January 4, 2004

Dear Friend of the Cognitive Science Journal,

In keeping with the last three years’ tradition, I am sending this annual editor’s report to the extended Cognitive Science family which includes reviewers, authors, editorial board members, and some selected other members of the cognitive science community. This report describes some recent changes to the journal and presents relevant statistics for 2004.

1) Changes and Updates
1A) Board changes
2005 will be the last year for the current editorial board. Our stint will be over on December 31, although some of the editorial board will most likely be asked to continue with the next editor. For authors this means that the next couple of months would be an excellent time to submit your manuscripts to us if you want to avoid getting caught between editorial boards. For potential editors, it is not too late to submit a nomination for the new editor of Cognitive Science. Here is the call for nominations:

The governing board of the Cognitive Science Society has opened nominations, including self-nominations, for the editorship of Cognitive Science. Candidates should have broad knowledge of cognitive science, and uphold the Cognitive Science Society’s goals of promoting an interdisciplinary science of minds and other intelligent systems. Desirable attributes for an editor include fairness, demonstrated editorial acumen, a strong research record, timeliness, diligence, dedication, and good judgment. The candidate should be available to receive manuscripts January 1, 2006. The initial term for the editor is three years, with an option to extend the editorship by an additional term. The governing board encourages participation by members of underrepresented groups in the publication process and particularly welcomes such nominations.

To nominate a candidate, prepare a support statement in one page or less and address it to the chair of the search committee: Douglas Medin. Informal inquiries can be emailed to medin@northwestern.edu. Nominations should be sent to the executive officer of the society, Thomas Ward, at editor@cognitivesciencesociety.org. The first review of nominations will begin February 1, 2005. Please submit nominations before then to assure full consideration.

In the meantime, the 2005 board will closely resemble the 2004 board, with only a few changes. The 2005 editorial board will be:

Associate Editors

John R. Anderson, Carnegie Mellon University
Nick Chater, University of Warwick
Marvin M. Chun, Yale University
Andy Clark, University of Edinburgh
Eve V. Clark, Stanford University
Gary S. Dell, University of Illinois, Urbana-Champaign
Shimon Edelman, Cornell University
Kenneth Forbus, Northwestern University
Dedre Gentner, Northwestern University
James Greeno, University of Pittsburgh
Dan Sperber, Institut Jean Nicod, CNRS, EHESS, and ENS, Paris
Joshua B. Tenenbaum, Massachusetts Institute of Technology

Board of Reviewers

Woo-kyoung Ahn, Yale University
Peter C-H. Cheng, University of Sussex
Susan Epstein, Hunter College and The Graduate School of the City University of New York
Cynthia L. Fisher, University of Illinois, Urbana-Champaign
Robert M. French, Université de Bourgogne
Adele E. Goldberg, Princeton University
Susan Goldin-Meadow, University of Chicago
Charles Goodwin, University of California, Los Angeles
Arthur Graesser, University of Memphis
Wayne D. Gray, Rensselaer Polytechnic Institute
Ray Jackendoff, Brandeis University
Frank C. Keil, Yale University
Stephen M. Kosslyn, Harvard University
Barbara Landau, Johns Hopkins University
Pat Langley, ISLE and Stanford University
Michael D. Lee, University of Adelaide
Richard L. Lewis, University of Michigan
Arthur B. Markman, University of Texas, Austin
Raymond J. Mooney, University of Texas, Austin
Terry Regier, University of Chicago
Steven Sloman, Brown University
Keith Stenning, The University of Edinburgh
James Tanaka, University of Victoria, British Columbia
Paul Thagard, University of Waterloo
Michael Tomasello, Max Planck Institute for Evolutionary Anthropology
Janet Wiles, The University of Queensland

1B) Special issues
Since 2001, the David E. Rumelhart Prize has been awarded annually to an individual or collaborative team making a significant contemporary contribution to the formal analysis of human cognition. The journal has instituted a policy of honoring Rumelhart prize winners by developing special issues of the journal around the prize winner’s research area and/or colleagues. In 2003 and 2004 we had special issues honoring the second and third Rumelhart Prize winners, Richard Shiffrin and Aravind Joshi. In 2005, Issue #3 will honor the fourth Rumelhart prize recipient, John Anderson. This issue will focus on recent advances in computational models using unified cognitive architectures such as ACT-R. The special issue authors include Rick Lewis, Marsha Lovett, Peter Pirolli, Dario Salvucci, Niels Taatgen, and of course Dr. Anderson himself. The fifth Rumelhart prize recipient is Paul Smolensky, and plans are afoot to have a special issue honoring him in 2006.

1C) Change of publisher
The Cognitive Science Society has signed a multi-year contract with Lawrence Erlbaum Associates to publish Cognitive Science starting with the first issue of 2005. The journal has never been published by LEA before, but in many ways, our relocation feels like a home coming. In 1976, Roger Schank approached Larry Erlbaum with the idea of forming a cognitive science journal. Mr. Erlbaum was highly receptive, and together with his partner Walter Johnson began publishing Cognitive Science in 1977 at the newly established Ablex Publishing Company. In the intervening years, LEA has published the Annual Proceedings of the Cognitive Science Conference. Beyond this, LEA has been a major force for the advancement of cognitive science through the impressive collection of books and journals they have published in the field. We are honored and delighted to be working with LEA. Lawrence Erlbaum, both the man and the company, has been exceptionally generous with his time, expertise, and support. Indeed, his contributions to cognitive science have been recognized twice by the Cognitive Science Society, as two independent governing boards have made him an honorary member of the society. As of August 2004, Lawrence Erlbaum has an official position on the Cognitive Science Society’s finances committee, but this position only formally acknowledges his long-time contributions to the society’s financial and intellectual well-being. We have every expectation that the journal will flourish with LEA. Through resources provided by LEA, the Cognitive Science Society will be able to significantly grow, offering its members new services for reasonable dues. LEA is also committed to the timely production of issues, Cognitive Science Society’s continued retention of copyrights, authors’ right to disseminate their articles on their own web sites, and continued growth of the journal.

2) Statistics (thanks to Caroline Verdier for assembling the critical data)
2A) Submissions
Submission rates to the journal continue to rise, reaching an all-time high level. In 2000, 2001, 2002, 2003, and 2004 we had 63, 126, 174, 188, and 210 submissions respectively. This increase will not come as a surprise to many of you on the editorial board. This increase, as welcome as it is as far as indicating the growing popularity of the journal, has certainly taxed our editorial and reviewer resources. Out of the 210 submissions in 2004, 136 were new manuscripts, and the remaining were resubmissions of previously submitted manuscripts. Of the 74 resubmissions, 40 were resubmissions of a manuscript previously given a “revise and resubmit” designation, and 30 were revisions from “accept with revision” designations, and 4 were prior “rejects.” Our number of new manuscripts only increased by 3 from 133 in 2003 to 136 in 2004. However, our resubmissions increased from 55 to 74. This is a natural demographic shift given that the current editorial board has been on duty for several years now, and several manuscripts have been wending their way toward the light of publication.

There are four major kinds of submissions: letters to the editor with about 1,000 words, brief reports with about 4,000 words, regular articles with about 12,000 words, and extended articles with about 18,000 words. Out of the 210 submissions in 2004, 2 were letters, 59 were brief reports, 115 were regular articles, and 34 were extended articles. The proportions of brief report in 2001, 2002, 2003, and 2004 were 18%, 26%, 29%, and 28% respectively. These numbers suggest that the word is finally out, and that people think of us when preparing brief, as well as regular-length, manuscripts.
The 2004 editorial board consisted of 12 associate editors and 25 review board members. I personally handled 54 manuscripts in 2004, and the associate editors handled approximately 13 manuscripts each. Each review board member reviewed an average of 8 manuscripts, and we had 214 ad hoc reviewers, many of whom reviewed multiple manuscripts. The increased submission rate has put additional strains on our editorial team, but I have been grateful and proud of their professionalism, thoughtfulness, and generosity in providing intelligent judgments and comments on manuscripts.

2B) Acceptance Rates
For accounting purposes, we classify every manuscript into one of four categories after an editor has acted on it: accept, accept with revision, revise and resubmit, and reject. Grouping the first two categories together, and counting each submission separately (i.e., counting a resubmission of a manuscript as a separate submission), the journal had an acceptance rate of 21% for manuscripts submitted in 2001, 28% in 2002, 29% in 2003, and 38% for the 158 manuscripts submitted in 2004 for which decisions have already been made. These acceptance rates include submissions for special issues, and so are somewhat artificially elevated. Although it seems as though our acceptance rate has increased dramatically from last year, this is because in 2004 we received a much larger proportion of revised manuscripts, and revisions have a substantially greater probability of being accepted. Also, all of these probabilities are elevated because many manuscripts are first conditionally accepted and then finally accepted. For this reason, a more appropriate way of calculating the probability of acceptance is to track each manuscript longitudinally. Using this method, the probability of a manuscript eventually being published is approximately 30% including invited special issue submissions, or approximately 24% if special issue submissions are ignored.
In 2002, there was an apparent trend for brief articles to have a lower acceptance rate (13%) than regular articles (35%) or extended articles (35%). This trend was not at all apparent in 2003, with brief reports, regular articles, and extended articles having acceptance rates of 29%, 24%, and 26% respectively. In 2004, the trend has reversed, with acceptance rates for brief, regular, and extended submissions at 43%, 34%, and 36% respectively. This may be another indication that brief reports have finally been accepted as part of the norm for the journal, by both authors and editors alike.

2C) Lag Times
The editorial lag time is defined as the number of days between when the on-line submission software registers a submission as complete (when the authors have supplied the manuscript, keywords, and manuscript designation) and when the software sends the author the completed editorial decision, action letter, and reviews. The table below shows the average lags in days:

The average editorial lag was 59.4 days. For the first year since I have been editor, in 2004 our average lag times for all submission categories were within our target ranges. For this accomplishment, I can only thank the associate editors and reviewers for their diligence and timeliness.
Another kind of lag that is relevant is the lag to print. We calculate print lag time as the number of days intervening between when we receive the final version of the manuscript to two weeks after the scheduled dispatch date (a date by which most people have received the journal). The overall average is 156 days, broken down by article category as follows:
Brief reports (9 manuscripts): 156 days
Regular articles (22 manuscripts): 162 days
Extended articles (3 manuscripts): 139 days

Although these lags may seem long, they are quite favorable compared to other journals, and are largely attributable to copy editing, assembling supplemental annex materials, final author changes, and author proofing. These lags are only minimally due to the backlog of articles waiting to be printed. Our average backlog for 2004 was only about 9 articles, up only slightly from 2003’s backlog of 6 articles. Another factor to consider is that the articles from 2004 were available on-line at Elsevier’s Science Direct service well before they appeared in hard-copy form. We will have a similar arrangement in 2005 with LEA using their on-line service. After an author has proofread the PDF file created by the copy editor, it appears on-line to all members of the Cognitive Science Society, or any member of an institution with an on-line subscription to the journal. If you have a personal subscription to the journal, then you can access articles in press at http://www.leaonline.com/. This site will also direct you to PDF versions of all published articles from 1995-present. LEA will soon be sending members of the Cognitive Science Society instructions on how to access Cognitive Science articles via LEA-Online. In addition, you can access articles from 1995-2004 via Elsevier’s site at http://www.elsevier.com/gej-ng/10/15/15/show/.
Another indication of our journal’s efficiency is provided by author comments. Andreas Ortmann and Peter Todd have put together an independent web site they call “Academic Journals Feedback Project” that allows authors to post their own comments about journals specializing in economics and psychology: http://www.cerge-ei.cz/multiversity/JFeedBack/index.asp. On this main page, if you select “psychology” as the field and press “submit” you will be able to look up authors’ feedback for Cognitive Science and many of its peer journals. By my admittedly biased eye, authors are fairly positive about their dealings with Cognitive Science.

2D) Topic Areas
We ask all of our associate editors and members of the board of reviewers to supply keywords for their domains of expertise and we also ask authors to supply keywords that fit their manuscripts. Rather than try to define whether cognitive science is characterized by its contributing set of fields, content topics, or methods, we include all three classification schemes in our list of keywords. The table below represents counts of individual members of the 2005 editorial board (listed as “Board”) and submitted manuscripts from 2001-2004 in the right 4 columns. Thus, for example, 14 members of the editorial board (out of a total of 39 board members) responded that they would be comfortable reviewing manuscripts in the field of artificial intelligence, and in 2003 and 2004 we had 35 and 28 manuscript submissions related to artificial intelligence, respectively. If our editorial board matches the manuscripts well, then we would expect the first column’s frequencies to be in roughly equal proportions to the last three columns. To a first approximation we find this to be true, although there are some possible exceptions. Taking an arbitrary criterion of the ratio of submissions to editorial board members exceeding 4 and a frequency of more than 8, the following keywords are underrepresented by the editorial board: biology, neuroscience, cognitive development, intelligent agents, motor control, perception, and computational neuroscience. Two of these terms, neuroscience and cognitive development, appeared on our list of underrepresented areas in 2002 and 2003 as well. At the very least, when the next editorial board is gathered for 2006, particular attention should be paid to these areas.

A perennial concern for the journal and conference is that cognitive science may be too dominated by psychologists. This is a concern not because of anything intrinsically wrong with psychology, but because many people read the journal to discover articles written for and by practitioners in a wide range of fields. This worry is not alleviated by an inspection of the table. Of the main fields that comprise cognitive science, psychology is by far the most often listed by authors: psychology (listed as a keyword by 52% of submissions), linguistics (15%), artificial intelligence (13%), neuroscience (10%), philosophy (8%), and computer science (7%) [the full set of field percentages sum to more than 100% because we do not restrict authors to designating only a single field]. I raised the same concern in last year’s editor’s report, and received a mixed response. Not all readers felt that the preponderance of psychology submissions was a negative occurrence or that proactive measures should be taken to more forcibly balance the submissions. I would be curious about other readers’ opinion, which I will pass along to my successor.

Fields                        Board         2001    2002     2003     2004
Anthropology                   5             1       2        6        3
Artificial Intelligence        14            21      25       35       28
Biology                        0             1       3        7        8
Computer Science               8             12      18       28       14
Education                      9             10      8        13       7
Linguistics                    12            13      23       37       32
Neuroscience                   5             6       20       24       21
Philosophy                     8             8       13       25       17
Psychology                     31            69      113      127      109

Topics:
Analogy                        10            3       14       9        5
Animal cognition               1             3       3        5        7
Attention                      3             5       9        10       11
Artificial Life                4             3       1        1        3
Case-based reasoning           7             4       1        0        0
Causal reasoning               13            6       8        11       9
Cognitive architecture         16            16      24       32       29
Cognitive development          4             16      18       29       20
Communication                  6             5       10       10       9
Complex systems                5             5       7        11       0
Computer vision                3             3       7        2        2
Concepts                       20            14      16       31       19
Consciousness                  5             1       4        13       8
Creativity                     4             3       5        5        4
Culture                        3             1       4        11       4
Decision making                9             11      15       19       8
Discourse                      3             3       11       15       9
Distributed cognition          6             4       13       10       8
Emotion                        3             2       0        6        3
Epistemology                   4             6       8        11       4
Evolutionary psychology        3             4       3        8        9
Expertise                      3             2       5        6        5
Human-computer interaction     7             4       5        7        6
Human factors                  3             1       2        4        4
Information                    1             3       5        11       7
Instruction                    6             7       4        5        3
Intelligent agents             2             8       4        8        11
Language acquisition           12            13      12       11       12
Language understanding         12            6       24       37       23
Learning                       14            25      25       35       24
Machine learning               10            7       11       11       8
Memory                         6             13      20       30       13
Motor control                  2             4       2        4        9
Music                          1             0       1        1        1
Pattern recognition            6             7       11       12       8
Perception                     4             12      23       22       25
Philosophy of computation      3             1       0        5        1
Philosophy of mind             6             8       8        19       11
Philosophy of science          8             3       5        16       3
Pragmatics                     2             2       7        6        3
Problem Solving                9             21      12       19       12
Reasoning                      14            19      22       29       23
Representation                 26            25      41       42       32
Semantics                      9             6       17       24       16
Situated cognition             4             4       8        9        11
Skill acquisition/learning     4             11      10       10       7
Social cognition               3             4       6        15       10
Speech recognition             1             2       1        5        3
Syntax                         4             6       7        12       5
Translation                    0             0       0        1        0

Methods:
Animal experimentation         1             1       3        1        3
Clinical methods               1             0       2        2        1
Computational neuroscience     2             1       6        9        9
Computer simulation            17            26      46       24       28
Cross-cultural analysis        3             0       1        5        3
Cross-linguistic analysis      3             0       2        1        1
Developmental experimentation  8             7       9        11       4
Electroencephalography (EEG)   2             0       2        0        4
Ethnography                    1             0       0        2        2
Human experimentation          24            28      48       65       52
Knowledge representation       12            16      28       33       25
Logic                          2             3       6        12       8
Mathematical modeling          8             13      23       22       10
Neural Networks                12            14      29       23       24
Neuro-imaging                  2             2       2        3        3
Robotics                       3             2       1        4        3
Single-cell recording          0             1       0        1        1
Statistics                     8             7       5        13       7
Symbolic modeling              11            10      11       13       15

Another way of capturing the some of the thematic trends in Cognitive Science is by a conceptual map. The concept map below reflects the ISI-assigned keywords for all articles published in Cognitive Science from January 2001 to August 2004. The size of a node reflects the number of times it was given as a keyword by an article. The thickness of a link shows the frequency with which two keywords were paired together in articles. Only keywords listed by more than one article appear in the map. This map was created by Weimao Ke and Katy Börner at Indiana University, and I thank them for letting me reproduce it here. Among other trends, the map shows the large role played by psychology and computational models in the journal, and also the strong link between the two methods.

2E) Quantitative Measures of Importance
Quantitative estimates of a journal’s importance are notoriously controversial. At the very least, they are important because people think they are important. The most common reported measures of a journal’s importance are ISI’s (http://www.isinet.com/isi/) Impact Factor and Cited Half-life. The two measures complement each other because the former measures the short-term impact of a journal’s articles while the latter measures long-term impact.

More specifically, the impact factor for a given year is calculated by dividing the number of citations in the given year to articles published in the two previous years by the total number of articles published in the two previous years. For example, Cognitive Science’s 2003 impact factor is obtained by dividing the number of citations to Cognitive Science articles appearing in 2001 and 2002 by the total number of articles appearing in 2001 and 2002.
Across all sciences, impact factor is assumed to be the most important factor in assessing a journal. However, in my opinion, this should only be the case for journals in fields that are so fast-moving that articles more than two years old are only marginally relevant to current research. It is important not to judge the field of cognitive science by criteria created for biomedical sciences and molecular biology. After all, one of the core fields comprising cognitive science is philosophy, where even 2300-year-old writings by Plato are considered pertinent to contemporary theorizing.
Cognitive Science’s recent impact factors are as follows:

1997: 1.53
1998: 1.96
1999: 2.27
2000: 1.63
2001: 2.44
2002: 1.52
2003: 2.67

In last year’s report, I reported that the low 2002 impact factor was due to a one-time change from our journal being published 4 times to 6 times per year. Despite my distrust of the impact factor measure, I was relieved that the one-time dip was indeed one-time. Our 2003 impact factor is the highest that I have record for, and places us as #10 out of 67 journals in our ISI-designated field of Experimental Psychology. Yes, I realize that this designation hardly does justice to the broad range of topics covered by Cognitive Science, but ISI insists on placing each journal into only a single category. Somebody should inform ISI that science is bipartite graph, not a hierarchy. Sanctioning the kind of cross-categorizations that ISI forbids, if we were in ISI’s category “Artificial intelligence,” our impact factor ranking would be #11 out of 77 journals, #1 out of 92 journals listed under “Education and Educational Research,” #2 out of 53 “Anthropology” journals, and #6 out of 46 “Multidisciplinary Sciences.”

One of the obvious limitations of the impact factor measure is that citations to articles that were published prior to 2001 have absolutely no influence on a journal’s 2003 impact factor. Given this obvious blindspot of the impact factor, it is useful to consider cited half-life, defined as the number of publication years from the current year which account for 50% of current citations received. For example, in 2003, 39% of the citations to Cognitive Science articles were to articles appearing between 1994 and 2003. The remaining citations were to Cognitive Science articles published at least 10 years prior to 2001. ISI has a single category “>10” that is used for all journals with a cited half-life of 10 years or more. With this long preamble out of the way, here are the 2003 impact factors for Cognitive Science and some journals that I selected as being within Cognitive Science’s peer group in a variety of disciplines covered by ISI.

Journal                            Impact Factor      Cited Half-life
Adaptive Behavior                     0.14                 7.0
Artificial Intelligence               2.45                 9.6
Brain and Language                    1.32                 8.8
Cognition                             4.30                 9.3
Cognitive Psychology                  3.68                 >10
Cognitive Science                     2.67                 >10
J. Artif. Intell. Res.                1.62                 6.4
J. Cognitive Neuroscience             5.01                 5.1
JEP:LMC                               2.52                 9.1
J. of Math. Psych.                    0.74                 >10
J. of Learning Science                1.60                 7.4
Language                              0.79                 >10
Learning and Instruction              1.30                 6.5
Machine Learning                      3.05                 7.4
Memory & Cognition                    1.41                 10.0
Minds & Machines                      0.34                 6.6
Neural Computation                    2.75                 6.4
Psychological Review                  8.36                 >10

In terms of these and other measures of influence, Cognitive Science looks very good indeed. The downloads of articles from Elsevier’s on-line ScienceDirect site were 5000, 12000, 20000, and 55588 in 2000, 2001, 2002, and 2003, and 19456 in the first four months of 2004, although this may speak more to the exponential growth in electronic downloads of scholarly content than it does Cognitive Science’s popularity per se. Still, in the first four months of 2004, the average number of downloads per on-line article were an impressive 50. If you are curious, the most downloaded Cognitive Science articles from 2003 were:
H. Cruse: “The evolution of cognition – a hypothesis”
R. Shiffrin: “Modeling memory and perception”
P. Verschure and P. Althaus: “A real-world rational agent: Unifying old and new AI”
W. Schneider and J. Chein: “Controlled and automatic processing: Behavior, theory, and biological mechanisms”
J. Waskan: “Intrinsic cognitive models”
S. Ainsworth and A. Loizou: “The effects of self-explaining when learning with text and diagrams”
A. Norenzayan, E. Smith, B. Jun Kim, and R. Nisbett: “Cultural preferences for formal versus intuitive reasoning”
S. Dasgupta: “Multidisciplinary creativity: The case of H. A. Simon”

3) Appreciation
In wrapping up this annual summary, I would like to personally thank our associate editors, the board of reviewers, and our ever-growing assembly of ad-hoc reviewers. The dedication, wisdom, and thoughtfulness of editors and reviewers drive the journal, and is the basis for its continued vitality. I thank the authors for considering Cognitive Science as an outlet for exciting developments, for their receptivity to suggestions, for the patience that we have had to occasionally tax, and for assuring that Cognitive Science will continue to stay on the forefront of innovations in the interdisciplinary study of minds and other intelligent systems. In 2004, the journal also benefited greatly from the professionalism, expertise, and proactive attitude of Elsevier’s crew, including Fiona Barron, Diana Jones, Niamh O’Hea, Elizabeth Purcell, and Mary Sheedy. In my dealings with LEA, I have been greatly impressed with their crew’s can-do spirit and creative problem solving, which includes Steven Cestaro, Jim Conroy, Lawrence Erlbaum, Rick Maffei, Joseph Petrowski, Steve Rutberg, and Nancy Seitz. James Greeno (the ex-editor of Cognitive Science), Bill Bechtel (the past chair of the publications committee of the Cognitive Science Society), Douglas Medin (the chair of new editor search committee, and hence my immanent liberator), Art Markman (the ex-executive officer for the society and continued spiritual guide), and Tom Ward (the current executive officer for the society, whom I bothered no fewer than 6 times with queries bearing on this annual report alone) have been very generous with their time, good ideas, know-how, and enthusiasm. Finally, I am indebted to Caroline Verdier, the managing editor for the journal, who has done a wonderful job of seeing to the daily care and feeding of the journal, as well as being mindful of plans for its long-term development.

Best wishes for a Happy and Thoughtful 2005,

Rob Goldstone, Executive Editor of Cognitive Science