Culturally diverse expert teams have yet to bring comprehensive linguistic diversity to intergovernmental ecosystem assessments

Abstract

•Linguistic diversity is essential for comprehensive evidence-based decision making•IPBES encourages linguistic diversity across its assessments•Despite encouragement, non-English-language evidence sources are rarely referenced•A systemic change in scientific culture can advance global assessment processes Synthesis of science and knowledge requires integration from multiple scales and diverse sources. Inherent biases and structural inequities within the scientific community favor English-language literature and Anglophone experts. We examined the linguistic diversity of assessment experts, references they consulted, comments they received, and the final reports of eight ecological assessments recently produced by the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES). We found that, despite encouragement, non-English-language literature was rarely consulted, even in linguistically diverse author teams. Such omission can bias assessments and perpetuate unequal power dynamics in science. The scientific community can work to be more inclusive. Methodological guidelines for these global assessments can facilitate this transition but, ultimately, systemic change is needed to democratize collection and representation of science and knowledge. Multicultural representation is a stated goal of many global scientific assessment processes. These processes aim to mobilize a broader, more diverse knowledge base and increase legitimacy and inclusiveness of these assessment processes. Often, enhancing cultural diversity is encouraged through involvement of diverse expert teams and sources of knowledge in different languages. In this article, we examine linguistic diversity, as one representation of cultural diversity, in the eight published assessments of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES). Our results show that the IPBES assessment outputs are disproportionately filtered through English-language literature and authors from Anglophone countries. To incorporate more linguistic diversity into global ecosystem assessment processes, we present actionable steps for global science teams to recognize and incorporate non-English-language literature and contributions from non-Anglophones. Our findings highlight the need for broad-scale actions that enhance inclusivity in knowledge synthesis processes through balanced representation of different knowledge holders and sources. Multicultural representation is a stated goal of many global scientific assessment processes. These processes aim to mobilize a broader, more diverse knowledge base and increase legitimacy and inclusiveness of these assessment processes. Often, enhancing cultural diversity is encouraged through involvement of diverse expert teams and sources of knowledge in different languages. In this article, we examine linguistic diversity, as one representation of cultural diversity, in the eight published assessments of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES). Our results show that the IPBES assessment outputs are disproportionately filtered through English-language literature and authors from Anglophone countries. To incorporate more linguistic diversity into global ecosystem assessment processes, we present actionable steps for global science teams to recognize and incorporate non-English-language literature and contributions from non-Anglophones. Our findings highlight the need for broad-scale actions that enhance inclusivity in knowledge synthesis processes through balanced representation of different knowledge holders and sources. English is the lingua franca of science,1Montgomery S.L. Does Science Need a Global Language? English and the Future of Research. University of Chicago Press, 2013Crossref Google Scholar especially in the areas of natural sciences.2Gordin M.D. Scientific Babel: How Science Was Done before and after Global English. University of Chicago Press, 2015Crossref Google Scholar Most journals indexed in Academic Rankings (i.e., with an impact factor) are written in English. Thus, publishing in English is often key to career development (e.g., citation rates,3Di Bitetti M.S. Ferreras J.A. Publish (in English) or perish: the effect on citation rate of using languages other than English in scientific publications.Ambio. 2017; 46: 121-127Crossref PubMed Scopus (91) Google Scholar job performance,4Vasconcelos S.M.R. Sorenson M.M. Leta J. Sant’Ana M.C. Batista P.D. Researchers’ writing competence: a bottleneck in the publication of Latin-American science?.EMBO Rep. 2008; 9: 700-702Crossref PubMed Scopus (37) Google Scholar mobility5Van Noorden R. Global mobility: science on the move.Nature. 2012; 490: 326-329Crossref PubMed Scopus (54) Google Scholar). There are advantages in having a common language in science and knowledge production. A common language facilitates communication across countries and cultures, which is essential in contemporary science and knowledge-building processes.6Woolston C. Osório J. When English is not your mother tongue.Nature. 2019; 570: 7760Crossref Scopus (8) Google Scholar In the absence of a common language, researchers from different regions would have difficulty working together. Ignoring linguistic diversity in science, however, can perpetuate hegemonic patterns of knowledge production by discounting the evidence base found in non-English-language publications or inhibiting it from being broadly shared, see, e.g., Meneghini and Packer (2017)7Meneghini R. Packer A.L. Is there science beyond English? Initiatives to increase the quality and visibility of non-English publications might help to break down language barriers in scientific communication.EMBO Rep. 2007; 8: 112-116Crossref PubMed Scopus (193) Google Scholar, 8Tietze S. Dick P. Hegemonic practices and knowledge production in the management academy: an English language perspective.Scand. J. Manag. 2009; 25: 119-123Crossref Scopus (44) Google Scholar, 9Vila F.X. Bretxa V. Comajoan L. ¿En qué lenguas se hace ciencia? La gestión del multilingüismo en el Parc Científic de Barcelona.Rev. Int. Organ. 2014; 13: 111-134https://doi.org/10.17345/rio13.111-134Crossref Google Scholar, 10Grandjean M. https://halshs.archives-ouvertes.fr/halshs-01526242/Google Scholar, 11Gradim A. Políticas da Língua e Comunicação de Ciência: a importância do multilinguismo no espaço Lusófono de conhecimento.Missões Rev. Ciências Humanas e Sociais. 2018; 4: 52-67https://www.academia.edu/38006452/POL%C3%8DTICAS_DA_L%C3%8DNGUA_E_COMUNICA%C3%87%C3%83O_DE_CI%C3%8ANCIA_A_IMPORT%C3%82NCIA_DO_MULTILINGUISMO_NO_ESPA%C3%87O_LUS%C3%93FONO_DE_CONHECIMENTOGoogle Scholar, Tietze and Dick (2009),8Tietze S. Dick P. Hegemonic practices and knowledge production in the management academy: an English language perspective.Scand. J. Manag. 2009; 25: 119-123Crossref Scopus (44) Google Scholar Vila et al. (2014),9Vila F.X. Bretxa V. Comajoan L. ¿En qué lenguas se hace ciencia? La gestión del multilingüismo en el Parc Científic de Barcelona.Rev. Int. Organ. 2014; 13: 111-134https://doi.org/10.17345/rio13.111-134Crossref Google Scholar Grandjean (2014),10 and Gradim (2018).11 Civil rights leader W.E.B. Du Bois's concept of “double consciousness”12Du Bois W.E.B. The Souls of Black Folk. A. C. McClurg & Co., 1903Google Scholar illuminates how non-Anglophone scholars often need to adopt the rules and structures of the systems that oppress their ways of knowing and the very foundations of their cultures to thrive in academia.13Meer N. W. E. B. Du Bois, double consciousness and the ‘spirit’ of recognition.Sociol. Rev. 2019; 67: 47-62Crossref Scopus (17) Google Scholar These systemic issues continue historic and ongoing colonization of thought.14Chilisa B. Decolonising transdisciplinary research approaches: an African perspective for enhancing knowledge integration in sustainability science.Sustain. Sci. 2017; 12: 813-827Crossref Scopus (64) Google Scholar Levels of linguistic representation differ across scientific disciplines.15de Pina Cabral J. Língua e hegemonia nas ciências sociais.Análise Soc. 2007; 42: 233-237Google Scholar,16Haße W. Peters S. Fey K.H. ¿Lingua franca impuesta o lenguas europeas de la ciencia en medicina? La opción del multilingüismo.[email protected] 2011; XII: 267-272Google Scholar For example, over a third of biodiversity conservation publications are in languages other than English.17Amano T. González-Varo J.P. Sutherland W.J. Languages are still a major barrier to global science.PLoS Biol. 2016; 14: e2000933Crossref PubMed Scopus (168) Google Scholar The number of non-English publications is arguably higher for research on Indigenous and Local Knowledge (ILK), which is often published only in local languages relevant to Indigenous Peoples and Local Communities.18Thaman R. Lyver P. Mpande R. Perez E. Cariño J. Takeuchi K. The Contribution of Indigenous and Local Knowledge Systems to IPBES: Building Synergies with Science. IPBES Expert Meeting Report, 2013Google Scholar Importantly, knowledge of Indigenous groups whose languages are endangered are also the least represented in the published literature, see, e.g., Cámara-Leret and Dennehy (2019).19Cámara-Leret R. Dennehy Z. Information gaps in indigenous and local knowledge for science-policy assessments.Nat. Sustain. 2019; 2: 736-741Crossref Scopus (20) Google Scholar Ignoring non-English-language knowledge sources can contribute to incomplete scientific understanding.20Mondada L. La science polyglotte: conditions et possibilités des interactions scientifiques plurilingues.Langues Prod. Du Savoir. Actes Du Colloq. L’académie Suisse Des Sci. Hum. 2002; : 33-42Google Scholar,21Martins M.de L. Revistas científicas de ciências da comunicação em Portugal: da divulgação do conhecimento à afirmação do Português como língua de pensamento.Intercom Rev. Bras. Ciências da Comun. 2012; 35: 233-251Crossref Google Scholar For instance, meta-analyses that omit a large proportion of literature because it is not in English could bias ecological evidence syntheses due to systematic differences in study characteristics (e.g., study species, ecosystem types) and statistical results (e.g., effect size).22Konno K. Akasaka M. Koshida C. Katayma K. Osada N. Spake R. Amano T. Ignoring non-English-language studies may bias ecological meta-analyses.Ecol. Evol. 2020; 10: 6373-6384Crossref PubMed Scopus (48) Google Scholar As one example, several studies have shown that there is extensive scientific literature on wildlife-wind farm interactions in languages, such as Spanish23Fernández-Bellon D. Limited accessibility and bias in wildlife-wind energy knowledge: a bilingual systematic review of a globally distributed bird group.Sci. Total Environ. 2020; 737: 140238Crossref PubMed Scopus (3) Google Scholar and German,24Illner H. Comments on the Report “Wind Energy Developments and Natura 2000. National Wind Watch, 2011https://www.wind-watch.org/documents/comments-on-the-report-wind-energy-developments-and-natura-2000/Google Scholar that are not broadly cited in English-language literature. Including such non-English literature would greatly amplify the sample size that conclusions are based on and may either confirm or repute conclusions based on English-language only studies. The bias also extends to global databases, which tend to be in English but require information generated worldwide to be complete. Consequently, it is not surprising that country-level data for such global databases (e.g., Global Biodiversity Information Facility, gbif.org) are more complete in countries with a higher proportion of Anglophones than those where English is rarely spoken.25Amano T. Sutherland W.J. Four barriers to the global understanding of biodiversity conservation: wealth, language, geographical location and security.Proc. Biol. Sci. 2013; 280: 20122649PubMed Google Scholar Importantly non-Anglophone policymakers and the broader public might miss relevant scientific discoveries that are only communicated in English. Several studies have shown that access to scientific information can be limited for certain groups if national languages are not used.3Di Bitetti M.S. Ferreras J.A. Publish (in English) or perish: the effect on citation rate of using languages other than English in scientific publications.Ambio. 2017; 46: 121-127Crossref PubMed Scopus (91) Google Scholar,17Amano T. González-Varo J.P. Sutherland W.J. Languages are still a major barrier to global science.PLoS Biol. 2016; 14: e2000933Crossref PubMed Scopus (168) Google Scholar,26Barath H. Indian initiatives aim to break science’s language barrier.Nature. 2019; 571: 7764Crossref Scopus (2) Google Scholar,27Lafforgue L. Science & actualité-Point de vue-Le francais au service des sciences.Pour Sci. 2005; 329: 8-9Google Scholar As a result, the transfer of scientific knowledge into local policies may be hindered.28Bortolus A. Running like Alice and losing good ideas: on the quasi-compulsive use of English by non-native English speaking scientists.Ambio. 2012; 41: 769-772Crossref PubMed Scopus (18) Google Scholar Furthermore, scientific discovery and its application can be slowed for non-Anglophones due to the linguistic burden of publishing in English.29Carli A. Calaresu E. Language and science.in: Hellinger M. Pauwels A. Handbook of Language and Communication: Diversity and Change. Handbooks of Applied Linguistics (HAL). vol. 9. Mouton de Gruyter, 2007: 525-554Google Scholar, 30Flowerdew J. Problems in writing for scholarly publication in English: the case of Hong Kong.J. Second Lang. Writ. 1999; 8: 243-264Crossref Scopus (200) Google Scholar, 31Flowerdew J. Attitudes of journal editors to nonnative speaker contributions.TESOL Q. 2001; 35: 121Crossref Scopus (206) Google Scholar People in countries where English is not widely spoken are less likely to read and publish ecological research in English-language journals,32Nuñez M.A. Barlow J. Cadotte M. Lucas K. Newton E. Pettorelli N. Stephens P.A. Assessing the uneven global distribution of readership, submissions and publications in applied ecology: obvious problems without obvious solutions.J. Appl. Ecol. 2019; 56: 4-9Crossref Scopus (49) Google Scholar,33Ramírez-Castañeda V. Disadvantages of writing, reading, publishing and presenting scientific papers caused by the dominance of the English language in science: the case of Colombian PhD in biological sciences.bioRxiv. 2020; https://doi.org/10.1101/2020.02.15.949982Crossref Scopus (0) Google Scholar which in turn can deepen global-level inequities around the access to science and implementation of sustainability actions. Language can be used as a proxy for broader ways of knowing.34Laitin D.D. What is a language community?.Am. J. Pol. Sci. 2000; 44: 142-155Crossref Scopus (102) Google Scholar,35Kassam K.-A.S. Biocultural Diversity and Indigenous Ways of Knowing: Human Ecology in the Arctic. University of Calgary Press, 2011Google Scholar The insistence on English as the language of science can exacerbate existing unequal power relationships36Turnhout E. Metze T. Wyborn C. Klenk N. Louder E. The politics of co-production: participation, power, and transformation.Curr. Opin. Environ. Sustain. 2020; 42: 15-21Crossref Scopus (158) Google Scholar,37Méndez García M.del C. Pérez Cañado M.L. Language and power: raising awareness of the role of language in multicultural teams.Lang. Intercult. Commun. 2005; 5: 86-104Crossref Scopus (17) Google Scholar and dominant epistemic cultures38Goyanes M. Editorial boards in communication sciences journals: plurality or standardization?.Int. Commun. Gaz. 2020; 82: 342-364Crossref Scopus (18) Google Scholar by reinforcing cultural imperialism.39Alves M.A. Pozzebon M. How to resist linguistic domination and promote knowledge diversity?.Rev. Adm. Empres. 2013; 53: 629-633Crossref Scopus (16) Google Scholar Such concerns have led to calls for scientists to develop mechanisms to overcome language barriers and be more inclusive of non-English-language literature, regardless of discipline.17Amano T. González-Varo J.P. Sutherland W.J. Languages are still a major barrier to global science.PLoS Biol. 2016; 14: e2000933Crossref PubMed Scopus (168) Google Scholar,40Rose D.C. Sutherland W.J. Amano T. González-Varo J.P. Robertson R.J. Simmons B.I. Wauchope H.S. Kovacs E. Durán A.P. Vadrot A.B.M. et al.The major barriers to evidence-informed conservation policy and possible solutions.Conserv. Lett. 2018; 11: e12564Crossref PubMed Scopus (50) Google Scholar,41Meneghini R. Packer A.L. Is there science beyond English?.EMBO Rep. 2007; 8: 112-116Crossref PubMed Scopus (0) Google Scholar Reaching beyond “tokenism,” institutions are seeking ways to establish more inclusive processes to incorporate diverse sources of evidence into knowledge production and synthesis.18Thaman R. Lyver P. Mpande R. Perez E. Cariño J. Takeuchi K. The Contribution of Indigenous and Local Knowledge Systems to IPBES: Building Synergies with Science. IPBES Expert Meeting Report, 2013Google Scholar,42Tengö M. Brondizio E.S. Elmqvist T. Malmer P. Spierenburg M. Connecting diverse knowledge systems for enhanced ecosystem governance: the multiple evidence base approach.Ambio. 2014; 43: 579-591Crossref PubMed Scopus (571) Google Scholar,43Sutherland W.J. Gardner T.A. Haider L.J. Dicks L.V. How can local and traditional knowledge be effectively incorporated into international assessments?.Oryx. 2014; 48: 1-2Crossref Scopus (60) Google Scholar The Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES) is a global science-policy body that aims to provide policymakers with the best available knowledge on the relationships between biodiversity and human well-being.44Diaz S. Demissew S. Carabias J. Joly C. Lonsdale M. Ash N. Larigauderie A. Adhikari J.R. Arico S. Baldi A. et al.The IPBES Conceptual Framework—connecting nature and people.Curr. Opin. Environ. Sustain. 2015; 14: 1-16Crossref Scopus (1136) Google Scholar It is the largest and most important institution of its kind. Here, we use IPBES as a case study to examine the extent of inclusion of non-English-language literature, in terms of participating experts and the knowledge consulted, in environmental assessment processes. IPBES explicitly operates on the principle of inclusion of diverse knowledge sources, facilitates dialogue between those with different values,45Stevance A.S. Bridgewater P. Louafi S. King N. Beard T.D. Van Jaarsveld A.S. Ofir Z. Kohsaka R. Jenderedijan K. Rosales Benites M. et al.The 2019 review of IPBES and future priorities: reaching beyond assessment to enhance policy impact.Ecosyst. People. 2020; 16: 70-77Crossref Scopus (22) Google Scholar and “recognize[s] and respect[s] the contribution of ILK to the conservation and sustainable use of biodiversity and ecosystems.”46Hill R. Adem Ç. Alangui W.V. Molnár Z. Aumeeruddy-Thomas Y. Bridgewater P. Tengö M. Thaman R. Adou Yao C.Y. Berkes F. et al.Working with indigenous, local and scientific knowledge in assessments of nature and nature’s linkages with people.Curr. Opin. Environ. Sustain. 2020; 43: 8-20Crossref Scopus (83) Google Scholar Thus, IPBES actively encourages use of non-English-language sources and even supports a task force specifically dedicated to facilitating the inclusion of ILK.47McElwee P. Fernández-Llamazares Á. Aumeeruddy-Thomas Y. Babai D. Galvin K. Guèze M. Liu J. Molnár Z. Ngo H.T. Reyes-García V. et al.Working with indigenous and local knowledge (ILK) in large-scale ecological assessments: reviewing the experience of the IPBES global assessment.J. Appl. Ecol. 2020; 57: 1666-1676Crossref Scopus (31) Google Scholar Several studies have already examined regional representation among the experts who participate in IPBES's different bodies and expert groups,48Montana J. Borie M. IPBES and biodiversity expertise: regional, gender, and disciplinary balance in the composition of the interim and 2015 multidisciplinary expert Panel.Conserv. Lett. 2016; 9: 138-142Crossref Scopus (31) Google Scholar, 49Timpte M. Montana J. Reuter K. Borie M. Apkes J. Engaging diverse experts in a global environmental assessment: participation in the first work programme of IPBES and opportunities for improvement.Innov. Eur. J. Soc. Sci. Res. 2018; 31: S15-S37Crossref Scopus (27) Google Scholar, 50Heubach K. Lambini C.K. Distribution and selection of experts in the intergovernmental science-policy platform on biodiversity and ecosystem services (IPBES): the case of the regional assessment for Africa.Innov. Eur. J. Soc. Sci. Res. 2018; 31: S61-S77Crossref Scopus (11) Google Scholar which, to our best knowledge, is the closest proxy we have to understand broader patterns of cultural diversity within IPBES. Our study widens the lens with which representation is examined in IPBES to include other aspects of cultural diversity, such as language (Note S1). Through five metrics, we analyzed linguistic diversity across eight IPBES assessments. Our results show that, despite having diverse expert teams, the IPBES assessment outputs are disproportionately filtered through English-language literature and authors from Anglophone countries. We examined linguistic diversity across four thematic assessments (Pollination, Scenarios and Modeling, Land Degradation and Restoration, Global) and four regional assessments (Africa, the Americas, Asia and the Pacific, Europe and Central Asia). We coded assessment experts, references (language and first author), comments, and final reports by language, nationality, and country of affiliation, as appropriate (Table 1; Figures S1–S7). We considered language, nationality, and country of affiliation here to be a proxy of cultural representativity. We identified Anglophone affiliations by the 18 countries recognized by the UK government as being “majority native English speaking” (listed in Note S2). These results can inform the inclusion of linguistic diversity in the second work program of IPBES and other global initiatives.Table 1Summary of metrics, methods, results, and recommendations regarding linguistic diversity representationMetricMethodologySummary of resultsRecommendations for representationAssessment expertsFor each of the eight assessments included in the analysis, we recorded the nationality/ies of each expert. The expert list included chairs, coordinating lead authors, lead authors, review editors, and fellows.The Americas assessment had the fewest nationalities overall (25) and highest proportion of Anglophone affiliations (7). The Global assessment had the highest number of nationalities (54). The Europe and Central Asia assessment had the smallest number of Anglophone affiliations (3).-Invite diverse expert teams through representative nomination and selection processes, including Indigenous and Local Knowledge (ILK) holders and experts.-Add contributing authors to fill in expertise gaps and broaden the diversity of knowledge sources consulted.-Provide best practice guidelines for improving group dynamics developed by those for whom English is not a first language.-Facilitate training opportunities for active participation among multicultural teams.Assessment referencesFor each of the eight assessments included in the analysis, we recorded language of the available references. We also randomly selected approximately 150 references per assessment and recorded country/ies of affiliation for the first author of the references.References examined totaled 27,891 across all 8 assessments, corresponding to 28 languages. English was, by far, the most common language (96.6% of references). The Europe and Central Asia regional assessment had the highest total number of languages represented by references (21 different languages), but the Americas regional assessment had the highest proportion of references in a language other than English (7%), and the Asia and the Pacific regional assessment had the least (0.15%). See Figure 1.In the subsample of references examined for first author country of affiliation, across all assessments 51% of references had a first author from an Anglophone country. The Scenarios assessment had the highest proportion of Anglophone first authors (62%) and the Europe and Central Asia assessment had the lowest (27%).-Facilitate searches for literature and knowledge in languages other than English.-Enable systematic review protocols that include local language search terms.-Provide guidance on how to include diverse forms of knowledge and evidence, including gray literature and ILK.Assessment commentsFor each of the seven assessments included in this analysis (comments were not publicly available for the Global assessment), we recorded country/ies of affiliation of the reviewer. We examined all reviewer comments for the First Order Draft, Second Order Draft, and the Summary for Policymakers (SPMs). We separately noted the number of reviewer comments made by government representatives and external reviewers. A total of 42,126 comments were coded.Ninety-four countries were represented by reviewer affiliation. A total of 32.9% of comments across all assessments came from Anglophone countries. The UK provided the highest number of reviewer comments (16%), followed by Germany (8.6%), the US (8.5%), Canada (5.50%), France (5.49%), South Africa (5.4%), and Switzerland (5.2%). The Pollination assessment received the highest number of comments (11,306) and the Scenarios assessment received the lowest (3,116).-Actively encourage non-Anglophones to provide comments.-Support submission of comments in any language.-Facilitate translation of input into multiple languages.Assessment documentFor each of the eight assessments included in this analysis, we recorded the languages in which the approved assessment reports are available for the public to view and/or download. Three versions of the assessment reports exist: the SPMs as plain text, the SPMs as a laid out version (i.e., visually friendly version), and the full report as plain text only.All plain text versions of the full reports and SPMs were available for download in English. All the SPMs could also be downloaded in the other five United Nations languages as plain text. Laid out versions of the SPMs were available in English for all assessments. In addition, the Pollination assessment was also available in Chinese and French, while the Scenarios and Modeling SPMs could also be downloaded in Chinese and the Global assessment was available in Czech and Japanese. None of the full reports (i.e., the detailed documents sustaining the findings reported in the SPMs) were available in any language other than English.-Publish assessment reports, or at minimum extended abstracts, in multiple languages.-Encourage synchronous interpretation during plenaries.Linguistic diversity was examined in eight assessments of the Intergovernmental Science-Policy Platform for Biodiversity and Ecosystem Services (IPBES). Open table in a new tab Linguistic diversity was examined in eight assessments of the Intergovernmental Science-Policy Platform for Biodiversity and Ecosystem Services (IPBES). Across the eight assessments, experts collectively represented 106 nationalities (54.9% of 193 United Nations member states). The majority of IPBES experts represented non-Anglophone countries with only ten Anglophone countries represented (9.4% of IPBES assessment expert nationalities compared with 9.3% of countries being Anglophone; Figure 1C). The Americas assessment had the fewest nationalities overall (25) and highest proportion of Anglophone countries (7). The Global assessment had the highest number of nationalities (54). The Europe and Central Asia assessment had the smallest number of Anglophone affiliations (3). Some countries, such as the US and UK, were disproportionately represented across all assessments compared with many countries in Africa and Asia. References across all assessments were overwhelmingly in English (96.6%; Table S1; Figure 1A), followed by some regionally important languages, such as Spanish for the Americas regional assessment (5.5%), Russian for the Europe and Central Asia regional assessment (4.5%), and French for the Africa regional assessment (2.3%). Linguistic diversity was particularly low among references cited in the Global assessment (99.6% of references were in English) and the Asia and the Pacific regional assessment (only 5 out of 3,368 references were in a language other than English; 0.15% of total) despite the existence of significant collections of non-English scientific publications in the region (e.g., Chinese and Japanese literature). Similar to reference language, first author affiliation for references revealed an overrepresentation of Anglophone countries when compared with Scimago Country Rank (www.scimagojr.com/countryrank.php) which tracks the number of scientific documents by country (Figure 1B). In the subset of references analyzed, 51% were first-authored by individuals in Anglophone countries, even though, worldwide, only 9.3% of countries are Anglophone. The proportion of Anglophone affiliations for first authors ranged from 27% (Europe and Central Asia regional assessment) to 62% (Scenarios assessment). The four regional IPBES assessments show some additional patterns, which, for the most part, align with their given regional foci (Table S2). The Americas assessment, for example, shows dominance of the US, UK, and Canada, with 36.7%, 10.6%, and 9.3% of references, respectively (this is the most unbalanced dominance of Anglophone countries of all eight assessments). A key component of the IPBES knowledge synthesis process includes the opportunity for scholars and stakeholders to review and comment on multiple drafts of the asse