Methods, tools and workflow

Linked resources

Wikidata was the central hub for this project. Wikidata is an open, collaborative and multilingual knowledge base providing structured data that can be read and edited by humans and machines (Vrandečić and Krötzsch 2014, Shafee et al. 2023). As a hub for other identifiers, it plays a central role in semantic linking and enriching of data (e.g. Groom et al. (2020), Groom et al. (2022), Page (2022), Page (2023)). Wikidata properties mentioned in the text are given in italics followed by the Wikidata entity link.

One advantage of using Wikidata as the hub for the project was the ability to use the Wikidata Query service to answer specific questions needed for improvement of the dataset. Pulling information from multiple sources and verifying and referencing each one to build the new dataset is time-consuming, but each interaction allowed us to query the resulting data (or subsets of it) with the Wikidata Query Service using SPARQL. For an example of such a SPARQL query, see Suppl. material 1. Wikidata also provides a the so-called Query Builder, a tool that empowers users with minimal coding/scripting knowledge to build simple Wikidata queries (e.g. botanists from a particular geographic area or plants named for fictional characters).

The key for answering the original questions posed were links between various elements in the dataset. The principal linkages used in creation of the dataset are shown in Fig. 1.

Linking plant genera and the women for whom they were named was the central link in our workflow. Once Wikidata items existed for all plant genera and all women (or female beings) honoured, these items were linked using the property named after. Each Wikidata item for a plant genus honouring a female was linked via a referenced statement to the item of the honoured woman or female being. It was this central link that allowed us to address our initial question.

To find information on eponymy contained in the original places of publication for generic names (protologues), we searched botanical databases, such as IPNI and Tropicos that often provide links to BHL. In other cases, we used openly accessible providers of digital publications, such as Google books, Hathi project and other digital libraries and websites that provide free access to such publications. BHL was invaluable to this research project as the majority of protologues (457, i.e. over 60%) were openly accessible through this digital library. Some texts were either in copyright or were missing from BHL as well as other providers of digital publications. Where we were unable to freely access the protologues digitally, we consulted the libraries accessible to us at our own institutions. During this process, we collated a list of all those publications that appeared to be in the public domain, but were missing from BHL. We passed on this list to BHL and requested these “missing” texts be scanned and added to BHL for the benefit of everyone. As of writing, BHL staff are working on this request.

Women for whom plant genera were named include both female beings from mythology (mythical women) and those who are documented to have actually lived; we termed these non-mythical women (see section on People below). For non-mythical women involved in collecting or identifying specimens, we created or improved their profiles in Bionomia and attributed specimens collected or identified by them. For deceased women, we made their Bionomia profiles public and added their Bionomia identifier to the women’s Wikidata items. If the woman were living and had registered for an ORCID, we were able to use this identifier to attribute specimens to her in Bionomia. However, while profiles of deceased people can be made public, living people have full control over their Bionomia profile and decide for themselves to make it public or remain private.

Over the course of our work, we found small errors in publicly available datasets that are used by many botanists. These publicly available datasets were created over decades using a variety of methods (Croft et al. 1999). They are immensely useful and the community is invited to improve them. Corrections or newly-researched information on taxon authors, protologues and eponyms was sent together with the corresponding references to IPNI and Tropicos and updated in these databases accordingly. As a result of our research into the protologues, we frequently contacted the IPNI editors with corrections (e.g. dates or page numbers) or enhancements, such as standardising author names or adding links to protologues that were difficult to track down. We also consulted the IPNI team to discuss difficult nomenclatural issues. We edited the Tropicos database directly with corrections and improvements. Missing IPNI and Tropicos genus identifiers were added to the Wikidata items of the respective genera.

Generic names of plants

The starting point for our list of plant genera named for women (both mythical and non-mythical) was the “Index of Eponymic Plant Names - Extended Edition” by Lotte Burkhardt (Burkhardt 2016, Burkhardt 2018) available online as a searchable PDF file. We manually extracted all genera honouring women (Fig. 1, Workflow A) and created a simple table. Burkhardt published an updated version in 2022 and kindly provided us with a list of additional genera included in that edition (Burkhardt 2022). This dataset was supplemented with data from IPNI (2023) and other sources including Baines (1981), Quattrocchi (1999), Charters (2015) and Mari Mut (2021), as well as suggestions received from colleagues and generated from our own research. We extracted the available etymology data in IPNI for all included genera to obtain any relevant information on the origin of a name. In some cases, we contacted the authors directly via email to ask who their name was honouring.

Other genera named for women were found using the code of nomenclature governing plant naming (ICN; Turland et al. (2018)), where clear rules are set out, in, for example, Rec. 60B. This recommendation sets out good practice for forming generic names by adding -a, -ea or -ia to the person’s name and Note 2 describes how to modify names to name different genera after the same person using a prefix or suffix. Recommendation 60H suggests providing the etymology or source of the name in the protologue when publishing the name of a new plant at any rank to give some indication of the derivation of the name.

Our dataset covers angiosperm genera (flowering plants) named after perceived women and female beings. Hybrid and fossil genera were excluded as well as names that were not validly published, while illegitimate names were included (see ICN for an explanation of these nomenclatural terms; Turland et al. (2018)). For genera published by French naturalist Constantine Rafinesque, we excluded those that appeared to have been named after nymphs that did not have any clear reference to Greek or Roman mythology. It has been suggested that these names were probably “made out of thin air” (Bernhardt 2008, Mari Mut 2019).

For all genera, the protologues were reviewed to confirm the etymology or eponymy. Newer publications usually include a specific “etymology” section explaining the origin of the assigned, in most cases, honorific name (Agosti et al. 2022). Sometimes a specific dedication text is given stating the role or specific achievements of the person. In other cases, the name of the person was not specified in the dedication, but could be indirectly implied, for example, as a collector (see Clemensia Merr.) or in the name of the type species (e.g. Dorothea talbotii Wernham). Dedications were copied in the original language into our working sheet accompanied at times with a translation. Where possible when citing the protologue in support of the named after statement in Wikidata, we quoted the protologue dedication text. For about a third of the genera (223), there was no ascription given in the protologue. When the etymology was not stated in the protologue, but by Burkhardt (Burkhardt 2016, Burkhardt 2018, Burkhardt 2022), we used these references to support the named after statement in Wikidata.

To be able to answer our research questions, we collected the following information as structured data: the year and place of publication (nomenclatural reference), the author(s) of the name and the digital link to the protologue or original publication (if available online) plus information on the honoured women (see section “People” below). A full list of data fields is given in the Data dictionary (see column field and description below).

Several genera were named after their similarity to other genera that were named for women. We classified such eponymy as “indirect”, but included them in our study since the derived eponyms are also honouring women. In Wikidata, these have been noted by adding named after statements linking the item of the new genus to both the genus after which it was named and to the woman/female being who was originally honoured and, therefore, honoured in the new genus indirectly. This information on indirect eponymy was entered in our data table in the column eponymy-derived-indirectly. For example, the genus Afrofittonia Lindau is named for the genus Fittonia Coem. and, thus, indirectly honours Sarah and Elizabeth Fitton, for whom Fittonia was named. In some cases, a replacement name (i.e. a name coined to replace a generic name that is outside the rules of the ICN) is coined honouring the same person as the original generic name; in these cases, two generic names, both honouring the same woman, apply to the same genus. These names, too, were treated as “indirect” eponymy with a reference in the column eponymy-derived-indirectly, since both generic names refer to the same taxonomic entity.

Wikidata items for the plant genera were updated where necessary or, in a few cases, newly created with information about the genus name, the author(s) of the genus, the year of the original publication and these statements were referenced using the original publication. If the protologue were available on BHL, the BHL bibliographic or page number was added to Wikidata, thus creating a digital link improving access to the protologue.

People

After the list of genera was established, we ensured all women being honoured had items in Wikidata (Fig. 1, Workflow C). We researched each person and her contributions, plus information on mythological figures where necessary. We enriched existing or created new Wikidata items using this research. Wikidata notability criteria were followed when creating new items. Our work also included disambiguation from other people with identical or similar names (Groom et al. 2020, Groom et al. 2022). We added any relevant identifiers for the woman to her Wikidata item, thus creating connections to other databases. Examples of such identifiers include library identifiers sourced from the Virtual International Authority File (VIAF), the Harvard Index of Botanists identifier, ORCID identifier (ORCID iD) and genealogical database identifiers, such as FamilySearch, WikiTree, Find A Grave and Geni.

An example of a woman whose contribution to science has been amplified with multiple links through this project is Clara Wehl (1833–1901), a German-born Australian botanical collector and sister of famous botanist Ferdinand von Mueller, who honoured her with the genus Wehlia F.Muell. About a year after her Wikidata item was created, links and connections added to Wehl's entry ensured she also became a ‘notable’ figure for the purposes of Wikipedia. An article about her was created, making her contributions more visible to users of that platform as well as to the global general public. If the women honoured in our dataset had written scientific publications, we linked the publication Wikidata items to the women’s Wikidata items via the author property.

We acknowledge that gender is a social construct; for this research, our statements on who was a woman or female being were based on how a living person publicly expressed their gender (e.g. through pronoun use). For deceased women or female beings, their implied gender was inferred (perceived gender) either through recorded pronoun use or gender statements in secondary sources. There were two levels of confidence in assigning gender to the person being honoured. High confidence arose when we could cite a source (the protologue or another published source), had direct confirmation from the author of the genus (such as with Lihengia Y.S.Chen & R.Ke) or knew the person honoured ourselves. If missing, the perceived gender was added to Wikidata.

We also included, but with less confidence, genera that appeared to have been named for women. Here, we recognised a genus name as likely honouring a woman, but had no direct evidence, such as a female pronoun or feminine grammar in some languages, given in the protologue. For example, the protologue for Joycea H.P.Linder merely states “a new genus, named Joycea after Joyce Vickery, who did such excellent work on the Australian grasses” (Linder and Verboom 1996: 606). In these cases, further biographical research was done to confirm that the eponymy honoured a woman. In addition, primary references for some names such as Genlisia Rchb. and Pandorea Spach, presumed to be named for the 18^th century Madame de Genlis and the mythical woman Pandora, respectively, were lacking. We, therefore, cited Burkhardt (2022) in Wikidata for these names. Burkhardt is continuing to update her list (pers. comm.) and some sources may be available after publication of this paper.

Occasionally, we found errors of attribution that had originated from our source data that we had entered into Wikidata as named after statements. In these cases, we left the original statement in the Wikidata item, but corrected the error by deprecating the original statement with a reference to the source of the correction. We then added a correct/new source of the etymology as a supporting reference for an additional named after statement. This happened for the genus Aster L., at first attributed to the goddess Astraea (Burkhardt 2018), but corrected to being named for its ‘star shape’ as there is no concrete evidence of its being named for Astraea rather than for its resemblance to a star.

We researched occupations of the honoured women and added or updated statements in Wikidata using the property occupation. Many of the more historic women did not have a formal education or a paid profession, but were still active and productive in various fields. This information can be used to analyse, for example, the group of female honourees and prevalent professions or occupations.

Our work also included enriching Wikidata items of taxon authors and compiling certain data to answer specific research questions. IPNI and Tropicos were searched for these author names first and then websites such as BHL, the Global Biodiversity Information Facility (GBIF) or other specialist databases were consulted. We compiled information, such as the nationality and gender of the taxon authors, as well as their relationships to the women after whom they named the genera.

Perceiving the gender of plant name authors was more challenging in the case of names in certain languages and cultures. In these cases, we searched for additional information about the person or contacted colleagues from the respective country. If the genus were recently published, we contacted an author directly if possible. Newly-obtained data were added to Wikidata items of the taxon authors using referenced statements and updated in the “Gendered Author List International Plant Names Index 2020” (Lindon et al. (2020) and subsequent updates to the dataset).

Results

Another publication analysing and discussing the dataset in detail is forthcoming. What follows is a brief summary of results. Our initial query in Wikidata resulted in about 40 plant genera named for women. As a result of our project, we have now a dataset of 728 genera honouring women or female beings. This was a nearly twenty-fold increase in the number of genera linked to women. A total of 31 genera were named after their similarity to other genera that were named for women, i.e. their eponymy is “indirect”.

Geographic coverage

Since we looked at all angiosperm genera named for women or female beings, our geographic coverage was automatically global and the plants included are distributed in all parts of the world. Nevertheless, there is bias related to the countries in which the honoured women were born and/or worked. There are 56 distinct countries associated with the women (country of birth and/or other close association) in our dataset: the ten most commonly occurring (see Fig. 2) are present in roughly 73% of all naming events (see definition of field eponymy-country in the Data dictionary, see Column field and description below).

The dataset includes 134 distinct angiosperms families that have at least one genus whose name honours a woman. As expected, the largest number of eponyms honouring women, both mythical and non-mythical, corresponds to the most taxon-rich angiosperm families (Fig. 3).

For published dataset: Creative Commons Public Domain Waiver (CC-Zero)

For published paper: Creative Commons Attribution 4.0 International (CC BY 4.0)

Angiosperm genera named for women

1