¹ Although the use of the term ‘predatory’ is widely used in scholarship, it has repeatedly been called into question (see Anderson, 2015; Eriksson & Helgesson, 2018). We have decided to privilege adjectives such as fraudulent, fake, or deceptive rather than predatory for two reasons. First, some legitimate commercial publishers can very much be considered to “prey” on researchers and libraries as they request high APCs or subscription fees (see Rentier, 2018, p.25), or both. Second, some authors publishing in fake journals “are aware that the journals do not adhere to accepted standards but choose to publish in them anyway, hence they are not ‘prey’” (Laine & Winker, 2017) ; they deliberately choose to publish in such fake journals for reasons that include fierce competition and a desire of career advancement (Shaghaei et al., 2018) .

² We are aware that using the terms blacklist and whitelist may have some racist overtones for some readers (Houghton & Houghton, 2018) . We would like to stress that our use of these terms is not meant to perpetuate racist culture. Rather, we use them because they are pervasive in the literature dealing with deceptive and fake academic publishing outlets. Moreover, our object of study in this paper uses this appellation (Cabell’s Blacklist).

³ Ulrich’s periodical directory covers more than 400,000 serial publications and is the source of bibliographic and provider information in Ulrichsweb™.

⁴ The last version of Beall’s blog as harvested by the Internet Archive on January 3, 2017, is available at: http://web.archive.org/web/20170103170903/ https:// scholarlyoa.com/ .

⁵ Crawford’s article count for 2016 takes only the January-June period into consideration.

⁶ Detail of codes can be found in the “Codes” tab appearing at the bottom of Crawford’s dataset (2016), which is available at: https://doi.org/10.6084/m9.figshare.4275860.v1 .

⁷ We would like to thank these primary data collectors: Olivier Borsus, Valérie Danthine, Ariane Ghislain, Thierry Jacques, Fabienne Prosmans, Sandra Rizzo, and Simona Stirbu.

⁸ To do so, you can apply the filter ‘yes’ in the column “Identical violations listed multiple times”. Alternatively, you can apply a filter in the journal column in inserting choosing one or several names of the journals mentioned in Figure 5.

⁹ The “no articles published” part of this violation also raises questions concerning Cabell’s possible reviewing of new authentic journals which may not have published any material yet. We assume that Cabell does not review newly launched journals but, for reasons of clarity, this may also be made explicit.

¹⁰ During our two-week trial period, we were not able to record details of violations for a significant number of journals. As a result, we were not able to determine whether or not some of Cabell’s criteria may never be used to blacklist journals.

¹¹ Detail of journals for which Cabell has failed to register a reviewing date can be found in our dataset using the filter “no date mentioned” in the “Last review date” column.

¹² Crawford (Crawford, 2017) argues that Shen and Björk’s numbers are grossly overestimated. While we agree with this, it should also be said that Crawford’s own numbers are probably underestimated considering that, as our study shows, a number of journals that he tagged as empty, and therefore did not take into consideration in his estimations, are not empty anymore.