were undertaken. Without the professional associations we would not have been able to succeed.
 

Frank Bisby:

I was assuming the items #6 and #7 were in some way related to the OECD Mega-Science Forum proposal for a Global Biodiversity Information Facility, and one of the main actions of that group would be to enhance the various facilities for bioinformatics and the resources to go with them. So I suppose the question for me is whether you are interpreting items 6 and 7 at a national or international level, or as a US contribution to a global system. But clearly that proposal is on the table.
 

Larry Speers:

Very few people are actually aware of what that proposal is. It's only been circulated to the people involved with the Mega-Science Forum and it won't be presented to the member countries until January. We are proposing a global biodiversity information facility as a distributed thing, with all the interoperability, and need for files, but I don't think that has gotten down to the details of how that will be organized or exactly what that will undertake. This obviously would have to be a component of that. How it's geographically or regionally organized, I don't think that's been determined. There have been suggestions, but nothing is concrete.
 

Stan Blum:

(#8) We discussed the notion that "things happen" through the work of particularly motivated and effective people; that it needs to be somebody's job to push things. Individual people make things happen, and organizations help in this only to the extent that they have, support, or provide such people. On the other hand, we need insurance against the "hit by a bus" phenomenon, where an important community effort can depend too much on a single person. So there is a role for community-level organizations and some degree of balance.
 

Moving on to #9, there was a recommendation that we contact the director of the digital library federation, which is a fairly new organization that coordinates activities that have resulted from the substantial funding of digital library research. The recommendation was that we should find out how they're managing coordinating the various activities in the digital library world. Comments on that?
 

With regard to intellectual property issues, there was an observation that the library authority files are considered in the "public domain", which seems to contrast with what's going on in taxonomic authority files. So there is a list of issues that need to be addressed, such as: funding to develop the files initially, professional rewards, and long-term support.
 

Walter Berendsohn:

Just a question, can you generate bibliographic citations from the authority files. If so, you say this is public domain. There are lots of titles in there. One clarification, are there article titles there as well or is this just monographs? [off mic.] Oh, just monographs. There is nothing underway to do articles?
 

Joan Swanekamp:

I would say that generally our catalogs are just monographs and serial titles, but in fact in many of our older catalogs there was a practice of recording article length descriptions back in the last century and well into the beginning of this century. We're finding huge numbers of those in our retrospective conversion process and we're trying to deal with them.
 

John Attig:

I think libraries are accustomed to providing access to "services" that index article-level information, but these are almost invariably not public domain databases. They're the ones you're familiar with.
 

John Riemer:

I've seen some new authority records for drug names that cite as source authorities (670 fields) articles instead of monographs.
 

Adam Schiff:

Just to echo that. When you're setting up an authority you cite information resources that support the heading you've chosen or provide variants, and you can use articles, but very few catalogers are likely to do that. They're likely to go to reference books first, and then to scientific monographs at the whole book level, and often we do use citations out of BIOSIS or Zoological Record, a usage of a taxonomic name, but it's unlikely that you'd find a specific article listed. Even in the citations you wouldn't cite the article, you would cite the journal and the volume, and not the actual author and the article within that journal.
 

John Mitchell:

To add to what Adam just said, for any new concepts that are just in the journal literature, that you wouldn't find in the monographic literature, we always look to the journal literature to make our citations.
 

Gary Rosenberg:

I know my institution catalogs articles by any author at our institution. They end up in our card catalog, but I was wondering if we could turn it around. Maybe systematists could start adding journal articles to library databases. If that were acceptable, there is already a structure in place to distribute that kind of information.
 

Stan Blum:

I think that's scaring the librarians, but I'm not sure.
 

Linda Ward:

I know that the Smithsonian Natural History Library doesn't want to see anything of our reprint collections, but a lot of us have gone ahead and cataloged the reprint collections individually, and in some cases have made them available on the Web or on CDs. But as a regular thing, they don't want to touch them.
 

Lowell Ashley:

As one of the catalogers in the Smithsonian Institution Library, I do see a good number of articles and we do catalog them as separate things and contribute them to the OCLC database. In the larger scheme of things, these would be just a drop in the bucket. There aren't very many of them in relation to the whole universe of articles out there, but this does happen in most libraries, I think.
 

Chris Thompson:

I don't know what the situation is in the botanical Code, but there is, at least in the ancient literature, a difference between what is a preprint, and what is an article in it. Preprints and separately published articles have different dates, and different nomenclatural implications and they should be treated as a monograph and should be cataloged separately. They are probably the hardest things for taxonomists to find, because they're ignored by some librarians.
 

Stan Blum:

So it sounds like there are some significant scope differences between our two domains. That might be an area for further work and coordination, so that we might work together better in the future. Which brings us to #11.
 

Joan Swanekamp:

Actually I was going to speak to #12, because I think in fact they go together. Our models for digital libraries are somewhat different than they have been for our traditional card catalogs, or even in our online catalogs. In fact many of our institutions are providing cataloging, regular cataloging, for locally created digital resources that might be preprints, that might be articles by professors, that might exist, in our case, in the Forestry School's web pages, that are very important to the Yale community and for which we're taking responsibility. I came from Columbia University, and the exact same thing was happening there—that reports that were generated in the Business School, or the School of International Affairs and elsewhere on campus, were being cataloged in a way that we might not have done before because we were having access to them in a ways that we hadn't before. So I think that this digital library model provides some new opportunities to provide access to this stuff that we haven't in the past.
 

Stan Blum:

So it appears that #11 and #12 might best be merged into a recommendation to get natural history libraries, natural history collections, and informatics efforts to get coordinated under the model of a digital library.
 

Joan Swanekamp:

Right. I mean, we view our collection differently now. Where it was just the stuff that came in, went through our acquisitions department, it's now the stuff that's available throughout our campus.
 

Adam Schiff:

It think it's important to point out that in many institutions, the catalogers are not the people who make selections of what to catalog. So I think for many of you who do want this stuff cataloged, it's important that you do talk to the people who make those selections, and tell the catalog department that they should be cataloged. In small institutions, they may be the same people, but in larger institutions they aren't necessarily the same people at all, so it's the public service science librarians who are making the selections about what books to buy and what electronic resources to catalog. So if you want those things, you need to let those people know, so that it gets to us, who actually do the cataloging.
 

Gail Hodge:

To follow up a little on what Joan said, one of the things that we're seeing with digital libraries, both within universities and government settings, is that there is an attempt to provide access to more resources, whether they're digital or not, by coming up with a mechanism for exchange of information that is a much smaller record structure than what you saw with MARC. There was mention of Dublin Core. I don't know how many of you know what that is, but it's essentially 15 very basic data elements that have been agreed to—spurred by OCLC and a consortium of folks who meet as the Dublin Core (that's Dublin, as in Ohio, which happens to be where the first meeting was held). They are now planning their fifth meeting in Washington, later this year, and the number of people involved in the Dublin Core project is growing. There are lots of pilot projects, international as well as national in this area. So it makes it more reasonable to consider including the kind of resources you're talking about, and to have them contributed by professional catalogers doing the cataloging, but having some of the metadata provided in Dublin Core format by the authors and then reviewed by catalogers and put up in a much more efficient mechanism to the digital library.
 

Nancy Morin:

Maybe we could look at this at an even higher level, because what we're talking about is a natural history digital library project because our authority files mostly don't exist just to exist by themselves; they're our way of giving people access to a lot more information: specimens, descriptions, phylogenies, all kinds of other things. So that might turn out to be a source of funding, if we could get it together.
 

John Attig:

I think the digital library model is very useful and that means we're talking about a different sets of standards (this is summarizing what other people have said) and that metadata is what we're talking about. I think that one of the strengths of this is that metadata doesn't have to be a unilateral, single, standard. Dublin Core is in fact sort of a "least common denominator" of common elements, that can be extended with specific elements that are needed to describe whatever domain that you're dealing with. So it gives you an awful lot of flexibility for not losing information as you're trying to provide access in a more general context. I think this a very useful model to consider for those reasons.
 

Greg Whitbread:

Are the metadata standards like Dublin Core going to be enough to tie the two communities together, or should we be looking at some other way to do it? It seems we have two enormous repositories of data, and they should be linked together somehow. Really that's what we should be talking about.
 

Jim Beach:

I'm working with one of the CIMI working groups looking at metadata for natural history collections, and we learning as we go, but to answer your question partly, I think that Dublin Core will provide sort of the loosest level of linking that will be possible in a networked environment. It's really for initial discovery of networked resources—that's really it's objective—so it won't provide anything like the high level of integration that say the Plant Names Project is shooting for, in terms of say transparent connections between databases and applications, but I think to answer your question more fully we'd have to consider the different kinds of linking and the objectives of those kinds of linkages; what the community needs and what would be useful.
 

Stan Blum:

I think we're getting into research, planning, and educational issues. These kinds of things seem like appropriate topics for presentations at TDWG meetings, on an on-going basis, so that people in our community can keep abreast of emerging models and technologies.
 

We have all of the intellectual property rights issues: credit (#13) and audit trailing (#14). Perhaps audit trailing is relevant to other areas as well, like what goes into making a record authoritative, but contributing records to a community database is something perhaps that appointments and promotions committees need to know about, and providing that feedback is one function that audit trailing can fulfill. We need sustained and appropriate levels of funding (#15). We had that issue of perfection versus availability (#16). I think I heard some convincing arguments about earlier availability actually resulting in a more rapid asymptote towards perfection. Perhaps I shouldn't characterize that as a group consensus, but I think the issues there were fairly well laid out.
 

It was noted that our authorities or data objects need item-level metadata to indicate status or support quality assessment (#17).
 

And finally there is #18 that a summary of this workshop should be published in journals, such as BioScience or Science.
 

Now what I would ask you to do is to write down what you would like to present as action items.