1. Introduction

There is a daunting array of other issues which are related to the publication activities of the Society, but the standing committee on publications considers many of these issues on a regular basis. Questions have been raised with us about specific operational issues such as the editorial policies of the Canadian Journal of Mathematics (CJM) and Canadian Mathematical Bulletin (CMB) and the composition of the set of Associate Editors. Such questions are to be expected from a diverse mathematical community and the standing committee is the appropriate forum for their consideration. 

Why is science public, with the exception of some research done at certain government and corporate laboratories? This is not the place to study these fundamental issues, but the importance of the continuation of a strong public system can perhaps be summarized as follows. 

(1) The openness and transparency of the research done, and the publishing of results, are in exchange for the considerable freedoms granted researchers by society, through salaries and tenure, etc., at public or partially public institutions. Implicit in this is also a mild form of control: access to privileges and finite resources is based on excellence, and not on less desirable criteria such as political or personal connections. 

(2) For reasons of efficiency, scientific results must be freely available to all researchers in a given discipline; this is to avoid wasteful duplication of effort, to enhance cross-fertilization and thus improve scientific progress. 

The CMS plays a major role in Canada with respect to scientific communication. Our report is aimed at helping the Society maintain that role. The primary area of long term concern is the revolution in dissemination through refereed journal publications (is there a revolution?). We discuss this in section 2. To get a snapshot of the current situation in Canada, we have compiled an inventory of Canadian Mathematics Journals in section 3 with the actual tables of data presented in the Appendix. In section 4, we consider the possibility for the addition of new journals in the future and conclude that the purely electronic format should be seriously considered. 

The TeX office located at the University of Manitoba is the technical heart of the Society's publication enterprise. We discuss this operation in section 5. The report concludes with a look at the various series of books published by the CMS. 

We make a total of 7 concrete recommendations. It was certainly tempting to make more, but others that we formulated were really too representative of the personal prejudices of task force members. We chose in the end to be restrained in our recommendations. The recommendations stand out in displayed italics in the text for ease of location. 

2. The Future of Scientific Publication

In a series of well-researched and well-documented papers, Andrew Odlyzko of AT&T Labs - Research gives compelling arguments in favour of a change to electronic means of communication. In his paper "The rapid evolution of scholarly communication" (preliminary version, March 19, 2000) he writes 

Usage of electronic forms of scholarly information has typically grown at 50 to 100 percent per year [...]. On the other hand, print usage has shown little change, as far as anyone can tell. Thus [...] a decade is about the length of time we should expect for new modes of electronic communication to become dominant, if current growth rates continue. (p. 4f.)

[...] the inertia of the scholarly publishing system is enormous, and so traditional journals have not changed much. They are in the process of migrating to the Web, but operate just as they did in print. However, we are beginning to see the sprouting seeds of new ventures that will lead to new modes of operations. Still, it will be a while before they become a sizable fraction of the total scholarly publishing enterprise. (Ibid., p. 16)

Our personal observations are consistent with much of Odlyzko's data. As with most major journals, the latest issues of CJM and CMB are available on-line well before they arrive in departmental reading rooms. The www.arXiv.org e-Print archive has become the primary mode for dissemination in some sub-disciplines of mathematics and mathematical physics. The number of papers posted to the mathematics part of the archive seems to be following a logistic growth pattern. In 1993 there were 18 postings per month, in 1997 there were 32 per month, in 1998 this jumped to 166 and since January, 1999, the average is close to 200 postings per month. We only have anecdotal evidence on readership of these postings, but one does often see papers being printed on the community printer with the vertical logo of the archive on the front page. 

One active researcher stated that in the last month, he may have read the abstracts of more than a hundred papers on-line from www.arXiv.org or postings of the latest issues of assorted journals, downloaded more than 30 for a closer look on the monitor and printed out 10 to take home for traditional reading with pencil in hand. On the other hand, this person reported only a few minutes spent in the reading room where recent print copies appear and no visits to the main library to look at older papers in the past month. 

There is a revolution going on in the way we access new results in our areas of interest, but it seems to be blending in with traditional publishing. Most papers in the e-print archive move on through the refereeing process to appear in a print journal or an electronic journal with a traditional editorial procedure. We expect that the need for libraries to hold print copies of journals will decrease as the issues are removed from the stacks less often. Those journals available in both formats will maintain their subscription levels; although increasingly libraries will chose the "electronic only" option. 

3. An Inventory of Journals

The move to electronic media can be considered as a partial solution to this crisis. However, there are also a number of challenges and as yet open questions. Perhaps the most important question concerns ownership of the published information. Is an electronic article in the public domain, or is it a commodity? 

It is in this uncertain and quickly changing environment that one should see the sub-group of Canadian mathematics journals in their present state, and in their future development. 

3.1 Definitions

Journal: The main criterion is that the publication be primarily devoted to refereed research articles. We therefore exclude society newsletters (such as the CMS Notes or the Gazette des Sciences mathématiques du Québec) even if they publish written versions of important lectures and similar material; such newsletters will only be listed at the end. However, we do include the purely electronic journal "Theory and Application of Categories". 

Mathematics: A definition of the "mathematics" component may be a matter of contention. We include the "Canadian Applied Mathematics Quarterly", but not the journals in operations research, statistics, and the history and philosophy of mathematics. However, we list the relevant journals in a separate category, without much further study. Similarly, journals primarily devoted to mathematics education and/or problem solving are listed separately. Not listed at all are publications in computer science. 

Canadian: For most journals in this study there is no doubt that they can be considered Canadian; this is certainly the case with all the society journals. However, in some cases it is by no means obvious whether a journal could (or should) be considered "Canadian". We used the criterion that the editor(s)-in-chief be located at Canadian institutions. This means that a journal may "migrate" to or from Canada. Two such cases were the move, in 1997, of the editorial office of "Aequationes Mathematicae" from Waterloo to Graz, Austria, and also in the 90s the change in editorship of "Order" from a Canadian to a US mathematician and institution. Since the publishers in both cases are large international companies, these journals can no longer be considered Canadian. We list them in a separate category. A more difficult case is "Historia Mathematica"; this journal was founded in Canada, but is now published by a big US company, with only one of the two editors-in-chief based in Canada. "Utilitas Mathematica" poses a similar problem; the journal's front cover actually shows "Durban, South Africa", recently changed from "Winnipeg, Canada". In this case we decided to still count the journal as Canadian since both the managing editor and the small publishing company are located in Winnipeg. 

According to these restrictions, then, there are 11 Canadian mathematics journals devoted to research; we shall refer to them as the primary journals. 

In the Appendix, there are tables providing a snapshot of the journal situation. The data in the tables should be seen as approximations only, although we tried to be as accurate as was possible or reasonable. In some cases a piece of information could not be obtained with a reasonable effort; those cases are indicated by question marks. A slash indicates cases where the corresponding entries were not available or not applicable. 

3.2 Discussion of the Primary Journals

Publishers: Four of the journals are published by three different societies, namely two by the CMS, and one each by the AMQ and the Royal Society. One journal is published by a North American consortium, sponsored by a Canadian society. Two are published by large multinational companies, with a third one distributed by such a company, while officially published by a university. Two journals are published by small independent (but interrelated) companies, and the last one is published electronically and independently by its editor. 

Specialization: 4 of the 11 journals are "general", 4 specialize in combinatorics and graph theory and related fields, and one each in universal algebra, applied mathematics, and category theory. 

B. Page charges
Most journals do not charge page charges. The CJM and the CMB abolished them in 1995; the loss in revenue at that time was made up for by a moderate increase in the subscription rates. Only the two least expensive journals have obligatory page charges, but payment is not a condition for publication. Utilitas Math. and Ars Combin. have a system of "priority page charges": if they are paid, the paper in question will be published faster by adding more pages to the next available issue. 

C. Subscription rates
Most journals have differential fee structures, with significant discounts for individuals or for society members, both individual and institutional. In some cases a journal subscription is part of the membership privileges. No attempt has been made to document the various fee structures; the tables show only the subscription rates charged to libraries that are not institutional members of an appropriate society. 

D. Circulation
We made no attempt to obtain circulation figures, but we did obtain the numbers of Canadian institutions (mainly university libraries) that subscribe, or subscribed, to a given title. The figures may not be entirely accurate, but may be useful for the purpose of comparison; they were obtained from the AMICUS database of the National Library of Canada. 

A few points are noteworthy: 

1. The erosion in subscriptions throughout the 1990's is not as dramatic as one might have expected, given the severe budget restrictions suffered by all libraries. Not surprisingly, the most inexpensive journals (such as the Ann. Sci. Math. Que. and the Math. Reports) were least affected, but even the very expensive commercial publications Algebra Universalis and Aequationes Math. were able to hold on to much of their Canadian library subscription base. A possible explanation is that libraries were reluctant to cut Canadian journals. 

While only the latest available figures (1999/2000) are shown in the tables, we did also compile the figures for 1990 and 1995. The drop in Canadian institutional subscriptions did not exceed 20% in the case of commercial publications, and 10% for society publications. 

2. It appears that new (print) journals have not been able to establish a viable subscription base, at least among Canadian research libraries. The trend is quite striking. Most journals were established in the 1970's or earlier; they are being subscribed to by respectable numbers of institutions. However, the numbers for the three newest print journals are as follows: "Order", founded as a Canadian journal in 1984: 13-15 libraries; Canad. Appl. Math. Quarterly (1993): 9-10 libraries; J. Algebraic Combinatorics (1992): 5 libraries. The last two are especially surprising since these are journals in areas of particular strength throughout the country. The Journal of Algebraic Combinatorics appears to be doing better internationally, while a low overall circulation is a matter of considerable concern for the Canadian Applied Mathematics Quarterly. 

E. Electronic access
Six of the 11 primary journals are available electronically; one of them is purely electronic. Access to these journals by individual mathematicians is fairly easy through campus licenses if a library subscribes to a journal, but this may depend on the particular arrangement at a given library. "Theory and Applications of Categories" is free for everybody. Two of the remaining journals (Ann. Sci. Math. Québec and Canad. Appl. Math. Quarterly) are open to the idea and/or envisage an electronic presence in the not too distant future. Two journals (Ars Combin. and Utilitas Math.) have no plans for an electronic version. Nine of the 11 primary journals have a homepage, as do all 9 journals in Tables 2-4. Titles and abstracts can be found on some of these homepages. 

F. Quality and impact
The quality of a journal is difficult to quantify objectively; it has many dimensions and depends on many factors. Within a given field there is usually a perception of which journals are "better" than other journals; we made no efforts to document this mainly anecdotal evidence. In some cases the "quality", however it may be defined, may not be the right, or the only, criterion for the value of a journal. For instance, the "Math. Reports of the Royal Soc." is a journal of rapid publication of short papers (limit: 6 pages) which may not have complete proofs. A certain compromise in quality is therefore inevitable and not necessarily harmful. Similarly, the "Ann. Sci. Math. Québec" started as a regional journal, and only fairly recently has tried to establish itself as an international journal with high standards. On the other hand, the "Canad. J. Math." and the "Canad. Math. Bull." have had decades to establish themselves as first-rate journals, supported by a stable subscription base, good technical and clerical support, and high-profile editorial boards with the backing of a major society. 

One attempt at a fairly objective measurement is the "Impact Factor" from the Journal Citation Reports published by the Institute for Scientific Information (the publisher of the Science Citation Index). The last column in the tables shows the Impact Factor for 1998, if it was available. It is the number of citations (by other journals) to articles in the 1996 and 1997 issues of the given journal, divided by the number of articles published in that journal. 

To put these figures into perspective, it may be interesting to consider the rankings within a discipline. Among the 138 journals listed under the "mathematics" heading, the Canad. J. Math. ranked 45th, just below the J. Algebra, and just ahead of the J. London Math. Soc., to mention two well-known non-Canadian journals. Leading the list was Acta Mathematica with an impact factor of about 2. The complete rankings are: 

1. Mathematics, 138 journals: 

2. Applied Mathematics, 127 journals: 

3. Other Disciplines: 

G. Purely Electronic Journal
The electronic journal "Theory and Applications of Categories" deserves special consideration because at this time it is the only journal in Canada of its kind. In spite of the obvious major differences from all the other journals, it does in many respects fit into the group of primary research journals. It is a peer-reviewed international journal publishing original research; it has an editorial board consisting of eminent researchers in the field, and perhaps most importantly, it serves as a "community centre" for an active branch of mathematics. The journal successfully took over this role from a well-established print journal (J. Pure Appl. Algebra) which in recent years was abandoned by this research community because of unacceptable increases in subscription rates. 

"Theory and Applications of Categories" is edited by its founder R. Rosebrugh at Mt. Allison University, but it appears that the journal has reached a state of maturity and acceptance in which its existence does not depend on this particular editor. The journal is free of charge, supported by the editor's unpaid work and implicitly by his university. 

H. The Non-primary Journals
Some explanation was already given above. More journals could probably have been added to Table 3 (such as the Ontario Mathematics Gazette); however, we included only the two listed because of their connections to CMS and PIMS, respectively. While "Crux" is well-established and has an international readership, the very new "Pi in the Sky" is expected to be distributed to high schools in Alberta. 

Miscellaneous

Newsletters
For the sake of completeness of this inventory, we mention the existence of the following society and institute newsletters:

• Bulletin AMQ
• CAIMS Newsletter
• CMS Notes
• CSHPM Bulletin
• Gazette des sci. math. du Québec.
• PIMS Newsletter

3.3. Challenges Facing Journals

Low Circulation

Too many or too few submissions

The opposite problem, while less of a threat to the future and the quality of a journal, leads to serious workload problems for editors, and sometimes very large backlogs. A journal loses its effectiveness as scientific medium if the time lag between submission and publication becomes too long. This problem is often addressed by increasing the number of pages published per year, and thus also increasing the subscription rates; this was certainly part of the reason behind the "serials crisis" of the 1990s. Other journals increase the rejection rate, thereby raising standards, while still others use a combination of both approaches. 

Changes of Editors

Most journals on our lists have successfully gone through changes in editorship. Some, such as the two CMS journals, have editors with fixed terms, and are therefore very resilient to changes. In other cases a change in editorship means that the journal "moves away" from Canada; this was the case recently with "Aequationes Math." and "Order". However, since mathematics is such an international discipline, these moves should not raise too much concern. 

The Future of Publishing

3.4. CMS Involvement and Recommendations

The Past

The Present, and Recommendations

B. The Canadian Applied Mathematics Quarterly
This is a matter of a different nature, and of more immediate concern. As was mentioned above, the "Quarterly" has had difficulties establishing a viable subscription base. The survival of the journal would benefit the Canadian mathematical community and would be in the interest of the CMS. The Society could help the journal in a variety of ways, from listing it on the membership renewal form along with the CMS publications, through providing technical or clerical support (e.g., through the TeX office), to taking it on as a CMS publication. 

4. New and/or Electronic Journals

• Authors want an alternative to commercial journals and there are few or no journals in their particular area sponsored by societies or universities.
• In recent years, the perceived changing editorial policies of the CJM and CMB have resulted in papers in their area no longer or rarely being published in a Canadian based journal.

In light of these arguments, we considered whether the Society should be open to expanding our journal offerings. But Table 1 shows that new journals introduced in recent years are not being subscribed to by most Canadian libraries. Although the cuts seem to be slowing down, it is not the time to introduce new print based journals with standard subscription rates. They will not reach a wide readership and, therefore, not attract a heathy base of authors. But print journals are not the only option. 

Electronic journals certainly appear to be here to stay. Most mainstream print journals now have an electronic version which is available free to institutions that receive the print version. More and more researchers check the contents of their favorite journals on-line before the print version of a volume arrives at the local library. Interesting articles are downloaded and scanned or even read completely on the computer before a choice is made to print a copy. The CJM and CMB are both operating very well in this format. 

Besides those print journals that now have an electronic version, there have been some successful purely electronic journals. One Canadian example is Theory and Applications of Categories; with a strong editorial board and traditional refereeing procedures, it is established as a major journal in the area. As such, it is successful. However, this journal is offered at no charge. 

The AMS offers two electronic only journals, Representation Theory and Conformal Geometry and Dynamics. These are free to institutions with subscriptions to the standard print journals of the AMS. Other institutions are charged $100 and individual members pay $25 per year for access to either of these journals. 

We think the CMS could be successful following the model of these two electronic journals of the AMS. However, there is an important condition to be satisfied before such an expansion occurs. There must be a strong research group that is international in scope but with a significant committed Canadian component interested in the establishment of a specialty journal. In anticipation of such a proposal coming forward, the Society can take some modest steps to prepare itself. 

5. The TeX Office

The office now produces camera-ready copy for both the Journal and the Bulletin and sends them off to the University of Toronto Press. In recent years, both journals have established a reputation for appearing on schedule and for their exceptional high quality of production. Much of the processing of manuscripts is now automated and Michael has developed documentation for the scripts and the process. 

In addition, the office also produces camera-ready copy of Crux Mathematicorum with Mathematical Mayhem and, under contract with the Royal Society of Canada, The Mathematical Reports of the Royal Society. The TeX Office has also produced several books, notably the three volumes of the fiftieth anniversary commemorative series for the Society and some conference proceedings on contract. There is the expectation that the amount of such work will increase. 

It is our opinion that the TeX Office represents an invaluable resource to the Society. We have been on sound financial footing for the last decade; even building up a substantial endowment fund. The TeX Office is a significant factor in that success. Michael Doob is hale, healthy and enthusiastic about continuing in his role as TeX Editor. We do not think it is time yet to worry about continuity in that position. However, we wonder if Michael might take on an apprentice. 

An apprentice TeX Editor could be located anywhere in the country. He or she could begin with a small, well-defined, component of Michael's work. As time went on, the capacity of the combined offices would increase so that more contract projects could be handled. This would provide a robustness to the TeX operations that would make all of us more comfortable.

6. Books

The CMS Book Series is published under a recent agreement with Springer Verlag; this replaces the CMS Advanced Book Series with Wiley. The first three books are due out in 2000 and the editors report several exciting projects underway. It is too early to determine how successful this new agreement with Springer will be, but our expectations are positive. 

The ATOM Series consists of booklets intended to stimulate the interests of young people in mathematics during the middle school and high school years. The booklets are very inexpensive and the initial volumes are very good value for the cost. We wish that there were more volumes being written. 

The CMS Conference Proceedings Series (CPS) represented the Society's vehicle for dissemination of papers read at conferences. In some disciplines, a refereed conference presentation is one of the primary modes of research publication - this is the case in computer science for example. Within some rapidly developing areas of mathematics, timely conferences are often the venue for the presentation of cornerstone results. Thus, there are good reasons for having a vibrant conference series whose volumes appear rapidly and are easily available to a wide readership. 

The CPS did not enjoy high sales volumes and, therefore, the papers published in those volumes did not reach a significant readership beyond those who attended the actual conference. In the last decade, 17 volumes were published. In contrast, since 1993 the Fields Communications series has published the proceedings of 24 conferences and, although this is very subjective, some have expressed the view to us that the institutes are attracting the proceedings of the best of the Canadian conferences. 

We should not consider ourselves in competition with the research institutes in attracting conference proceedings. However, we feel there is still a valuable role to be played by the society in providing a vehicle for timely publication of conference proceedings. We return to the ideas of Andrew Odlyzko discussed in the introduction. He argues that ease of access is the crucial factor in determining volume of readership and that the electronic format is increasingly providing ease of access. As an aside, we observe that this is so in the developing world as well as, or perhaps even more than, the developed world. Information technology enables poorer countries to bypass the need for a large and expensive print collection. 

It has been suggested that the CPS could be revived in an electronic format. We find this suggestion has several merits: We have argued that articles would reach a wider audience; publication should be faster; conference organizers would be offered a clear choice of format for publication; and we have the expertise in our TeX Office and Camel to carry this out with relative ease. 

The new CMS Tracts in Mathematics series is planned to publish monographs that would be shorter than those usually considered by the Books Series and coherent collections of survey articles. Such a survey collection could come out of a conference. The tracts will be published in paper cover form at a reasonable price. 

The introduction of the CMS Tracts in Mathematics and an electronic Conference Proceedings Series will mean that the Society is providing the mathematics community with a comprehensive set of options for book publishing. 

Appendix

Table 1: Primary journals:

Table 2: Journals in Related Fields:

Table 3: Educational Journals:

Table 4: Formerly Canadian Journals