“Books, books, a hundred thousand books …”

“Books, books, a hundred thousand books …” – covering some 94,000 accessions, the number of objects under examination in the project dedicated to identifying Nazi-looted books at Bonn University and State Library falls just short of the number cited in the famous appeal by Professor Abronsius of Königsberg. Nevertheless, the sheer volume of items requiring autopsy within this book-centred provenance project forms the focus of the following reflections. Object autopsy constitutes the most important and therefore fundamental component of provenance research. The very highest degree of thoroughness is essential here, since otherwise indispensable clues to the provenance or biography of an object may be lost.

Based on observations made in the Bonn project, this article therefore seeks to offer a plea for the most meticulous possible object autopsy, illustrated based on selected examples, and to help close a certain gap that is noticeable in handbooks relating to book-centred provenance projects. It should be noted that the aspects presented here do not amount to a systematic evaluation but to observations made incidentally in the course of an ongoing project. What is more, the author is of course aware that he may simply be telling librarians what they already know. The hope remains, however, that the following thoughts may provide a stimulus for discussion or encourage further engagement with the topic. The need for such discussion has emerged clearly from many conversations with colleagues, and their suggestions have also provided input for this article.

It is frequently the case that a comparatively large number of individual items require investigation on grounds of suspicion, so this can typically be described as a characteristic feature of book-centred provenance projects – especially in contrast to projects working with other types of object or collection. Large looted collections have rarely survived intact en bloc; more often, looted items are found in differing proportions in numerous sections of a library’s historical holdings, if not all of them. In such large-scale collections (Zuschlag 2022, pp. 95 f.), this frequently results in an increased need for autopsy of those holdings for which initial suspicion has been raised. The work involved is further intensified by the materiality of books as objects: unlike paintings or sculptures, a complete investigation entails the examination of significantly more physical material. 

Just as it would be inconceivable not to examine a painting or a sculpture from every angle in the course of provenance research, the same approach must also be applied to books, even if they sometimes contain more pages than researchers might wish. Faced with a thousand-page large-format encyclopaedia volume, it is easy to succumb to the temptation of limiting scrutiny to the binding and preliminaries; but nobody would dream of considering only the front of a painting or neglecting to look beneath a sculpture simply because the effort seemed too great (“finding on the reverse side”, cf. Zuschlag 2022, p. 85).

How, then, to address the problem of sheer mass? After all, even after applying the method of exclusion – i.e. designating volumes as “not suspicious” based on sources such as accession registers or other finding aids – a large number of books often still remain potentially suspicious. A still greater problem arises when such supplementary sources relating to the object are partly or entirely lacking.

If, in search of a best-practice example, one consults standard handbooks on provenance research, one is struck by the absence of any normative guidelines in this regard. An evaluation of four widely used, high-quality handbooks and introductions shows that, while they naturally stress that an autoptic examination must be undertaken and indicate which features (usually only selectively) may be found in or on a book, they do not state how precisely such an examination is to be carried out or what specifically requires attention (Albrink 2006, esp. pp. 161–163; Alker 2017, esp. pp. 16–25; Leitfaden 2019, esp. pp. 45–56; Zuschlag 2022, esp. p. 85 f.). Only the importance of recording “all indications” is emphasised, i.e. all provenance features present in a book (Alker 2017, esp. p. 17). A similar picture emerges in published research outcomes that also address the methodology underlying the analysis, such as the work on Nazi-era provenance research at Salzburg University Library (Schmoller 2012, esp. pp. 129–133) – though this is by no means intended as a criticism!

A three-stage procedure is proposed here which has been successfully implemented at ULB Bonn. Once a collection has been selected for investigation, the number of books to be examined more closely can first be divided into library acquisitions from before and after the benchmark year 1933. Acquisitions between 1933 and 1945 are of particular interest; for post-1945 accessions, only antiquarian purchases or accessions to stock are relevant.

The second step can then involve carrying out a superficial autoptic examination. By surveying the external parts of each book (binding, title pages, etc.), further items can be cleared of suspicion. If the year of publication is compared with the year of accession, for example, it is possible to determine whether an item was a new purchase by the library. Other combinatory exclusion criteria are conceivable here but these depend entirely on the specific characteristics of the holdings.

In the third and final step, all volumes not yet cleared of suspicion are subjected to a detailed autoptic examination, i.e. a perusal of the book from all sides, page by page, so as to enable, where relevant, a definitive and well-founded suspicion of looted material to be articulated. In a page-by-page examination, it is not sufficient to check only the margins left blank beside the text area (gutter, head, fore-edge and tail margins), since features may also be found in the text block itself, placed there to protect them from being cut off during rebinding, for instance. If a book has already been trimmed, only minimal traces of formerly present features often remain.

Finally, features that have been completely removed must also be documented, even if little or nothing of the original remains: washed-out stamps can leave cockled paper; scraped-out marks can leave rough or translucent areas. A high degree of thoroughness is required in the examination, despite the inherently taxing nature of the process.

Admittedly, statistically speaking, provenance features are most often found on the outer parts of a book – on the binding and the edges, on the pastedowns, on the title page and on the final page – since these areas were the most accessible both for the original application of features and for their discovery today. This is where the bulk of features occurs, but by no means the entire corpus. At times these parts are no longer extant, however, for example because of rebinding, during which the original cover and pastedown were removed without fraudulent intent and discarded as not worth documenting (cf. also Alker 2017, p. 18).

Such conservation-driven measures whose primary aim is of course the preservation of the text-bearing block may have been necessary for various reasons: normal mechanical wear and tear in library use or – as was particularly the case at ULB Bonn, damage or contamination caused by war, including fire, water or evacuation. Other forms of damage can also account for missing components, such as paper decay, in which the outer parts of a book disintegrate because they are most exposed to air, while previously loose internal parts are easily lost in the course of routine library operations.

For the above reasons, a page-by-page autoptic examination is indispensable for the most complete possible identification of all cases involving looted material, whereas a superficial examination – though much quicker and therefore applicable to a larger number of volumes at the same time – will often not suffice. Here are three examples:

1. The first case concerns a looted volume with a social democratic context of seizure, which found its way to Bonn University Library in September 1935 through the mayor of Oberwesel (ULB Bonn, shelfmark 51/3461+2; Fig. 2). At first glance this book displayed no suspicious provenance features on the outer parts mentioned above. The fact that closer scrutiny might be worthwhile was partly due to its belonging to the left-wing political spectrum (Karl Kautsky, Terrorismus und Kommunismus. Ein Beitrag zur Naturgeschichte der Revolution, 2nd edition, Berlin, J. H. W. Dietz Nachfolger G.m.b.H., 1925). What is more, the volume had been rebound at an unknown but relatively recent date, with the original binding and pastedowns and the flyleaves no longer extant. This suggested that provenance features might have been destroyed in the process but could still have survived within the text block. The effort of examining all 329 pages was not in vain: on pp. 61 and 241 there were clear impressions of a circular stamp (39 mm, blue ink) bearing the inscription: Sozialdemokratische Partei · Bezirksvorstand / Bezirksverband / Obere Rheinprovinz [Social Democratic Party · District Executive / District Association / Upper Rhine Province]. In-house research then revealed that the missing front sections of the book had at one time, before rebinding, been copied in another context. This simple copy, fortunately preserved and now of unique evidential value, shows that the same stamp had once appeared on the recto of the flyleaf. On this basis, the remains of an erased blue-pencil entry (probably the number “50”) visible on the original printed wrapper title that had been affixed to the new binding acquired fresh meaning: such numbers were normally applied by the confiscating authority in order to facilitate the allocation of books to confiscation lists. The seizure of books of this provenance in the course of the National Socialist measures against social democratic institutions has since been corroborated by archival material, and the relevant findings have been deepened accordingly. The resulting page-by-page examination of the entire consignment delivered by the mayor of Oberwesel to Bonn University Library (35 volumes) revealed a further six volumes from the holdings of the SPD, as well as five from the library of the social democrat Wilhelm Schneider (†1995) and one from a trade union provenance (Deutscher Eisenbahner-Verband, Ortsverwaltung Bingen [German Railway Workers’ Union, Local Branch Bingen]). Without a page-by-page autopsy these would not have been identified.

2. A similarly complex case with respect to provenance features involves 17 valuable scholarly works published between 1797 and 1889, including writings by Alexander von Humboldt, looted from Ukraine during the Second World War. These volumes did not find their way to ULB until 2016, by way of a donation. Provenance features from the front sections of the books had been cut out by a previous owner (e.g. ULB Bonn, shelfmark W4’2021/503 (2,1); Fig. 3), which made detailed examination necessary for the reasons already discussed. This brought to light a variety of provenance features that had survived inside the books and which had apparently been overlooked when the earlier erasures were carried out. In addition to provenances from Kyiv, there were in particular ownership stamps originating from the library of the School of Agriculture and Horticulture in the Ukrainian city of Uman (oval stamp, 27 x 47 mm), which still exists today as Uman National University of Horticulture. In one volume (ULB Bonn, shelfmark W4’2021/505 (1,1); Fig. 4) a historical library slip from Ukraine was even preserved, inserted at p. XLIII. The value of page-by-page autopsy is evident here: the seizure, and hence the objects’ eligibility for restitution, has since been confirmed through compelling biographical research into a former owner, which cannot be presented here due to space limitations.

3. The case of another “insertion” in a book suspected of being looted further demonstrates the value of page-by-page autopsy when it comes to compiling the fullest possible object biography. In a volume belonging to the context of books seized from German prisoner-of-war camps (ULB Bonn, shelfmark Fc 516/3; Fig. 5), an originally inserted fragment was found between pp. 170/171. This was identified as a cutting from the newspaper Le Trait d’Union, no. 90, dated 8 June 1941. Beyond the known information that the book had originally belonged to the camp library of Stalag XII/A – Limburg an der Lahn and was accessioned by Bonn University Library in 1943, a further valuable conclusion can now be drawn: the French-language biographical novel by well-known Parisian author Francis Carco (†1958) (Le roman de François Villon [Le roman des grandes existences 4], Librairie Plon, Les petit-fils de Plon et Nourrit, Paris [1926]) found its way to the camp by 1941, if not before, and was read there by a French-speaking prisoner. The latter evidently used a piece of newspaper as a bookmark. This was from Le Trait d’Union, a free newspaper for French prisoners of war published in Berlin between 1940 and 1945 and contained German propaganda (Laska 2003, p. 237). Why the prisoner was only able to read the book up to p. 170/171, or marked these pages as memorable, must remain a matter of speculation due to the user’s anonymity. Nevertheless, the fragment uncovered by means of page-by-page autopsy opens an impressive window onto everyday life at the Limburg prisoner-of-war camp some 84 years ago.

Had page-by-page examination not been undertaken in these three cases – which account for only a small segment of the approximately 360 suspect provenances recorded at ULB Bonn to date – the volumes concerned would not have been identified as cases meriting suspicion.

Only two common types of provenance features have been discussed here, namely ownership stamps and insertions, but the question of which other types may be found within books can, in condensed form, be answered as follows: any book-typical feature not inherently bound to the outer parts of the codex (binding, endpapers, title page, etc.) may occur at any point within the text block (supralibros, shelf labels, bookplates, etc.).

The T-PRO Thesaurus der Provenienzbegriffe offers a reliable overview of the possibilities to be expected; in conclusion, however, attention may be drawn to some of the most frequently encountered internal features.

Once again, the provenance mark of the stamp must be mentioned, which may have been applied not only to the external parts of the book but also within the text block as a so-called “secret stamp”. This was a security device “to protect library property from theft or to identify it as one’s own in other hands” (Lexikon des gesamten Buchwesens, vol. 3, p. 116 f.). Often smaller than the “main stamp”, such marks were usually applied by institutions (and less frequently by private collectors) at the same place in all their books, for example on p. 33, on the first leaf of the second gathering, or on the reverse of plates or maps. The advantage for provenance researchers is that by their very nature, such features have a higher chance of survival than marks on the outer parts of the book, which are more easily destroyed intentionally or lost accidentally. The disadvantage is that institutions followed differing practices in terms of how, how frequently and where they applied their secret stamps. For example, the main library of the former Nazi Ordensburg Vogelsang always stamped the last page of the main text, while Bonn University Library always stamped p. 33.

Moreover, many institutions used secret stamps but did not do so consistently, applying them at different places in different books. This enhanced their function as security marks, but made them harder to locate today. One such institution which used “unsystematic” stamping was the library of the SPD Bezirksverband “Obere Rheinprovinz” [SPD District Association Upper Rhine Province], which in one book placed its ownership stamp on pp. 3 and 89 (ULB Bonn, shelfmark Rf 148/561) and in another on the half-title and pp. 33 and 97 (ULB Bonn, shelfmark Ll 702/138; Figs. 6 and 1). This also shows that secret stamps were not used “only by major libraries”, as is stated in the Lexikon des gesamten Buchwesens, but were also applied by smaller libraries, such as that of the editorial office of the trade union journal Der Holzarbeiter, organ of the Zentralverband christlicher Holzarbeiter Deutschlands [Central Association of Christian Woodworkers of Germany] (cf. ULB Bonn, shelfmark 50/5163, editorial stamp on front cover, title page and pp. 5, 23, 27 and 99). Other examples that can be cited from the experience gained in the Bonn project include lending libraries (e.g. ULB Bonn, shelfmark Fb 638/18), the library of the former Reichsschule NSDAP in Bernau/Berlin, and the Agency for Intellectual Relief in Germany.

Here too, page-by-page autopsy of suspect books proves indispensable. The positions of such marks are not known in advance or made public for obvious reasons, and in the case of inconsistent stamping no such information may exist at all. Attention must also be drawn to the “human factor” as the instance responsible for provenance features: this element always needs to be taken into consideration in object examinations.

A second major group of provenance features are the so-called insertions, a term that can encompass anything that has been physically placed inside a book – such as letters, photographs, or any object used as a bookmark, as illustrated above in the case of the newspaper cutting. Such objects within objects serve firstly as direct provenance evidence, offering an address on a letter, a date on a photograph or even a portrait of the reader, for example. Secondly, they may also serve as indirect evidence, contributing to the reconstruction of the object’s history (tickets, publishers’ advertisements, borrowing slips, lottery tickets, newspaper clippings). Another example concerns a volume from the context of seizures from Jehovah’s Witnesses (Zionslieder für alle christlichen Zusammenkünfte, Verlag der Internationalen Vereinigung Ernster Bibelforscher, Brooklyn, NY, [et al.], 1923. ULB Bonn, shelfmark Gm 267/435). Accessioned by Bonn University Library on 7 July 1936 after confiscation in the region under the authority of the Kreuznach Landrat [district administrator], it bears on the front pastedown the handwritten ownership note of a “Frau Thomas” [Ms. Thomas]. Several simple slips of paper inserted into this hymn book (pp. 48/49, 86/87 and 110/111) attest to its regular use. Still more strikingly, its continued use in the early weeks and months of persecution of Jehovah’s Witnesses in the Nazi state is evidenced by calendar leaves inserted at various points: leaves for 14/15 May 1932 (pp. 36/37), 16/17 May 1932 (pp. 42/43), 18/19 May 1932 (pp. 28/29) and finally 8 May 1933 (pp. 40/41; Fig. 7).  

“Handwritten notes” can cover ownership entries, dedications and many other elements. While these are often found in the outer parts of a book, they may also appear elsewhere, as illustrated by a volume originating from a French-speaking territory occupied in the Second World War, most likely Belgium (Georges Rency, Le Jardin des Images, Vanderlinden, Brussels 1931). ULB Bonn, shelfmark Fc 506/14). This volume, which most likely found its way to Bonn University Library in 1942/43 via the Reichstauschstelle [Reich Exchange Office] was owned by the schoolchild Raoul Piérart, who not only noted his name and school year in ink on the original printed wrapper title (now damaged by trimming) but also entered pen trials in the form of a stylised signature several times within the book (pp. 45 and 100).

To some extent this latter example touches on the final area to be addressed here, namely the so-called traces of use. These comprise all marks left in a book by its users: annotations, underlinings, markings or other entries that are not necessarily, but often can be, related to the reader’s engagement with the text. Such traces need not in themselves be connected with a context of seizure, but in certain circumstances they can be attributed to such a context, for example by comparing handwriting, ink, or by criteria based on the content of the work. They therefore constitute not merely a useful or illustrative by-product of book autopsy, but in some cases can make a genuine contribution to object research. However, it is true that traces of use are often difficult to attribute with certainty, and that it is usually only possible in limited and qualified ways to distinguish between contemporary and earlier or later additions, some of them made by “ordinary” library users after 1945.

The overall conclusion is thus that in projects centred on printed materials, in order to conduct an exhaustive autopsy of a corpus of books it is essential to undertake an examination literally from every angle, however many individual items may be involved. Admittedly this is always a time-consuming and demanding task, one that calls for perseverance against fatigue. It is also technically delicate, depending on the state of preservation of the volumes. The results can be frustrating, too, when after days of work on supposed cases of suspicion in volumes that may extend to thousands of pages, nothing is ultimately found. Yet persistence is rewarded by those discoveries that would have remained hidden had a more superficial approach been taken: these are a source of motivation to continue searching.

It may be acknowledged that it is legitimate to conduct a superficial autopsy in order to gain an initial overview of the corpus under analysis or to narrow it down by way of exclusion, especially in the case of large-scale collections. However, this legitimate method must not give way to an aim of “making up numbers” quickly by means of a cursory glance. Lessons learned from the examination of Bonn ULB’s historical collections show just how many provenances and how many of their features would not have been identified and would not have been recorded as cases of suspicion without page-by-page autopsy. Beyond the scholarly standards to be upheld, it is ultimately the moral considerations arising from the context of seizure that oblige us to pursue this methodical approach with transparency.

It is equally clear that many institutions are simply not in a position to carry out page-by-page autopsy of large or even medium-sized holdings. This means that the additional time and staff resources that may be needed for object checks in libraries should be factored into project proposals submitted by applicants and duly taken into account by funding bodies when assessing them.

In conclusion, it should be noted that the author is of course aware that the above remarks are not to be taken as a rigid formula to be applied slavishly: every collection has its own history and specific features, just as the volumes it may contain stem from a wide variety of provenances. Even so, it is hoped that these reflections might help ensure that books receive the same attention in terms of autoptic examination as other cultural assets – despite the extra work involved and the sometimes very different financial value attached to them. The challenge is to master the sheer mass of books: after all, to return once more to the Königsberg professor cited at the outset, “a life without books would be torture”!

Dr. Tobias P. Jansen is a research associate working on the Looted Cultural Assets project at Bonn University and State Library.

Literature:

Albrink 2006
Albrink, Veronica, Leitfaden für die Ermittlung von NS-verfolgungsbedingt entzogenem Kulturgut, in: Die Suche nach NS-Raubgut in Bibliotheken. Recherchestand – Probleme – Lösungswege (Schriften der Universitätsbibliothek Marburg 126), Bernd Reifenberg (ed.), Marburg 2006, p. 150-180.

Alker 2017
Alker, Stefan et al., NS-Provenienzforschung und Restitution an Bibliotheken (Praxiswissen), Berlin and Boston 2017.

Laska 2003
Laska, Andreas, Presse et propagande allemandes en France occupée: des Moniteurs officiels (1870-1871), à la Gazette des Ardennes (1914-1918) et à la Pariser Zeitung (1940-1944), Munich 2003.

Leitfaden 2019
Provenance Research Manual to Identify Cultural Property Seized due to Persecution during the National Socialist Era, German Lost Art Foundation et al. (ed.), [no place of publication], 2019.

Lexikon des gesamten Buchwesens
Lexikon des gesamten Buchwesens, Severin Corsten et al. (eds.), second edition, vol. 3, Stuttgart 1991.

Schmoller 2012
Schmoller, Andreas, Einleitung. Erste Fragen, erste Recherchen, erste Ergebnisse, in: Buchraub in Salzburg. Bibliotheks- und NS-Provenienzforschung an der Universitätsbibliothek Salzburg, Ursula Schachl-Raber et al. (eds.), Salzburg and Vienna 2012, pp. 124-133.

Zuschlag 2022
Zuschlag, Christoph, Einführung in die Provenienzforschung. Wie die Herkunft von Kulturgut entschlüsselt wird, Munich 2022

To the project Ermittlung von NS-Raubgut in der Universitäts- und Landesbibliothek Bonn [Identification of Nazi-Looted Books at Bonn University and State Library] (December 2020 to February 2026)