After Ernst Gombrich failed to get the Bilderatlas Mnemosyne into a publishable form in the 1930s, a number of lists identifying individual images remained in the Warburg Institute Archive. Dorothée Bauerle was able to draw on these sources when she conducted the first comprehensive investigation of the Bilderatlas in the late 1970s. In the late 1980s, she left her research notes to the Transmediale Gesellschaft Daedalus. The Viennese group (Werner Rappl e.a.) began its research in around 1990 and published the entire Bilderatlas with a near complete set of captions in 1994. Another decisive advance in knowledge about the origins of the Bilderatlas during this period was Peter van Huisstede’s PhD thesis, De Mnemosyne beeld-atlas van Aby M. Warburg: een laboratorium voor beeldgeschiedenis (1992). The Daedalus captions appeared in expanded form in the Bilderatlas that Martin Warnke published in 2000 in the context of Warburg’s Gesammelte Schriften, but they still contained various errors and omissions. These were only sporadically corrected in the translations—into French, Italian, Spanish, Hebrew, Japanese, and Polish.
In May 2012, the Forschungsgruppe Mnemosyne in Hamburg launched an extensive series of events and exhibitions on the Bilderatlas that took place at the venue 8. Salon, in Hamburg. The group’s members included among others Marcel Hüppauff, Jochen Lempert, Christian Rothmaler, Katha Schulte, Philipp Schwalb, Kolja Gollub, Regine Steenbock, Knut Wittmaack, and us. Above all, our focus was on developing a commentary about each panel; the captions were only occasionally checked and corrected. In spring 2016, we asked the Warburg Institute if we could search its Photographic Collection for the original images for two panels from the 1920s. The idea was to show panels 32 and 48 as part of an exhibition of the reconstruction of the Bilderatlas by the Forschungsgruppe Mnemosyne planned for the following summer at the ZKM │ Zentrum für Kunst und Medien, Karlsruhe, at the invitation of Peter Weibel. Most of the images that Warburg had handled himself were located relatively quickly, so we were encouraged to expand the experiment and search for all of the “original” panels. In autumn 2016, we gained the support of David Freedberg, then Director of the Warburg Institute, who gave us the go-ahead for the project. He had intended to show the original version of the Bilderatlas in London, but resigned as director just a few months later, bringing the cooperation to a halt for the time being.
In 2018, HKW in Berlin responded positively to our proposal to stage an exhibition of the recovery of the original version, and this initiative led to a revival of the cooperation with the Warburg Institute. Bill Sherman, the new director, put great effort into finally making our undertaking possible. Some periods of intense research in the Photographic Collection in 2019 achieved the result we had hoped for, identifying about 80 percent of the originals for the Bilderatlas as a whole. We were supported by the Institute’s archivist, Claudia Wedepohl, and Paul Taylor, and Rembrandt Duits from the Photographic Collection team, and crucially by Lorenza Gay. Mateusz Sapija, Susanne Förster, and Amirkhan Saifullin also worked with us at various times in the Photographic Collection. At the same time we reviewed and amended the captions systematically, with Joacim Sprung and Giovanna Targia supplying valuable pointers from afar.
In the end, less than a dozen of the 971 images still lack certain data on location, place of publication, or authorship. The numbering system used in the Gesammelte Schriften was retained on the grounds that it has become the international standard, even if it is inconsistent in places. Where it was not possible to find the originals in the Warburg Institute we drew on the Photographic Collection and the Warburg Institute Library to identify substitutes. Apart from anything else, in certain cases we were forced to use “originals” that had been altered in the interim—for example, cropped or remounted. Keeping the provisional character of Warburg’s work was uppermost. Thanks to the kind support of Bill Sherman and Uwe Fleckner we were also able to access material from the collections of the Zentralstelle für wissenschaftliche Sammlungen, Universität Hamburg. We also drew on data from Daedalus and from the archive of Forschungsgruppe Mnemosyne / fluid.
With the publication of the Folio volume Aby Warburg: Bilderatlas Mnemosyne – The Original in April 2020 we were able to present the state of our research.
Roberto Ohrt and Axel Heil, curators