Why Books are Important in the Scholarly Communication System in Social Sciences and Humanities

Books are relevant outputs in research in Social Sciences and Humanities (SSH) and need to be considered properly in research evaluation. Books are different from journals and the scholarly book publishing sector remains quite unknown within evaluation agencies. This paper describes the state of the scholarly publishing sector, beyond the large publishing groups and imprints, focusing the attention to the role of smalland medium-sized scholarly publishers at the national level. Bibliodiversity and multilingualism are essential values for representing plural and diverse SSH research. They are also linked to societal impact. Academic Book Publishers project and Helsinki Initiative are described to this regard. Finally, some remarks and recommendations for addressing the assessment of books are provided.


Introduction
Books are an integral part of the scientific communication in the Social Sciences and the Humanities (SSH hereafter). They are the natural space for much of the research carried out in those fields which, usually, do accumulate knowledge and hence, give rise to new advances .
Many indicators presented in various studies and developed in different periods underline the central role of books in the SSH. But before presenting data, an expressive analysis has to be done. A simple review of the catalogues of scholarly publishers shows that there are research topics which naturally adopt the form of books, which are ' called to be' books and which are not well adapted to other communication channels without losing their meaning or becoming superficial.
Consider the following subjects and publishers: a. Kinship and Gender (Routledge) b. Storia dell'Adriatico (Il Mulino) c. Docencia y cultura en el exilio republicano español (UNAM) d. El vidrio en la pintura del Museo Nacional del Prado (Editorial CSIC) e. La despoblación del mundo rural (Aranzadi) f. La mirada de la Justicia (Antonio Machado Libros) The titles from this list, including their treatment of topics, do not seem to be suited for publication as a journal article. Because of their approaches, chronological periods under study, depth and degree of analyses, and the amount of materials each will cover, they do not fit well with the structure of a scientific paper. The book allows many SSH research topics to be covered through a state-of-the-art analysis, corresponding contextualization, the development of theories, detailed perspectives, with a reproduction and/or partial quotation of original information sources, including reproduced images or other graphic materials. According to Clemens et al. (1995), a book may be considered as a 'genre', restricted to few scholars, established and reputable researchers, with a long career, and representative, thus, of a certain 'status' in the discipline. Moreover, "The option of publishing books or using them as preferred sources in research is not a whimsical choice: it is a necessary choice which responds to the requirements of research in these areas" (Giménez Toledo, 2013).
Some basic notions underlying this article are: 1. SSH need a multiform communication system as a reflection of the diverse and plural nature of SSH and its object of research, in particular a system that acknowledges the multi-lingual nature of SSH. 2. Monographs and edited books are essential communication media in SSH, that are fit for purpose in the process of knowledge production in SSH. 3. Assessment of SSH research needs multi-dimensional research evaluation methods that do justice to the pluralist nature of SSH.

Academic books
Several studies on the publication patterns have been showing for a long time ago that in the Social Sciences and the Humanities the publication of books, book chapters and articles in national scientific journals is fundamental, and numerically prominent (Broadus, 1971;Cullars, 1992;Thompson, 2002;Hicks, 2004;Franceschet, 2011;Leydesdorff, L., & Felt, U., 2012;Engels et al., 2012;Gorraiz et al., 2013;Kousha et al. 2013;Zuccala et al. 2015, Tanner, 2016, among many others). It is, in fact, a distinctive feature of the communication of research results in these disciplines. A review of the literature which has shown how the book does occupy a relevant place in the scientific output of the Humanities and the Social Sciences is given in the introduction of a study by Giménez Toledo et al. (2016). More recently, an European study analyzing the publication patterns in eight different countries  using complete databases (Current Research Information Systems, CRIS), has established that an average of 56,5% of the research results in SSH are journal articles, while books are a 43,5%. That percentage includes book chapters (33,4%), edited volumes (4%) and monographs (6,2%).
Since there is no CRIS at the national level, Spain did not participate in that study, but taking into account the data from the Spanish National Research Council -the largest research institution in the country-it can be observed that, for the SSH area in 2018 (CSIC, 2019;Bordons et al., 2018), the output of book chapters (430) was very similar to that of articles (455). The sum of chapters and monographs amounts to 54% of the output in SSH. As a point of comparison, books and chapters are only 4% of the output in Biology and Biomedicine, within that same institution. Also, the only report from the CRUE (Conference of Rectors of Spanish Universities) which included the totality of the scientific output of Spanish universities by compiling data from each of them, yielded similar percentages (Michavila, 2012). In the Arts and the Humanities 56% where book publications (including both, books and book chapters) and in the Social Science, this percentage was 47%.
The data prove, once again, that a book holds weight in those fields. They also show that the publication patterns are intrinsically linked to the nature of the fields, and also that the evaluation policies can modify them only up to a certain degree, but are unable to change them in a fundamental way. Although publications in indexed scholarly journals have been prioritized in many countries, the few diachronic studies that exist on the weight of books in the scientific output of a country or region are conclusive. In a study concerning scientific production in Flanders (Engels et al., 2012), as well as a more recent one, pertaining to books published in five European countries , findings show that the production of books and book chapters remains stable. Only in the case of Poland a 'reactive effect' has been identified with regards to the evaluation policies, which count with negative incentives for book publications (Kulczycki, 2018).

The scholarly book sector
Publication patterns show the relevance of the book in the SSH. Here, we reflect upon certain aspects of the publishing sector that support the publication of scholarly books, both at the national and international level, and which are the implications of that characterization for the research evaluation processes. The report The Global Publishing Industry in 2018 (WIPO, 2020) points out "The scientific, technical and medical (STM) sector represents the most complex and internationalized sector in the global publishing industry. It is dominated by a number of large international companies, incorporated in different countries". These publishers have a very large editorial production, both in journals and in scholarly books. In fact, scientific journals are an essential pillar of the academic and scientific publishing market. The journals are so relevant that they often blur the analysis on books. The report points out the difficulties of a global analysis of the scholarly publishing sector given the lack of comprehensive publication databases, and, hence, it does only present an analysis on journals in the Web of Science, not including data on scholarly books.
Roughly, without intending to present an exhaustive taxonomy of publishers, it can be said that the scholarly publishing sector is made up of: a) large publishing companies and scholarly imprints which are highly relevant in the international framework; b) university presses; c) big publishers at the national level or regional level; and d) small and medium-size independent publishers from each of the countries.

Large publishing companies and scholarly imprints
The report The Global 50 World Ranking of the Publishing Industry annually identifies the largest publishing companies in the world based on revenues obtained in previous years. In its 2019 edition, the following publishers are among the 20 first publishing companies active in the publication of scholarly books: RELX Group (Reed Elsevier), Pearson, Thomson Reuters, Wolters Kluwer, Springer Nature, Wiley, McGraw Hill and Oxford University Press. These and other similar publishers include scholarly imprints which jointly account for an important part of the publishing business and might be defined as global publishers. Their production volume, market position, marketing strategies, digital and open publication programs or their deals with libraries allow them to lead the scholarly publishing sector worldwide and that influences directly in the prestige they hold in academia in general. These publishers have their headquarters in the United Kingdom, The Netherlands, United States or Germany, and they publish mainly in English.

University presses
University presses are a relatively heterogenous group, although in this taxonomy they are presented as a group since they are linked to a given university and, therefore, to its mission. They are present in all countries worldwide and publish in various languages about local, regional or international topics. There are very small publishers linked to small universities as well; middle-sized ones but with ambitious editorial projects and models; and large publishers, based on various publishing models, operation and funding schemes. To the latter group belong publishers such as Oxford University Press and Libros de la Universidad Nacional Autónoma de México (Books from the Autonomous National University of México).

Big publishers at the national level or regional level
Imprints such as Cátedra in Spain (Hachette Livre Group) or Fondo de Cultura Económica in Mexico are very relevant publishing projects in the Spanish-speaking area. Those are very consolidated and prestigious imprints in the region, they publish mainly in Spanish, are influential and have good distribution and marketing structures.

Small and medium-size independent publishers
Finally, it is possible to distinguish a large group of scholarly publishers comprised of small and medium-sized private publishers, established in each country, which includes an heterogeneous set of publishing projects and business models but having some features in common: they publish in the language or languages of the country, usually focused on topics of local, regional and national interest, and they are nurtured by original works produced by researchers in the country. They make a significant contribution to the GDP.
In the case of Spain, scholarly books are 17% of all registered ISBN . In Colombia, the publications by University Presses represent 27.7% (Unilibros, 2019) and in Latin America as a whole, 13% (CERLALC, 2018). It is also worth taking into account that the book industry is the second most relevant cultural industry in the world.

Bibliodiversity in scholarly publishing
A review of this taxonomy is relevant for undertaking an evaluation of books, book series and/or publishers which constitute an important part of SSH outputs. Know the publishing sector of scholarly books is key in order to value the various types of scientific output, for establishing evaluation models, and for creating support tools (metrics, among others) in research evaluation processes. Lack of insight implies practically that there is going to be an adequate recognition for those books published by the large publishing companies with an international profile but not for the more national oriented and smallest ones. The analysis of the publishers in which the humanists and social scientists publish their works offers a surprising vision, probably unknown to many science policy managers. It allows deepening into the scholarly publishing segment and its diversity, established not only as per the number of publishers, but also in terms of size, typology, catalogue features and publishing languages. It allows observing, in sum, the bibliodiversity and the diversity of topics and the plurality of approaches that it implies. Such bibliodiversity is an identity sign of the research carried out in the Humanities and the Social Sciences, since the objects of study can be addressed from multiple perspectives and approaches, with different methodologies, in different languages and on international, national and local topics. Scholarly books respond to all the aforementioned features.
Tanner (2016) showed surprise when he found so many types of different publishers in the analysis of the contributions by British researchers to the REF2014. In his report, he shows that 8513 published books were sent for evaluation, published by 1183 different publishers, although the distribution is very unequal: 46% of the books were published by 10 publishers only, whereas the remaining 54% is distributed among a long tail of different publishers.
Such bibliodiversity can be clearly observed in the data analysis carried out on some European Current Research Information Systems (CRIS). The publication channels used by researchers allow identifying a significantly varied publishing landscape: 203 publishers in Norway, 231 in Italy and 137 in Denmark. 1 Studies on publishers' prestige carried out in different countries confirm this outcome (Giménez Toledo, 2013 andBorges de Oliveira, A., 2018). In the Spanish case, consulted humanists and social scientists identified 505 Spanish publishers relevant for their research, whereas in Brazil the number of Brazilian publishers was 447, and 167 in the case of Colombia. In sum, beyond scholarly publishing with an international profile there is a varied and diverse publishing landscape in each country which guarantees the richness of the topics and approaches.
A relevant example of the different publication profiles of the various publishers is provided by the study developed for the entire scholarly publishing activity in Spain (Giménez Toledo, ed., 2019). From a metadata analysis of all published scholarly books it was possible to observe (Figure 1) the volume of production in the various scientific disciplines, and also the differences in the publication pattern between the university presses and the commercial/private publishers. For example, some university presses share with private/commercial publishers their focus on large fields such as History, Law, Education or Economy. But also university presses contribute to the diversity in contents and languages through a considerable production of titles in smaller fields such as Archaeology and Prehistory, Linguistics, Philology, Literature and Sociology which, in relative terms, are less interesting for the private/commercial sector. Therefore, taking into account the differences between catalogues, the interest in preserving bibliodiversity becomes clear, as well as the need to recognize it in the processes of scholarly assessment. The leadership in the market of large commercial imprints should not diminish the function or recognition of the professional scholarly publishers which in each country take care of the publication of rigorous scholarly contents, closer to local topics and in national languages. Global and local research have different spaces when it comes to publishing. What is published in international publishers is not the same as what is published by national publishers, and does not reach the same readers. Furthermore, the publication of results in local, regional or national publishers can generate an impact in society which is impossible to reach for international publishers. Some of these ideas are pointed out by Sivertsen (2018) and also by Tanner (2016), when he discusses the distribution of publishers obtained in his analysis of the outputs of REF 2014: "Appendix B provides a full list of all the 1.180 uniquely named book publishers listed in the REF2014 Panel D submissions. The long list of publishers suggests that many of the works published may be relatively obscure in terms of scholarly retailing. This is borne out by closer inspection of works from publishers with only one book submitted in a subject. These books, whilst relating important research outputs, appear to have been published with smaller publishing houses for academic reasons spanning: regional, political, linguistic, research partner, economic, digital, OA and ideological considerations." (Tanner, 2016) Being conscious of this issue, the project Academic Book Publishers: a global and interactive register emerged in the framework of the COST action ENRESSH (European Network for Research Evaluation in the Social Sciences and the Humanities), aiming to show the diversity of scholarly books that follow quality standards applicable in scientific research, and to claim the function that they fulfill in the generation of plural contents (Giménez Toledo et al., 2019). The objective is to favor the recognition of books in the scholarly output (Hicks et al. 2010) without biases derived from the type of publisher, its country of origin, or the language of publication. All these publishers are nourished by the research results produced by humanists and social scientists in the various countries. Existing data show that there are contents for all of them, although not all with the same quality or relevance, of course. The analysis of catalogues also shows the differences in titles, foci and languages between each of them. The heterogeneity can be observed also in the publishing practices, the funding, dissemination and commercialization models, their professionalization and the adoption of digital technologies.
It is important to recognize their contribution to the preservation of the diverse and plural research in the humanities and social sciences, as well as to the transference of results to society. Both questions are linked to the existence of publishing structures in each country. All this is part of the claims of the Helsinki initiative on multilingualism in scientific communication (Helsinki initiative, 2019). Also the Leiden manifesto mentions the need to recognize the relevant local research which is published in journals (and books) in languages different from English: "Metrics built on high-quality non-English literature would serve to identify and reward excellence in locally relevant research" (Hicks et al.). In this sense the works of Dahler Larsen (2018) and Sivertsen (2018) claim the various itineraries of publication for the researchers.
The research topics and their potential audiences are conditioning factors when choosing a publication channel: books, journals and others, national or international. There are also factors related to the evaluation of the researcher, since not all publications have the same weight and, in many cases, priority is still given to indexed, international journals. Finally, the decision where to publish responds also to the researcher's ethical principles and his or her engagement with society: the author must take care of making his or her research results available to those who need them. If necessary for the attainment of the aforementioned goal, the researcher would have to publish with a national publisher or even in a technical report. In such a case it should be guaranteed that such decision will not be penalized. After all, the researcher would be prioritizing the societal impact of the research over scholarly impact. Both forms of impact are compatible, although it is not always an easy task. For these reasons, science policies ought to be congruent: if they intend to make research reach society it is important to give an adequate value to the channels that allow that to happen, scholarly books and reports among them. 2

Scholarly books and they recognition in the research evaluation processes
Documents such as the Leiden Manifesto (Hicks et al., 2015), San Francisco Declaration on Research Assessment (2012) or ENRESSH Policy Brief Research Evaluation (Ochsner et al. 2020) emphasize the need of a proper consideration of the differences between scientific fields, the forms of scientific output as well as the various sources of information and indicators. This avoids biases which could derive from the coverage of the various sources in terms of document coverage, topics or languages.
Research evaluation based on expert judgements, complete publication data sources (covering the full set of the scholarly output) and using multidimensional methods are three fundamental and demandable keys in the evaluation of the scholarly activity nowadays. All three lead to a responsible evaluation with regards to those under evaluation (researchers and institutions) which live up to the dedicated, intensive and long term work which is the writing of a book, as well as to the publishing structures of each country, which support and enable the publication of research results. This implies, in the specific case of books, a greater connection between the associations of scholarly publishers in a given country and their scientific evaluation agencies, and a better knowledge of their respective action frameworks.
Carrying out research on the publishing sector at the national and international level is an essential step in the adequate assessment of the publishers. It allows the experts to include more precise information for decision making as well as to raise awareness on the work of the researchers and the publishing industry in keeping those publishing structures. Without knowing the national publishing sector, its working practices, specialization and catalogue features, it is not possible to carry out a fair evaluation nor to establish appropriate metrics.
Isolated metrics, without knowledge of the application context, can produce substantial errors in the evaluation. The amount of data provided by the digital environment for books -from reviews to sales, including also citations and downloads -are a temptation for data analysts and a huge risk if those data are not interpreted within the context of the characteristics of the publishing sector. For example, it is necessary to know that many small and medium-sized scholarly publishers have not begun a solid digital transition, which implies that their web visibility is limited, their books show little discoverability, are difficult to buy, download or read online. In this sense, little can be known about the circulation of their books or their impact. The metrics that can be obtained are insufficient. The reasons for not having developed that transition can be related to the lack of coordinated policies, lack of support for small publishers, and very limited capacities for investment in innovation. Without doubt, these are pending challenges. Their possibilities of competing, in every sense, with the large publishers are trivial. Nevertheless, they publish research books which are socially relevant (post-conflict in Colombia, femicide in Latin America or the resignification of places and historic memory, just to mention some examples). Therefore, there is a lack of actions which facilitate the visibility and accessibility of the bibliodiversity on internet. In order to counteract this weakness it would also be important to consider the world of scholarly publishing as a whole and not only the visible part of it (which is so for various types of reasons).
As pointed out by Engels et al. (2018) "Research evaluation systems need to take book publishing fully into account because of the crucial epistemic and social roles it serves in an SSH". In recent years the research on scholarly books and the methodologies for its evaluation has notably increased (Zuccala et al. 2018;Zuccala & Robinson, 2019;Bonaccorsi, ed., 2018;Kousha et al. 2011.;Kulczycki, E., Korytkowski P., 2020). In an analysis of the institutional evaluation practices (and in some cases at the individual level) in 19 countries (Giménez Toledo et al. 2019), different evaluation models were detected: from strictly qualitative ones based on expert panels, to those using validated lists and/or publishers rankings and quality labels for books, series or publishers. According to this study, there are three common features shared by many of the models: • The evaluation of books based on the evaluation of the book publisher. The individualized evaluation of books does not seem to be feasible -at least in evaluation processes at the institutional level-because of the resources required and the amount of scientific output which has to be evaluated. Basing the evaluation on the publisher provides a specific value to the work of the publisher concerning the selection filters applied in what Calasso (2015) calls 'the Art of the publisher', that allows him/her to develop a differentiated catalogue. • The evaluation of the publishers by expert panels or massive consultations to researchers, which implies the recognition of the publishers by their readers and authors. • The manuscript selection processes, not always based on traditional peer review used in scientific journals (Sivertsen et al., 2017) but on various methods implying validation and filtering of scholarly contents. The way in which a catalogue is constructed is one of the most frequently valued aspects, as well as one of the least known by the evaluation agencies. This is why one of the main working axes of the project Academic Book Publishers is the analysis of the manuscript selection processes used in books scholarly publishing, with the objective of increasing knowledge of the working processes within publishing houses and making these processes transparent to authors and evaluators.
What has been learnt from the evaluation of scholarly journals is not completely applicable to the evaluation of books, series or scholarly publishers. The scholarly book sector and its publishing practices ought to be a specific object of study. By doing so not only would it be possible to understand the nature of what is published, but also the processes along which it is published. Both variables are essential in order to face the evaluation of this form of scholarly output with rigor and precision. By doing so, furthermore, a part of the cultural industry and, not less important, a plural research will be protected.