Science Gossip: new dataset now available

cover of science gossip

Science Gossip is one project within the AHRC-funded Constructing Scientific Communities initiative, led by Professor Sally Shuttleworth at the University of Oxford, together with Gowan Dawson (Leicester) and Chris Lintott (Oxford) between 2014 and 2019. This project brought together historical and literary research in the nineteenth century with contemporary scientific practice, and looked at the ways in which patterns of popular communication and engagement in nineteenth-century science can offer models for the present day. Science Gossip (which was led by Dr Geoffrey Belknap in collaboration with the Missouri Botanic Garden) takes as its source material the pages of 16 natural history periodicals from the nineteenth century, digitised by the Biodiversity Heritage Library. Using the citizen science platform Zooniverse, nearly 10,000 online volunteers tagged more than 160,000 illustrations in these digitised pages, adding information on subjects, places, species, artists and engravers. A document setting out the methodology and format for data entry is available here:

The resulting classifications – numbering nearly half a million – constitute a valuable resource for historians investigating why, how often, and who made images depicting a whole range of scientific subjects in the Victorian period. The Science Gossip dataset, combining both images and classifications, is freely available on the SDS platform for browsing, searching and re-use: Each digitised page has a separate record: click on a record to view the image and associated metadata as reported by the citizen scientists.

The dataset can be directly searched here: It is a free text search; use quotation marks to search for specific names or phases, e.g. “Edward Kite” (omitting the quotation marks would return any records containing the words Edward or Kite). It is also possible to search keyword tags directly, using the format :keyword: Oxford, or :keyword: Edward Kite (no quotation marks needed). Within each record, keyword tags are clickable to return other records with matching tags. More information about searching the platform can be found on the Figshare website.

The project dataset can be explored in its entirety, or through a series of subfolders, one for each periodical, as listed below.