The issue of multilingualism vs lingua franca in science (and in society) is certainly very complicated, but the recent article by Gregory Crane raises some questions and a few concerns. In general, I think everybody would agree with Miran’s appeal on Humanist: “Let us invest in language diversity”.
There are countless documents supporting multilingualism in science, but it is interesting to look at the conclusions of the 2011 European Science Foundation Young Researchers Forum on Changing publication cultures in the Humanities:
While also aiming for multilingualism, we should preserve linguistic diversity in humanities research. Diversity has to do with creativity, which is a precondition for the humanities. A key instrument of diversity is translation.
I think we will never stress enough the value of translation. I’m currently reading the fascinating story of the Jesuit Orientalist Yves Raguin, for whom philology, spirituality and cultural awareness coincided. It is the responsibility of dominant languages and cultures to translate from marginal or less influential languages. In our forthcoming book The Digital Humanist. A Critical Inquiry (Punctum Books) there is a quote by Luisa Maffi that could help us to understand the vital issues at stake:
“[…] a lowering of both cultural and biological diversity has been found to correlate with the development of complex, stratified and densely populated societies and of far-reaching economic powers. … From ancient empires to today’s globalized economy, these complex social systems have spread and expanded well beyond the confines of local ecosystems, exploiting and draining natural resources on a large scale and imposing cultural assimilation and the homogenization of cultural diversity. (Maffi 2010, 8).
In other words: cultural and biological richness does not necessarily match material wealth.
The same should apply to scholarly endeavor of any kind and at all latitudes: there is no necessary relationship between international visibility, language of publication and research quality. English native speakers get a free ride, but the incommensurate economical, rhetorical and semiotic power of Anglophones undermine and inhibit the right to express ideas in our own native language. If biology is a model, then we should remember that monoculture is pushing species towards extinction in the most effective way.
A colleague and I have an article coming out on the relationship between the language of DH publications and the languages of sources (i.e. bibliographic references and citations). Our data, although gathered from a relatively small sample (seven main DH journals worldwide), show that DH is monolingual regardless of the country and/or working institution/affiliation of authors. The table down here summarizes the main data we’ve collected on the use of sources in DH journals (more details will be available soon on this blog).
Number and percentage of sources in most used languages*
|Sources||% on Journal total||Total of Journal sources|
|Digital Humanities Quarterly||4630||97%||17||0.4%||77||2%||13||0.3%||15||0.3%||4766|
|Jahrbuch für Computerphilologie||326||39%||11||1%||1||497||59%||843|
|Digital Studies / Le champ numérique||2213||97%||66||3%||1||2281|
|Digital Medievalist **||357||83%||23||5%||7||2%||13||3%||4||1%||430|
|Literary & Linguistic Compunting||5706||94%||76||1%||63||1%||111||2%||28||0.5%||6092|
These data show that the real problem is not that English is the dominant language of academic publications (and of DH), but that both Anglophone and a high percentage of non-Anglophone colleagues barely use/quote non-Anglophone sources in their research. On the long run, this trend could have a devastating effect on Humanities research as a whole, and lead to the disappearance of cultural diversity (at least in academic publications…). In educational institutions worldwide we keep hearing “go English if you want to be international”, a mantra that can be also translated as “your local language is useless for intellectual expression”.
I have also some concerns about the reliability of the sources used by Crane to draw some of its conclusions. As for Scopus (but the same could be said for other private centers of academic legitimation), I will summarize here information that everybody can easily found on the web. For a more accurate discussion on the academic performance indicators see this study by the Italian economist Michele Gazzola.
First of all, we should not forget that Science Citation Index, the Social Sciences Citation Index, and the Arts & Humanities Citation Index are heavily biased in favor of journals in English. That was one of the main reason behind the creation of the European Reference Index for the Humanities project. Although many disagreed with its ranking criteria, ERIH emerged from an urgent need:
The background for launching this project in 2001 was the need felt by European researchers for better databases than the existing ones, which are limited in coverage (even some of the best English language journals are missing from the database and there are even bigger gaps in the coverage of journals in other European languages) and tend to be centred on Anglo-American publications. Research conducted in national (especially in so- called lesser used) languages is either not adequately covered or not covered at all. Furthermore, there are specificities in the Humanities in terms of cultures of publication and traditions of citations, which make it meaningless to work with the evaluation tools used in science.(Ferenc Kiefer, “ERIH’s Role in the Evaluation of Research Achievements in the Humanities” in P. Dávidházi (ed.), New Publication Cultures in the Humanities. Exploring the Paradigm Shift, Amsterdam: Amsterdam University Press, 2014, pp. 173-182.)
There is of course more to say about Scopus-Elsevier. You will remember the international boycott campaign against Elsevier nasty business practices. Some recent reflections on the conflict of interest with (public?) academic research can be found here.
Not long ago on Researchgate.net there was a discussion about PROs and CONs in adopting SCOPUS as a source for bibliometric measures of scientific publications. Most of the answers were critical of SCOPUS. I summarize them:
1) Since SCOPUS is run by a publishing company, it cannot reasonably being assumed as independent of commercial interests.
2) Given the commercial nature of Scopus there is no way to act on it to have conferences or journals included (i.e. less-than-full transparency of inclusion decisions).
3) When the DB became the official source for bibliometrics to evaluate [national] academics, authors found out that only few of the citations to their works were correctly indexed. Scopus was hence flooded with correction requests.
4) Many citations in Scopus are negative citations (to incorrect results) but are counted as favouring the original author! It’s similar to Facebook only having “like” buttons: How do you express sympathy to a person who’s sister has just died? Press “like”???
“There are so many CONS compared to PROS that the answer to the original
question is a resounding “NO!”
But apart from these general remarks, one figure attracted my attention in Crane’s paper (see his Appendix), and it was the percentage of English articles published in Italian journals indexed by Scopus: 66% (see Figure 1 on his paper). This figure surprised me because I thought that most of Italian Humanities journals published articles written in Italian. So I did some research, and found a detailed work made in 2010 by two Italian scholars (published in Italian… sorry).
Briefly, although the number of Italian journals has increased since 2011 (151 titles in 2013), still Scopus (and WoS) indexes a small portion of Italian Humanities journals that publishes mainly in Italian. This is quite understandable, as is the interest of more “international” journals to appear in Scopus or other databases, while a journal with higher local impact (but published in the local language) has much less interest (and resources) to be indexed. This could explain the majority (66%) of English articles published in Italian Scopus-indexed journals. However, it is easy to demonstrate that this figure is a misrepresentation of the real dimension and relevance of Humanities research in Italy. First of all, the number of Italian Humanities journals far exceeds 151 titles (and an increasing percentage includes open access and online refereed journals). Many sectors (for example History of Art) are underrepresented, and it is not credible for an “international” database that UK and USA have ten times Humanities journals included.
Secondly, and perhaps most importantly, relevant research in the humanities, especially in Romance languages contexts, has been traditionally based on the monograph rather than the journal article. Knowledge in the Humanities tends to be accumulative rather than experimental, as it showed for example by the life-long endeavor of preparing a critical edition. But the advent of national evaluation agencies (in Italy the ANVUR, ANECA and ANEP in Spain, etc.) changed this scenario and imposed on humanists the journal article-based model of the hard sciences. So the truth today is that if I publish an English article in an Anglo-American journal I will get a “higher score” than if I make a critical edition of Petrarch with a small local “quality” publisher. It’s the evaluation, stupid!
In conclusion, while numbers allow us to affirm that “English is clearly the dominant language of publication”, any connection between the representation offered by these private databases and the reality – let alone the “quality” – of research in a given cultural space / region should be abandoned.
Italian or Italy in fact are the least problem: as Scopus itself admits, “looking at the regional diversity it is clear that Europe as a whole is best represented and Central & South America and Asia Pacific are underrepresented with respect to Humanities content.”
If this picture is confirmed, using Scopus as source of legitimization of either visibility and quality of research in Humanities and Social Science could only lead to the reinforcement of a vicious circle. The over-representation of US and UK Humanities titles (1178 and 1113 respectively) in Scopus, as well in Web of Science, will always support arguments in favor of using English as the lingua franca, and the misrepresentation of knowledge production and geopolitical inbalance will continue to thrive.
* In few cases some languages are missing (i.e. Dutch in DM or Japanese and Swedish in DHQ) so the sum does not get to 100%. Besides we did not include figures below 0.1%.
** In DM we found 22 sources in Dutch (22% of the total) and 4 in Latin (1%).