2019Q3 Reports: CL Journal Editor

In the past 6 months, we have received 85 first submissions and the average time to first decision for the papers first submitted (excluding desk rejects) was 108 days. The journal has now developed a pipeline of accepted and conditionally accepted papers to fill at least two upcoming issues (till December 2019). The numbers of published papers in the last four issues of CL journal in each of the paper categories are as follows: regular papers (14), special issue papers (7), survey papers (1), squibs (4), and book reviews (2).

The special issue on “Multilingual and Interlingual Semantic Representations for Natural Language Processing” received 14 submissions by the submission deadline of 21 March 2019. The guest editors of this special issue are Marta R. Costa-jussà, Cristina España-Bonet, Pascale Fung, and Noah A. Smith. First-round notifications for all submissions have been sent out. We are targeting the publication of this special issue in the March 2020 issue of CL journal.

In order to shorten the reviewing time of CL submissions and to solicit the views of ACL members on presenting CL papers at conferences, a survey was conducted during 24 – 30 June 2019. A total of 258 ACL members responded to the survey, which comprises three questions:

Reponses to Question 1: “When deciding whether to submit a paper to CL Journal, how important to you is the length of the reviewing time taken by the journal?” Scale: 5 (very important) to 1 (very unimportant) Responses: count (percentage) 5: 121 (47%) 4: 91 (35%) 3: 34 (13%) 2: 7 (3%) 1: 5 (2%)

Responses to Question 2: “Would you be willing to serve as a standing reviewer for CL journal?” Responses: count (percentage) Yes: 188 (73%) No: 70 (27%)

Responses to Question 3: “If you have a paper accepted by CL Journal, how interested are you in presenting it at one of the *ACL or EMNLP conferences (i.e., in a similar arrangement as TACL)? Note: Presentation at these venues is not compulsory.” Scale: 5 (very interested) to 1 (very uninterested) Responses: count (percentage) 5: 173 (67%) 4: 49 (19%) 3: 24 (9%) 2: 7 (3%) 1: 5 (2%)

The survey outcome indicates that 82% of the respondents view fast reviewing time as important or very important; 188 respondents volunteer to serve as standing reviewers; and 86% of the respondents are interested or very interested in presenting their CL journal papers at conferences.

Based on the survey outcome, CL journal will implement the following changes:

1. Set up a pool of standing reviewers to speed up reviewing time, which is valued by most respondents.

2. Since reviewing will now be done by the standing pool of reviewers, the responsibility of the editorial board members will no longer be reviewing papers. Instead, editorial board members will become action editors whose role is like area chairs in a conference. That is, action editors assign reviewers to submitted papers, and recommend decision on a submitted paper to the editor-in-chief, whose role is like a program chair in a conference. The name of the action editor who recommends the acceptance of a paper will be listed on each published CL journal paper.

We also plan to work towards allowing CL journal papers to be presented at *ACL/EMNLP conferences for authors who are interested in presenting their accepted papers (not compulsory), since this is what most respondents preferred. We propose the following two categories of submissions to CL journal:

(A) A submission is an expanded version of previous conference publications. In this case, we use single-blind reviewing (since the authors will need to openly declare their previous publications), and an accepted paper of this kind will *not* be given a presentation slot at *ACL/EMNLP (since part of its content could have been presented at a prior conference).

(B) A submission is new and none of its content has been published before (similar to conference submissions). In this case, we use double-blind reviewing, and an accepted paper of this kind will be given a presentation slot at *ACL/EMNLP (oral or poster presentation) if the authors so desire. Requiring double-blind reviewing for this category of CL submissions would mean that all papers presented at *ACL/EMNLP conferences will have undergone double-blind reviewing (i.e., treated in the same uniform manner).

Hwee Tou Ng 28 July 2019