1Nietzschesource describes itself as a ‘web site devoted to the publication of scholarly content on the work and life of Friedrich Nietzsche’. So far, two editions of Nietzsche’s work have been made available on Nietzschesource: the Digitale Kritische Gesamtausgabe Werke und Briefe (eKGWB) and the Digitale Faksimile Gesamtausgabe (DFGA) of the Nietzsche Estate. The eKGWB is based on the latest critical print edition of Nietzsche’s works, the Gesamtausgabe edited by Giorgio Colli and Mazzino Montinari (Colli-Montinari) and follows both the text as well as the categorisation of Nietzsche’s writings established by that edition, which are the following:

• Published works (18 volumes in the eKGWB)
• Private publication (3 volumes)
• Authorised manuscripts (3 documents)
• Posthumous writings (8 documents)
• Posthumous fragments
• Letters (covering 38 years)

2The DFGA is an independent contribution of Nietzschesource, at the moment comprising 9000 digitised pages and aiming at providing high-resolution colour-scans of the entire contents of the Nietzsche estate. The scholarly team behind Nietzschesource consists of a group of ten people headed by Paolo D’Iorio; the website is currently managed by the Association HyperNietzsche, a non-profit organisation hosted at the École normale supérieure in Paris.

3Nietzschesource is not the first attempt to produce a digital edition of Nietzsche’s works and, as we will see, it falls behind some of the achievements of these earlier attempts. In the world of digital editions, Nietzsche has had a relatively long history, starting with a CD-ROM version by Malcolm Brown published by de Gruyter in 1994 (Brown). Since then there has been the HyperNietzsche project (D’Iorio 2001), whose website first appeared online in 2001, and now Nietzschesource. HyperNietzsche and Nietzschesource, both being directed by D’Iorio and managed by the Association HyperNietzsche, are two closely connected projects. Furthermore, navigating to http://www.hypernietzsche.org will redirect you immediately to Nietzschesource, and hence it appears that Nietzschesource is supposed to be the successor of HyperNietzsche and a further development. At the same time, much of the content of HyperNietzsche is still available if one uses a search engine to bypass the start-page. A good entry point is http://www.hypernietzsche.org/surf_page.php?type=scholarly. The relation of HyperNietzsche and Nietzschesource is a very interesting issue from a number of viewpoints, but a detailed comparison of the two projects is beyond the scope of this review. I will, however, refer to HyperNietzsche at some points of my criticism of Nietzschesource and it will emerge that, at least in some aspects, HyperNietzsche has to be considered to be superior to its successor project.

4 To be sure, Nietzschesource is certainly a very good starting-point for the study of the works of Nietzsche. The texts in the eKGWB are presented in an intuitive interface boasting four languages and a lucid, modern design. The screen is always divided into three parts: an area at the top containing the search-function, a navigation-path and the print-button; a collapsible drop-down menu on the left that initially shows the different categories of Nietzsche’s writing and unfolds down to section-level; and the main reading view, appropriately taking up the largest part of the screen. The search-engine allows simple as well as advances queries, which can include rudimentary regular expressions. Typing in a search window will bring up auto-complete suggestions based on a word-list, also showing the number of occurrences of a particular term in the corpus. The advanced search allows different and well-known restrictions-methods: any/all terms, case-sensitivity, and a complex work-selection that can be either set to a certain time-period, a particular work or, in the case of the letters, to a specific recipient or a location from which and/or to which the letter has been sent. In the search-results screen, hits are grouped by work in the left pane and a preview is shown in the central area. Rather unexpectedly, clicking on a particular hit will display the section containing the hit in isolation from its context. A button located in the top area will bring the reader to the normal work-view, but it would serve purposes better to immediately show the hit in its context and guide the eyes of the user by appropriate highlighting of both the section and the search-terms.

5Browsing the contents the eKGWB is straight-forwarded and can hardly be improved. After selecting a work from the left navigation pane, the entire text is loaded into the central reading area while the navigation pane unfolds into a chapter-browser. This is convenient, for it allows the readers to avail themselves of the inbuilt search-function of his browser. Reading longer parts of the text from the screen can prove somewhat exhausting, due to line-spacing, the choice of font and the fact that the length of the lines is determined by the browser window, which, on modern big screen results in very long lines. However, this is a minor point and thanks to the incorporation of numerous corrections on the text of the Gesamtausgabe, the eKGWB can pride itself of being the most up-to-date edition of Nietzsche’s works available. All these features should make Nietzschesource the first stop for students of Nietzsche’s work and a great contribution to Nietzsche scholarschip.

6Nietzschesource developed a reference system that allows referring to works, chapters, aphorisms or fragments with human-readable, fixed and therefore citable internet-addresses. For instance, section 18 of the prelude of the Gay Science, ‘Schmale Seelen’, can be referred to with the URL http://www.nietzschesource.org/#eKGWB/FW-Vorspiel-18. At the moment, to obtain such a reference-URL, the reader has to select the relevant section from the navigation pane and then copy the URL from the browser address bar; it would be more convenient to also have appropriate citation information in every referable section in the main reading area.

7The DFGA offers high-res colour facsimiles of several of Nietzsche’s writings, but a quick view of the current contents makes it evident that some work still has to be done. At the moment, the reader will find facsimiles of some published works, as well as proof sheets, manuscripts for printing, loose sheets, notebooks and notepads, but the digitising of such an enormous corpus evidently takes a long time and hence a lot is still unavailable. Studying the facsimiles is facilitated through an image-viewing app that allows not only to zoom in and out, but also to rotate the image and modify contrast, brightness, hue and saturation, features that certainly will be appreciated by those trying to decipher Nietzsche’s handwriting, which, especially when it comes to his notebooks, can be a tricky business.

8 In general, a reader unfamiliar with digital editions will feel at home right away when working with Nietzschesource. Hardly any use has been made of the possibilities digital editions offer in favour of classical forms to present editorial action, well known from printed editions: added letters are angle-bracketed and incorporated corrections highlighted with a different background colour. Only the additional information regarding the type of correction offered in a popup window prove to be exceptions to this observation.

9 So much for the upside (if, indeed, the last point is considered a plus). The downside looks a bit bleaker. First of all, the most disappointing thing about Nietzschesource is the fact that eKGWB and DFGA are, at present, completely unconnected. To compare the edition with the respective facsimile it is necessary to open two browser windows and find the right facsimile manually (if it is available at all, of course). The idea of layers of texts, which enable the reader to go from an edited text to a diplomatic transcription and from there to the underlying facsimile, is absent from the current version of Nietzschesource. In this respect one cannot help to acknowledge the loss of some features which were already realized in HyperNietzsche: there, several layers were available in a synoptic view (cf. (Saller 2003)), and the manuscript-editions even offered a rhizome-view providing a representation of the evolution of certain works.

10 Secondly, the mark-up of existing textual phenomena is extremely limited. Apart from the above-mentioned addition of missing letters, it seems that the only one other phenomenon that has been marked is letter-spacing. Of course, editors have to make a choice regarding the phenomena they want to record, especially if confronted with a corpus as large as Nietzsche’s. However, some phenomena, for instance strike-through, are so fundamental for a critical edition that the omission to record these has to count as a considerable minus; again, HyperNietzsche did much better in this respect.

11I would like to add a note on the phenomenon of letter-spacing. In the eKGWB, letter-spacing is frequently used both in texts belonging to the category of printed works, as well as in handwritten material such as the letters. If we compare passages of the printed works featuring this phenomenon with their respective manuscripts, we find that Nietzsche, writing by hand, did not space the letters, but underlined them. Thus, his underlining was meant to tell the printer to emphasize the respective words and probably can also be interpreted as an expression of his wish that these words should be letter-spaced assuming, as is plausible, that Nietzsche knew that the printers would make use of this kind of emphasis. The eKGWB edition of the letters also feature letter-spacing and we can assume (since there are no facsimiles of the letters yet, I was unable to confirm this assumption) that the manuscripts of the letters also underline the words that the eKGWB represents as letter-spaced.

12Hence, one could say that the eKGWB is consistent in representing underlining with letter-spacing. However, there is a difference. For while in the case of manuscripts that were meant to be published Nietzsche used underlining as an indication for the printers to emphasize the respective words and in all probability knew that the printers would not use underlining but letter-spacing to realize this indication, this is not the case for the letters. This brings up the question how an editor should deal with textual phenomena such as underlining. The editors of Nietzschesoure have chosen to represent the semantics of the phenomenon: underlining in the case of the published works as well as in the case of the letters was (probably correctly) interpreted as emphasis and this semantic value is, in the edition, represented with letter-spacing. Alternatively, an editor can choose to preserve the textual phenomenon and mark the respective words as underlined. They can then, in a further step, add an interpretation of this phenomenon and decide that underlining has the semantic value of emphasis. The latter method has the advantage of greater flexibility: for the reading texts of the published works it might make sense to use letter-spacing for words marked as underlined in order to stay close to the printed works, which, since Nietzsche saw and approved of them, count as autographs. But for the letters, it might be better to represent underlined words by underlining them. In both cases, given appropriate documentation, the process is transparent to reader and hence open to criticism. This consideration is not meant as criticism of Nietzschesource, for such a criticism would be nit-picking. Also, it is possible that the editors indeed went half the way of the second method: according to their introduction the texts are encoded using TEI and the normal way to encode underlining in TEI would be <emph rend=”underlined”>; I cannot confirm that this mark-up has been used because the TEI files are not publically available, but even if this was the case, Nietzschesource makes no use of the separation between textual phenomenon and semantic interpretation by just printing letter-spaced without providing the reader with information regarding the underlying textual phenomenon.

13But at the end of the day, if the eKGWB is consistent in always representing underlining with letter-spacing, nothing is lost to the reader and that is all that counts. But I think the consideration shows an important (possible and desirable) feature of digital scholarly editions, namely to preserve and present textual phenomena and keep them and the editor’s interpretation of them apart (a striking example is the use of underlining in medieval manuscripts: words were underlined for a number of reasons including deletion, indication of proper names and quotes, just to mention a few. Clearly, keeping the textual phenomenon and its interpretation apart is instrumental when editing such a text.)

14This brings us, thirdly, to the important question of documentation. As the case of the letter-spacing shows, the readers are forced to speculate about the textual evidence (or, if available, they have to check the facsimiles themselves), even in case of existing mark-up because no explanation regarding these issues is provided within Nietzschesource (another example is the occurrence of ‘[+ + +]’ in the letters). The same is true for information regarding the technical background of Nietzschesource. We learn from the four-paragraph introduction to the eKGWB that TEI was used to encode the data, but nothing else is revealed (the XML files themselves are not accessible). All in all, ‘minimalistic’ would certainly be a euphemism as far Nietzschesource’s technical and editorial documentation is concerned. In fact, it is a basic editorial duty to make transparent to the reader all symbols used to indicate editorial action, but even after browsing Nietzschesource for hours I could not obtain these information.

15 Furthermore, it is regrettable that one of the greatest possibilities of digital editions has been neglected, which might be called deep indexing or information enrichment, i.e. the adding of information relevant to the text, for example the identification of named entities and the connection to entities and identifiers from authority files. Especially in the case of letters and diaries this technique can be put to great use, since the readers are usually confronted with a plethora of persons, and about which they are eager to learn more. To accomplish such a task, an increasing number of databases is already available to date, such as GND/PND or VIAF or, for editions mostly concerned with German speaking persons, the Biographie-Portal, which merges ADB, NDB and a few other registers (see http://www.biographie-portal.eu/). Even Wikipedia provides a convenient way of linking to their data with the use of unique person-IDs obtainable, in the case of the German Wikipedia, which would probably be the most important for Nietzschesource, through a web interface: http://toolserver.org/~apper/pd/. Thus, linking a name with the appropriate ID, and with a little scripting, the student of Nietzsche’s letters could immediately see who Erwin Rohde or Friedrich Ritschl was. A similar point can be made for the occurrence of references to certain writings, especially those of Nietzsche himself: where in the letters is he referring to his works (completed or in progress)? This question could certainly be of interest to the historian of philosophy who tries to trace the genesis of a particular text. Hence, two important functionalities become available as soon as textual entities like persons and works are appropriately tagged: the readers can conveniently obtain contextual information that helps them to understand the text; and it is possible to automatically create indices that can be used to systematically study the text with respect to an indexed item.

16Since it is evident that this sort of additional information can be extremely helpful, it should, if at all possible, always be provided in a digital edition. But there are even more basic kinds of metadata, and it seems that Nietzschesource is not providing them, either. A striking example is the Idyllen aus Messina. Opening this work from the list of published works in the eKGWB will give you only the title itself and then immediately the first piece, ‘Prinz Vogelfrei’. But we are not told – at least in the eKGWB, while the DFGA, remarkably, contains this data – that this work was published in the fifth issue of Ernst Schmeitzner’s Internationale Monatsschrift. Zeitschrift für allgemeine und nationale Kultur und deren Litteratur or any other relevant information regarding its publication. Nietzsche scholarship has put in an effort to provide these information (Schaberg 1995) and an extract of such findings or at least a reference to it would be of great value to any reader. Again, had work titles been indexed it would have been possible to provide the readers with links to documents relevant for the publication of the Idyllen, as e.g. Nietzsche’s letter to the editor of the Monatsschrift, Schmeitzner (cf. http://www.nietzschesource.org/#eKGWB/BVN-1882,227). One could object that the case of the Idyllen is a rather exceptional example and that basic information like the place of publication is available for all or most of the other works. However, the real problem is that they are not available as formal metadata, but only as text on the cover-pages of the individual works. This is another example of missing mark-up and results in the impossibility, just to give one example, of automatically creating registers that include the different publishing houses and editors Nietzsche published his works with.

17To sum up, Nietzschesource will certainly be welcomed by scholars working on Nietzsche or related subjects: it is a free of charge, publically available edition of all of his works offering the best text available to date, thanks to the incorporation of the errata; a large number of facsimiles can be studied and the search engine can be put to great use. Apart from that, and in view of the criticism advanced above, I conclude that Nietschesource is rather a digitised critical edition than a true digital scholarly edition (Sahle 1:148-155). There are, despite D’Iorio’s claim that Nietzschesourve ‘is not a “digital photocopy” of the Gesamtausgabe’ (D’Iorio 2010), only very few features that could not have been accomplished by a printed edition, and some features that a good printed edition surely would have provided, are missing. Compared to other digital editions projects, like for instance the Diaries of Robert (Graves-Petter-Roberts 2003), the Correspondence of Carl Maria von Weber (Allroggen et al. 2013), the Correspondence of Alfred Escher (Jung 2012), or the Diary of William Godwin (Myers-O’Shaughnessy-Philip 2010). Nietzschesource has not availed itself of the possibilities that would allow it to meet expectations generally held for digital scholarly editions nowadays. Without access to the XML files or a sufficiently detailed technical documentation, it is unclear to me that it would be possible to implement the following suggestions, but I feel these would immeasurably enhance the existing project: first and most importantly, establishing links between eKGWB and DFGA; secondly, incorporating a layer-model including a diplomatic view; and thirdly, enriching the texts with metadata. Moreover, making the XML files publically available would allow specialists to systematically analyse the corpus as well as make sure that the work the Nietzschesource-team has put into the creation of a TEI-version of Nietzsche’s texts can function as the basis for a future development.

18Given this evaluation of Nietzschesource, the question of its relation to HyperNietzsche is raised once again, especially since D’Iorio has been involved in HyperNietzsche as well. Most of my criticism would have been inapplicable to HyperNietzsche, and we have to wonder why, even if it had proved impossible to carry through the extremely ambitious task HyperNietzsche set for itself, Nietzschesource is significantly below its precursor in terms of methodological aims and design. One could suspect that this is due to the integration of Nietzschesource into the community of other *-source projects, like Wittgensteinsource, or ModernPhilosophiesource (see http://www.discovery-project.eu/philosource.html), foisting the Nietzsche-project into a generic framework that cannot accommodate for many features, the lack of which has been criticised above. However, one would expect that something as basic as the mark-up of persons and works should be available in any framework designed to host a great variety of philosophical writers. At any rate, it seems that the different *-source projects really are pretty independent of each other and that Wittgensteinsource, for instance, offers a very neat configurable synoptic viewer including diplomatic transcriptions, just to name one feature (http://tinyurl.com/cbd8nga).

19To conclude, Nietzschesource is a great resource for students and scholars of Nietzsche. It offers the best text available to date and is free of cost. It already provides many facsimiles that will allow researchers to challenge the existing editions and hopefully this collection will grow over time. As a digital edition, though, it is, at least in its current state, a disappointment, for the reasons outlined above. The author of this review hopes that the future will allow the team around Nietzschesource to pick up some of the ambitious and fascinating ideas that made HyperNietzsche such an interesting project.