Sign In

Communications of the ACM

Point/Counterpoint

The Pros and Cons of the 'PACM' Proposal: Counterpoint


View as: Print Mobile App ACM Digital Library Full Text (PDF) In the Digital Edition Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook
The Pros and Cons of the 'PACM' Proposal, illustration

Credit: iStockPhoto.com

ACM has been a longtime leader and innovator in the ecosystem of computing publications. ACM produces journals and proceedings of the highest quality, and it continually finds ways to address the publishing needs and concerns of the computing research community, most recently with its innovative Author Rights framework. It is therefore with no small amount of trepidation that I write to argue against the creation of a new proceedings-focused journal series.

In their editorial, Konstan and Davidson identify two fundamental problems motivating the need for a new journal series. First, the merit and funding systems of some countries do not accord conference publications the status they deserve. Second, conference publication has hard deadlines, page limits and limited review, and it exhausts the reviewing capacity of the best researchers. I concede these problems exist to some extent. I also concede the proposed journal series, by virtue of being journals, might help circumvent some of the problems. However, I feel the proposed journal series fails to attack the root causes of these problems.

With respect to the first problem, I feel it is counterproductive for our community to distort itself merely in order to live up to a false image devised by non-experts who refuse to understand and accept the way our community works. (As Emerson wrote, "to be yourself in a world that is constantly trying to make you something else is the greatest accomplishment.") Konstan and Davidson rightly note the NRC and CRA have been very successful in educating U.S. institutions (and many international institutions, I would add) on the importance of conference publications.a Thus, rather than create a new journal series, a more natural way to attack this problem is for ACM, as the world's leading professional society in computing, to lobby vociferously on the community's behalf about the importance of conference publications.


"I feel it is counterproductive for our community to distort itself merely in order to live up to a false image devised by non-experts who refuse to understand and accept the way our community works."


In addition, I feel it is overly simplistic to say improper recognition of conference papers is a problem of countries. The reality is far more complicated, and it relates less to countries than it does to the relative quality differential between institutions, the degree of enlightenment among institutional administrators, the additional expectations placed on researchers at some institutions (such as explicit annual publication quotas), and the level of readiness (and unreadiness) of authors to publish to ACM's high standard of quality.

Regarding the second problem, several subcommunities have been experimenting successfully with approaches to marrying conferences with journals in a way that mitigates some of the problems of conference publication. SIGGRAPH/ACM TOG and HiPEAC/ACM TACO are two prominent examples in the ACM family, and ACM TOSEM and IEEE TSE are embarking on a "journal-first" initiative in conjunction with some of the major software engineering conferences. These examples show the problem can be addressed with existing journals, making it unclear what additional benefits this new journal series would bring. Even so, tying journal publication to annual conference cycles arguably still encourages an annual cycle of submissions from authors (particularly authors operating under quotas), even when the related journal and conference do away with fixed deadlines (as in the case of VLDB).

One big problem not mentioned by the co-chairs is the continued proliferation of publication venues in our community. A new journal series will only further exacerbate this problem, without any obvious compensating benefit.

But all these problems are closely bound up with many additional drivers of the difficulties we face, including increasing pressure on researchers to produce, and increasing demands from research-funding agencies to plan for and quantify impact before new research even begins. Conferences now routinely receive numerous submissions from the same author in the same year (with one author having submitted 11 papers to a recent software engineering conference), which begs the question of just who exactly is complaining that conferences are too deadline-driven!

I am a strong believer in the power of incentives. As a community, we need to step back, return to first principles, and decide for ourselves what behavior we desire, and then develop appropriate incentives to produce that behavior. Do we want to encourage more journal publication? Do we want to devalue the esteem associated with conferences and conference reviewing? These are things we need to examine deeply before creating yet another series of journals that need to be nurtured, grown, and sustained.

Back to Top

Author

David S. Rosenblum (david@comp.nus.edu.sg) is a professor of computer science and Dean of the School of Computing at the National University of Singapore.

Back to Top

Footnotes

a. In fact, I would argue we find ourselves in our present predicament of dying journals and overloaded conferences in part because of that success.


Copyright held by authors.

The Digital Library is published by the Association for Computing Machinery. Copyright © 2015 ACM, Inc.


Comments


R Oldehoeft

As a department chair I was able to educate a Dean of Natural Sciences about the quality of CS conference publications, widely adopted software distributions, and the ways that CS research differs from work in other sciences. It can be done.

However, the proliferation of low-quality conferences threatens to pollute the status of CS conference publication. At the least, ACM might maintain a list of conferences that meet (still to be developed) criteria as being worthy for the purposes of evaluating conference publications. But do not tie that to ACM's involvement in a conference--I have seen that contribute inordinately to the workload of conference organizers.

Rod Oldehoeft
Emeritus Chair and Professor
Computer Science Department
Colorado State University


Shriram Krishnamurthi

Preach it, brother!

The one major point I wish you had made more strongly is that our journals have a genuinely different (and better) review model than our conferences, and if we equate the two, we will not "average" them or anything else: we'll simply have conference-quality reviews (and I definitely mean to put scare-quotes around the word quality).

I think there are three convergent streams here:

1. People who genuinely care about people who're stuck in what I can only term as "dumb" countries (because they're needlessly hurting their own scientists and science).

2. People who genuinely believe journals are no better, maybe even a bit worse, and anyway not really different, from conferences. I suspect many of them haven't written (or read) a true journal paper in a very long while.

3. People who're cynical and using these other reasons on the surface to get rid of what they see as a nuisance.


Displaying all 2 comments