petersuberJul 30, 2023

In September 2020 I started what became a long #Twitter thread on #MultilingualResearch.
https://twitter.com/petersuber/status/1307774697531113474

Starting today, I'm stopping it on Twitter and continuing it on #Mastodon.

Here's a rollup of the complete Twitter thread.
https://resee.it/tweet/1307774697531113474

Here's a nearly complete archived version in the @waybackmachine.
https://web.archive.org/web/20220908060944/https://twitter.com/petersuber/status/1307774697531113474

Watch this space for updates.

#Academia #Multilingualism
@academicchatter

šŸ§µ

Peter Suber (@[email protected]) on X

New study: "More than 90% of the scientific articles published by Colombian researchers are in English....Publishing in a 2d language creates additional financial costs...&...problems with reading comprehension, writing ease & time, & anxiety." https://t.co/ZLDvIdNZSq

X (formerly Twitter)
Show threadpetersuberAug 4, 2023

Update. The cost of translating #consent forms into other languages limits the participation of non-English speakers in clinical trials.
https://www.nature.com/articles/s41586-023-06382-0

Summary this study.
https://www.science.org/content/article/non-english-speakers-are-being-shut-out-clinical-trials

#Multilingualism #MultilingualResearch

Consent document translation expense hinders inclusive clinical trial enrolment - Nature

The availability of translated consent documents improves recruitment of patients with limited English proficiency to clinical trials, indicating a potentially modifiable barrier to the inclusion of patients with limited English proficiency.

Nature
Show threadpetersuberAug 5, 2023

Update. The Journal of Electronic Publishing (@JEPub) just issued a call for papers to appear in a special issue on #Multilingual publishing.
https://journals.publishing.umich.edu/jep/news/71/
(#paywalled)

The special issue will accept papers in accept papers in English, Spanish, or French.

#Multilingualism #MultilingualResearch #CFP

Show threadpetersuberAug 27, 2023

Update. New study: "Most journals [in #biology] offer minimal support for scientists whose first language is not Englishā€¦Only 8% of the journals made their complete guidelines to authors available in at least one language other than English; less than 7% allowed authors to publish articles in languages other than English; and a mere 10% explicitly approved the use of references published in a language other than English."
https://www.nature.com/articles/d41586-023-02529-1

#Multilingualism #MultilingualResearch

Scientists who donā€™t speak fluent English get little help from journals, study finds

An analysis of hundreds of journals finds limited efforts to accommodate scientists who are not native English speakers.

Show threadpetersuberSep 3, 2023

Update. "Despite English being the lingua franca of science, getting an excellent level of English is a privilege of few, restricted to high-income countries or those who can pay for intense language training in Global South countries. Yet, most of the hiring calls from Global North research institutes urging inclusion conflate being a good scientist with being good (or excellent) at English. This isnā€™t true."
https://www.scientificamerican.com/article/english-may-be-sciences-native-language-but-its-not-native-to-all-scientists/

#Multilingualism #MultilingualResearch

English May Be Scienceā€™s Native Language, but Itā€™s Not Native to All Scientists

There are talented scientists worldwide who do not speak fluent English. We have to accommodate the language barrier or risk losing their potential

Scientific American
Show threadpetersuberSep 3, 2023

Update. Good line from the previous piece (this thread, above).
https://www.scientificamerican.com/article/english-may-be-sciences-native-language-but-its-not-native-to-all-scientists/

"When it comes to language, we often think of Gloria Pritchett, a character in the popular American television show Modern Family. Gloria is a native Spanish speaker living in a fast-paced English-speaking family. 'Do you even know how smart I am in Spanish?' she says in one episode.ā€

#Multilingualism #MultilingualResearch

English May Be Scienceā€™s Native Language, but Itā€™s Not Native to All Scientists

There are talented scientists worldwide who do not speak fluent English. We have to accommodate the language barrier or risk losing their potential

Scientific American
Show threadpetersuberSep 11, 2023

Update. Can #AI solve the problem of #language #discrimination (good research by non-native speakers rejected from English-language journals on the ground of weak English)?
https://www.timeshighereducation.com/blog/ai-writing-tools-will-not-fix-academias-language-discrimination-problem
(#paywalled)

Natalia Kucirkova argues no. The best bots may be good enough. But they are not accessible to all who need them, and "non-native researchers still have to invest extra time & more dollars to fix their papers than their native colleagues."

#Multilingualism #MultilingualResearch

AI writing tools will not fix academiaā€™s language discrimination problem

Affordable AI-powered writing software offers some hope to scholars unfairly criticised for their imperfect English, but more radical change is required, says Natalia Kucirkova

Times Higher Education (THE)
Show threadpetersuberSep 18, 2023

Update. All too often scholars researching global attitudes toward something will write a survey in English and only English. Because I've criticized this practice, I'm happy to spotlight and applaud this research team for writing its survey (on data sharing and re-use in the field of traumatic stress) in seven languages ā€” and publishing its abstract in three languages.
https://www.tandfonline.com/doi/full/10.1080/20008066.2023.2254118

#Multilingualism #MultilingualResearch

Show threadpetersuberOct 9, 2023
Update. New study: Chinese scholars decide whether to publish in Chinese or English based on the "perceived value" of the "best-fitting" journals.
https://www.frontiersin.org/articles/10.3389/fpsyg.2023.1249857/full
Publishing in English or Chinese: a qualitative analysis of Chinese researchersā€™ academic language choice

Non-native language scholars often struggle to choose between English and their native language in scholarly publishing. This study aims to identify the mechanism by which journal attributes influence language choice by investigating the perspectives of 18 Chinese scholars through semi-structured interviews. Drawing on grounded theory, this study develops a model for how journal attributes influence researchersā€™ language preferences. We find that journal attributes influence researchersā€™ perceived value which, in turn, affects their particular language choice, with contextual factors playing a moderating role. By examining the motivations underlying Chinese scholarsā€™ language choice, this study provides a critical understanding of the factors shaping their decision-making processes. These findings have significant implications for Chinese scholars, policymakers, and journal operators, shedding light on the issue of discrimination in academic publishing. Addressing these concerns is crucial for fostering a fair and inclusive academic environment.

Frontiers
Show threadpetersuberOct 29, 2023

Update. I missed this from 2020: "Among the researchers [from seven European countries] who published at least three [#SSH] journal articles [in 2013-15] over one-thirdā€¦had written their work in at least two languagesā€¦Research is international, but multilingual publishing keeps locally relevant research alive with the added potential for creating impact."
https://asistdl.onlinelibrary.wiley.com/doi/10.1002/asi.24336

#Multilingualism #MultilingualResearch

Show threadpetersuberOct 31, 2023

Update. The Confederation of #OpenAccess #Repositories (#COAR) just released an important set of recommendations on managing multilingual and non-English language content in #OA repositories.
https://www.coar-repositories.org/news-updates/managing-multilingual-and-non-english-language-content-in-repositories/

#GreenOA #Metadata #Multilingualism #MultilingualResearch

Show threadpetersuberNov 1, 2023

Update. I missed this from last year: "We broadly review the advantages and limitations ofā€¦machine #translationā€¦and propose that translation can serve as both a short- and a long-term solution for making science moreā€¦accessible, globally representative, and impactful beyond the academy. We outline actions that individuals and institutions can take to support multilingual science."
https://academic.oup.com/bioscience/article/72/10/988/6653151

#Multilingualism #MultilingualResearch

Overcoming Language Barriers in Academia: Machine Translation Tools and a Vision for a Multilingual Future

abstract. Having a central scientific language remains crucial for advancing and globally sharing science. Nevertheless, maintaining one dominant language

OUP Academic
Show threadpetersuberNov 1, 2023

Update. See some new comments on the previous article (this thread).
https://www.universityaffairs.ca/news/news-article/machine-translation-a-game-changer-in-science/

"The real benefit of machine translation lies elsewhere, said [Vincent] LariviĆØre. 'In an ideal world, the #metadata, #abstracts and titles of scientific articles would be available in [many] languages, increasing their #discoverabilityā€¦Readers could then choose to translate the text into any language they want. We wouldnā€™t always have to use English.' "

#Multilingualism #MultilingualResearch

Show threadpetersuberNov 10, 2023

Update. "Non-native English speakers face additional barriers to scientific publishing, from journal guidelines accessible only in English to higher rates of language-related manuscript rejection. Journals, congresses, and others are beginning to offer support ranging from English language mentoring programmes, training, and buddy systems to free AI proofreading tools."
https://thepublicationplan.com/2023/11/07/language-barriers-in-scientific-publishing-how-many-hurdles-are-there/

#Multilingualism #MultilingualResearch

Language barriers in scientific publishing: how many hurdles are there?

Find out more about the barriers non-native English speakers face in scientific publishing, and initiatives that are starting to help.

The Publication Plan for everyone interested in medical writing, the development of medical publications, and publication planning
Show threadpetersuberDec 7, 2023
Update. "Much scientific evidence is still published in non-English languages, and excluding non-English-language evidence can introduce biases in evidence synthesis."
https://ecoevorxiv.org/repository/view/6342/
Assessing diverse values of nature requires multilingual evidence

Show threadpetersuberDec 18, 2023
Update. 2nd Helsinki Initiative Webinar on #Multilingualism in Scholarly Communication
https://www.helsinki-initiative.org/en/events/2nd-helsinki-initiative-webinar-multilingualism-scholarly-communication
2nd Helsinki Initiative Webinar on Multilingualism in Scholarly Communication | helsinki-initiative.org

Show threadpetersuberJan 24, 2024
Update. "To what extent are minority languages excluded from scholarly publishing infrastructure and initiativesā€¦This is the provocative question we ask in this opinion piece, drawing on our experience of publishing the worldā€™s only academic journal in our minority language [SĆ”mi]."
https://insights.uksg.org/articles/10.1629/uksg.640
Show threadpetersuberJan 27, 2024

Update. "GPT detectors frequently misclassify non-native English writing as #AI generated, raising concerns about fairness and robustnessā€¦GPT detectors could spuriously flag non-native authorsā€™ content as AI #plagiarism, paving the way for undue harassment."
https://www.sciencedirect.com/science/article/pii/S2666389923001307

#LLMs #Multilingualism #MultilingualResearch

Show threadpetersuberFeb 29, 2024

Update. "The need to publish in English even when it is not the local language affects the type of research undertaken & further consolidates the global North-centric view of scientific approach. The bibliometric databases on which assessments of universities and journals are based are owned by two large [corporations], and this concentration of the market has in turn concentrated the research environment. #OpenInfrastructure offers an alternative option."
https://osf.io/preprints/socarxiv/c8yq3

#Monopoly

OSF

Show threadpetersuber
Update. "Most journals [make] minimal efforts to overcome language barriers. The impact factor of journals [#JIF] was negatively associated w/ adopting a number of inclusive policiesā€¦Ownership by a scientific society tended to have a positive association. Contrary to our expectations, the proportion of both #OpenAccess articles & editors based in non-English speaking countries did not have a major positive association w/ the adoption of linguistically inclusive policies."
https://doi.org/10.1098/rspb.2023.2840
Mar 17, 2024 at 7:22pmWeb
Show threadpetersuberMar 24, 2024

Update. The authors of the previous piece in this thread just released a summary.
https://theconversation.com/prestigious-journals-make-it-hard-for-scientists-who-dont-speak-english-to-get-published-and-we-all-lose-out-226225

"Linguistically inclusive policies come in many forms, and can be implemented at each stage of the editorial process. They might aim to make publishing more #multilingual. Alternatively ā€“ if sticking with English ā€“ they may aim to reduce the burden on non-native English speakers."

#Multilingualism #MultilingualResearch

Prestigious journals make it hard for scientists who donā€™t speak English to get published. And we all lose out

A study of 736 biological science journals showed only a small fraction are making efforts to foster a multilingual scientific community.

The Conversation
Show threadpetersuberMar 28, 2024

Update. New study: "Academics who perceive high #pressure to publish tend to employ instrumental publication strategies rather than normative onesā€¦Publishing results in open-access outlets or in native languages other than English is less important for those under pressure."
https://academic.oup.com/rev/advance-article/doi/10.1093/reseval/rvae011/7634754

#Multilingualism #MultilingualResearch #OpenAccess

The impact of researchersā€™ perceived pressure on their publication strategies

Abstract. This article investigates researchersā€™ publication strategies and how their perceived pressure to publish and to obtain external funding are rela

OUP Academic
Show threadpetersuberApr 3, 2024

Update. "Too often, scientific research in any language other than English is automatically seen as second tier, with little consideration for the quality of the work itself. This harmful prejudice ignores the work of those involved, especially in the humanities and social sciences. It also profoundly undermines the global academic communityā€™s ability to share knowledge with society."
https://theconversation.com/english-dominates-scientific-research-heres-how-we-can-fix-it-and-why-it-matters-226198

#Multilingualism #MultilingualResearch

English dominates scientific research ā€“ hereā€™s how we can fix it, and why it matters

Around 98% of all research is published in English, posing a serious problem for the global scientific community.

The Conversation
Show threadpetersuberApr 25, 2024

Update. Results of two surveys of authors whose first language is not English, mostly from China, on the use of #AI tools to publish in English-language journals. "The top reported use cases for AI tools were grammar corrections/proofreading, rewriting, translation, and drafting abstracts/lay summaries."
https://www.csescienceeditor.org/article/ai-what-the-future-holds/

#Multilingualism #MultilingualResearch

Artificial Intelligence: What the Future Holds for Multilingual Authors and Editing Professionals - Science Editor

Observations and predictions based on two rounds of surveys among primarily Chinese-speaking researchers conducted in early and late 2023. A Tumultuous 2 Years The increasing availability of generative artificial intelligence (AI)-based tools such as ChatGPT for writing and editing, among other academic tasks, has prompted considerable debate among researchers, universities, publishers, journals, and other stakeholders over the boundaries separating the ethical and unethical use of such tools. Some, such as the Science family of journals and the University of Hong Kong, initially imposed strict restrictions on the use of ChatGPT and other AI tools, both deeming the use of AI-generated text as plagiarism.1,2 These restrictions have since been relaxed,3-5 reflecting a general movement in academia, both in education and research, from outright bans to embracing productive and ethical use.  Authors for whom English is not a first language (hereafter, multilingual authors) face greater barriers to publication than their native English-speaking counterparts.6,7 Multilingual authors have traditionally relied on editing or translation services to ensure their manuscripts meet journalsā€™ requirements for high standards of English. However, not all can afford professional language services, and their use increases the costs associated with publishing. Generative AI tools can be a game-changer for multilingual researchers, bringing much-needed equity to academic publishing and eliminating English fluency as a barrier to research dissemination.8  As managers at a company providing (100% human) academic editing services, we were interested in how the authors we work with, the majority of whom are multilingual, perceived the recent developments in AI and [ā€¦]

Science Editor
Show threadpetersuberApr 25, 2024

Update. Wonderful project. "Climate Cardinals is a youth-led nonprofit thatā€™s accomplished quite a bit with almost no funding: translating 2 million words in four years to make scientific literature more accessible to non-English speakers." #Google just gave the project $400k.
https://www.forbes.com/sites/jeffkart/2024/04/22/climate-cardinals-plans-to-double-its-science-translation-efforts-with-new-google-backing

#Climate #Translation

Climate Cardinals Plans To Double Its Science Translation Efforts With New Google Backing

The youth-led nonprofit has accomplished quite a bit with almost no funding: translating 2 million words in four years to make scientific literature more accessible to non-English speakers.

Forbes
Show threadpetersuberApr 28, 2024

Update. This new study follows up on three recommendations from #COAR on #multilingualism in #repositories ā€” "[1] declaring the language(s) of the resource andā€¦its #metadata, [2] writing personal name/s using the writing system used in the deposited document while providing a persistent identifier to disambiguate author/sā€¦andā€¦[3] enabling UTF-8 support so as to promote use of the original alphabet / the writing system whenever possible."
https://zenodo.org/records/11060284

#GreenOA #PIDs

Dealing with multilingualism and non-English content in open repositories: Challenges and perspectives

Abstract Several organizations and initiatives have recently called for more support of multilingualism in research to promote epistemic plurality and raise awareness of the adverse effects of an anglocentric research ecosystem. But this support for and practice of multilingualism and linguistic diversity cannot happen in a digital or technological vacuum. Open repositories can play an important role in ensuring that research infrastructures have the ability to implement and promote multilingualism at scale in an Open Science environment. This implementation, however, is complex and does not come without its own theoretical and technical challenges. One of these challenges is to recognize that the implementation of multilingualism in open repositories can hardly be dissociated from wider concerns of discoverability, research assessment practices, and the anglocentric nature of digital infrastructures and metadata standards or protocols. Drawing on the COAR (Coalition of Open Access Repositories) recommendations report produced by the COAR Task Force on Supporting Multilingualism and non-English Content in Repositories, this article presents and critically examines how and why three particular recommendations of this document are particularly well suited to support a decolonial trajectory for the management of multilingualism in open repositories. More specifically, this article discusses the decolonial aspects and praxis underlying guidelines such as declaring the language(s) of the resource and of its metadata, writing personal name/s using the writing system used in the deposited document while providing a persistent identifier to disambiguate author/s identification and, overlapping with the latter, enabling UTF-8 support so as to promote use of the original alphabet / the writing system whenever possible, without negating the possibility to transliterate metadata by means of recognized standards (e.g. ISO). In so doing, we argue that these recommendations enable a multifaceted technology and politics of recovery that promotes a form of linguistic revitalization and strengthens linguistic diversity.  

Zenodo
Show threadpetersuberMay 2, 2024
Update. Next week the Public Knowledge Project and the U of Alberta Library are hosting a conference on #OpenAccess #multilingual publishing.
https://ualberta.libcal.com/event/3797697
Multilingual Publishing: Editorial Lunch & Learn

Join us for the next Open Publishing Editorial Lunch and Learn on Multilingual Publishing. This session will be presented by Emma Uhl from the Public Knowledge Project (PKP) and...

UAL Calendars/Bookings
Show threadpetersuberMay 9, 2024
Update. "This paper presents the challenges language ideological research still faces despite ongoing efforts to tackle #multilingualism and offers possible solutions for language ideological research to become truly inclusive."
https://www.degruyter.com/document/doi/10.1515/eujal-2024-0006/html
Whose language counts?

This position paper makes a critical intervention in one of the mostly salient sociolinguistic debates of the recent years, that of the integration of multilingual speakers in language ideological research. Although many scholars now recognize the need for decentering the ideal(ized) ā€˜monolingual native speakersā€™, they remain the default norm in language ideological research. Indeed, despite our efforts to dismantle our own language ideologies as linguists, the implicit focus on native speakers and monolinguals proceeds to the erasure or invisibilization (Irvine & Gal 2000) of multilinguals. Drawing on research on gender-inclusive language as well as my own empirical study on L2 speakers of German, I show why integrating a higher variety of language users is necessaryā€”not only on ethical grounds, but also on theoretical ones. Altogether, this paper presents the challenges language ideological research still faces despite ongoing efforts to tackle multilingualism and offers possible solutions for language ideological research to become truly inclusive.

De Gruyter
Show threadpetersuberMay 11, 2024

Update. "Chinese-language journal editors expressed a lack of motivation to adopt #OpenAccess, unless there is a reform in the mechanism of academic publishing or a policy is imposed. On the other hand, the English-language journal editors [in #China] acknowledged that they have no other choice but to adopt open access."
https://www.mdpi.com/2304-6775/12/1/1

h/t @hannaSH

Going Open Access: The Attitudes and Actions of Scientific Journal Editors in China

This study aims to investigate the attitudes and actions of scientific journal editors in China towards open access. Semi-structured interviews were conducted with 17 Chinese editors from various scientific journals during September and October of 2022. The results indicate that the editors generally possess knowledge of open access and have implemented an appropriate open access model for their respective journals. However, the Chinese-language journal editors expressed a lack of motivation to adopt open access, unless there is a reform in the mechanism of academic publishing or a policy is imposed. On the other hand, the English-language journal editors acknowledged that they have no other choice but to adopt open access. This study helps us learn about Chinese editorsā€™ understanding and attitudes towards open access, the current status of open access in Chinaā€™s scientific journals, and the mechanisms of academic publishing in China.

MDPI
Show threadpetersuberMay 13, 2024

Update. "The #Translations and #OpenScience project explores the potential of technology-aided translation to help produce and disseminate research in multiple languages."
https://zenodo.org/records/10972986

#Multilingualism #MultilingualResearch

Exploratory studies for the creation of a technology-aided collaborative translation service in open scholarly communication

Since the Helsinki Initiative in 2019, language diversity and multilingualism have become key concerns in scholarly communication. Among the initiatives working towards a sustainable multilingual science, the Translations and Open Science project explores the potential of technology-aided translation to help produce and disseminate research in multiple languages. This report presents an overview of the four exploratory studies conducted as part of the Translations and Open Science project in order to lay the foundations of a technology-aided collaborative translation service for open scholarly communication. More detailed information can be found in the specific deliverables of each study, cited as references in the present report.

Zenodo
Show threadpetersuberMay 13, 2024

Update. "This study explores Chinese early career returnee academicsā€™ motivations for writing and publishing in English."
https://www.nature.com/articles/s41599-024-02846-4

#China #ECRs #Multilingualism #MultilingualResearch

Why English? Exploring Chinese early career returnee academicsā€™ motivations for writing and publishing in English - Humanities and Social Sciences Communications

To improve their research performance in international league tables, many universities in non-English language dominant settings recruit academic returnees in the hope that they will increase the quantity and quality of articles published in English-medium internationally indexed journals. This study explores Chinese early career returnee academicsā€™ motivations for writing and publishing in English. Utilising ecological systems theory, the findings show that the microsystem is reflected in the early career returnee academicsā€™ interaction with collaborators, while national policies constitute the exosystem. The academic culture has a noteworthy impact at the macrosystem level. This study contributes to the understanding of early career returnee academicsā€™ motivations to write and publish in English which will assist policymakers and university administrators to create a more beneficial environment to promote the accomplishments of academic returnees.

Nature
Show threadpetersuberJun 18, 2024

Update. "We first introduce #translingualism and #ELF [English as a lingua franca] as equally legitimate multilingual theories. Next, we identify [their similarities and differences]ā€¦Finally, we lay out a vision for how this transdisciplinary, collaborative dialogue can help realize more equitable #multilingualism."
https://www.degruyter.com/document/doi/10.1515/eduling-2023-0012/html

#MultilingualResearch

Towards equitable multilingualism: promoting transdisciplinary, collaborative dialogue between English as a lingua franca and translingualism

Inspired by Hultā€™s (2010. Theme-based research in the transdisciplinary field of educational linguistics. In Francis M. Hult (ed.), Directions and prospects for educational linguistics , 19ā€“32. Springer) discussion of theme-based research within educational linguistics that focuses, first and foremost, on solving problems rather than adhering to disciplinary boundaries and conventions, this article calls for transdisciplinary collaboration to resolve educational problems related to monolingualism, native-speakerism, and the hegemony of colonial ideologies. More specifically, we make a case for transdisciplinary dialogue between English as a lingua franca (ELF) and translingualism. We argue that these two research fields, with their increasingly converging interests and goals as well as distinct strengths, can make a greater impact in addressing persistent educational problems through targeted collaboration, research, and (language) pedagogy based on equitable multilingualism (e.g., Ortega, Lourdes. 2019. SLA and the study of equitable multilingualism. The Modern Language Journal 103. 23ā€“38). In this conceptual paper, we first introduce translingualism and ELF as equally legitimate multilingual theories. Next, we identify shared key tenets and orientations as well as distinctive vantage points of each approach that researchers and educators can take advantage of for collaboratively addressing problems within classrooms and beyond. Finally, we lay out a vision for how this transdisciplinary, collaborative dialogue can help realize more equitable multilingualism.

De Gruyter
Show threadpetersuberJun 28, 2024

Update. Recent work on #AI summaries is mostly monolingual, focusing on English. "In this paper, we consequently explore how state-of-the-art neural abstract summarization models based on a multilingual encoderā€“decoder architecture can be used to enable cross-lingual extreme summaries of scholarly texts."
https://link.springer.com/article/10.1007/s00799-023-00373-2

#Multilingualism #MultilingualResearch

Cross-lingual extreme summarization of scholarly documents - International Journal on Digital Libraries

The number of scientific publications nowadays is rapidly increasing, causing information overload for researchers and making it hard for scholars to keep up to date with current trends and lines of work. Recent work has tried to address this problem by developing methods for automated summarization in the scholarly domain, but concentrated so far only on monolingual settings, primarily English. In this paper, we consequently explore how state-of-the-art neural abstract summarization models based on a multilingual encoderā€“decoder architecture can be used to enable cross-lingual extreme summaries of scholarly texts. To this end, we compile a new abstractive cross-lingual summarization dataset for the scholarly domain in four different languages, which enables us to train and evaluate models that process English papers and generate summaries in German, Italian, Chinese and Japanese. We present our new X-SCITLDR dataset for multilingual summarization and thoroughly benchmark different models based on a state-of-the-art multilingual pre-trained model, including a two-stage pipeline approach that independently summarizes and translates, as well as a direct cross-lingual model. We additionally explore the benefits of intermediate-stage training using English monolingual summarization and machine translation as intermediate tasks and analyze performance in zero- and few-shot scenarios. Finally, we investigate how to make our approach more efficient on the basis of knowledge distillation methods, which make it possible to shrink the size of our models, so as to reduce the computational complexity of the summarization inference.

SpringerLink
Show threadpetersuberJul 4, 2024

Update. "This paper analyses the attitudes of Polish linguists towards English for research publicationā€¦[The dominance of English contributes] to the sense of disadvantage associated primarily with manifold costs incurred by publishing in English."
https://www.sciencedirect.com/science/article/abs/pii/S1475158524000717
(#paywalled)

#Multilingualism #MultilingualResearch

Show threadpetersuberJul 7, 2024

Update. From 670 #OpenAccess journals published in #Africa, "only 57 journals allow publication in languages other than English [#LOTE]. From these journals, between 2010 and 2020, only 26 articles appeared in an African language. The Google Scholar site reports the highest citation for articles written in the aforementioned African languages is one."
https://www.tandfonline.com/doi/full/10.1080/14790718.2024.2369172

#Multilingualism #MultilingualResearch

Show threadpetersuberJul 28, 2024

Update. "Linguistic hegemony and language biasā€¦can pose significant threats to the scientific community, as language morphs from being a tool of communication and exchange to becoming an obstacle. Consequently, the broader scientific community misses out on knowledge generated by non-English-speaking researchers, and these researchers, in turn, face the injustice of unequal access to the broader audience that only high-impact journals can reach."
https://criticalcarescience.org/wp-content/plugins/xml-to-html/include/lens/index.php?xml=2965-2774-ccsci-36-e20240084en.xml&lang=en

#Multilingualism

Critical Care Science (CCS)

Show threadpetersuberAug 22, 2024

Update. "Multilingual Data Science: Ten Tips to Translate Science and Tech Content"
https://datasciencebydesign.org/blog/multilingual-data-science

One nice tip I haven't seen before: If you write a glossary to accompany your research, make it a bilingual or multilingual glossary.

#Multilingualism #MultilingualResearch

Multilingual Data Science: Ten Tips to Translate Science and Tech Content ā€“ Data Science by Design

Yanina Bellini Saibene & Natalia Soledad Morandeira

Show threadpetersuberSep 4, 2024

Update. "The preeminence of English as the lingua franca in global science has led to English-dominant publication practices, even in non-English-speaking countries. We examine the complex dynamics of language use in scientific publications in #Chinaā€¦The findings underscore the risk of underestimating Chinaā€™s scientific output by only counting English-language publications."
https://academic.oup.com/rev/article-abstract/doi/10.1093/reseval/rvae036/7746712

#Multilingualism #MultilingualResearch

The anglicization of science in China

Abstract. The preeminence of English as the lingua franca in global science has led to English-dominant publication practices, even in non-English-speaking

OUP Academic
Show threadpetersuberSep 7, 2024
Update. "Tony Liddicoat shows that English-language research publications about #multilingualism are by and large a monolingual affairā€¦I expand on Liddicoat's quantitative findingsā€¦Specific aspects ofā€¦English-monolingual ways of seeing multilingualismā€¦include perceptions of multilingualism as generic and context-free; as characteristic of the present; and as constituted in textual products rather than processes of production and reception."
https://www.tandfonline.com/doi/full/10.1080/17447143.2015.1102921
Show threadpetersuberSep 21, 2024

Update. "This emphasis of publishing in 'international' (i.e., English-language) journals has multiple consequences for non-native English speakers, including problems with having the quality of their English reviewed over the academic content of their work and
limited access to scientific research for those who do not speak English."
https://journals.publishing.umich.edu/jep/article/id/4847/

#Multilingualism #MultilingualResearch

Research Assessment Systems and the Effects of Publication Language: Manifestations in the Directory of Open Access Books

Research assessment is a major driver of research behavior. The current emphasis on journal citations in a limited number of journals with an English focus has multiple effects. The need to publish in English even when it is not the local language affects the type of research undertaken and further consolidates the Global North-centric view or scientific approach. The bibliometric databases on which assessments of universities and journals are based are owned by two large corporate organizations, and this concentration of the market has in turn concentrated the research environment. Open infrastructure offers an alternative option for the research endeavor. The OAPEN online open access library and the Directory of Open Access Books form part of this infrastructure and we consider the pattern of languages present in the directories over time.

The Journal of Electronic Publishing
Show threadMike Taylor šŸ¦•Aug 22, 2024
@petersuber I don't disagree, but whenever I see new requirements like this, I always find myself asking "What do you want me to do this INSTEAD OF?". Because obviously we can't all just keep adding more and more requirements.
Show threadMike Taylor šŸ¦•Jul 8, 2024
@petersuber That's surely very easy to understand. In terms of dissemination, an article written in English has a potential audience orders of magnitude greater than one written in an African language other than Swahili.
Show threadFrank SonntagApr 25, 2024

@petersuber
fwd to @TranslateScience
in case you've not heard about them (Climate Cardinals).

Maybe there is room for cooperation.

https://fediscience.org/@petersuber/112333155819842365

petersuber (@[email protected])

Update. Wonderful project. "Climate Cardinals is a youth-led nonprofit thatā€™s accomplished quite a bit with almost no funding: translating 2 million words in four years to make scientific literature more accessible to non-English speakers." #Google just gave the project $400k. https://www.forbes.com/sites/jeffkart/2024/04/22/climate-cardinals-plans-to-double-its-science-translation-efforts-with-new-google-backing #Climate #Translation

FediScience.org
Show threadLuis VillaApr 25, 2024

@petersuber I have heard anecdotally from professors who teach a lot of students from outside English-speaking countries that ChatGPT was getting used a lot as essentially a very sophisticated grammar checker, rather than the content creator that lots of people are concerned about.

That raises its own set of concerns, of course, but also easier to see how it can be a net positive for both student (less anxious, more focused on substance) and faculty (easier to read and grade).

Show threadš“¼š“®š“»š“Ŗš“¹š“Ŗš“½š“±ć€ćƒ„怑ā˜®(šŸ“šŸ‡ØšŸ‡¦)Apr 4, 2024

@petersuber

one language we all learn is just easier.
if everyone did it in their language it would be harder to do scientific work. doesnt matter what one language would be used, but for historical reasons we ended up with english and it doesnt really matter imho.

Show threadThomasMar 28, 2024
@petersuber that sounds plausible. Publish-or-perish is what made many a brilliant mind quite science in horror, and old-timers don't mince their words. Susan Haack's rant still echos in my ear.