Berlin-Brandenburgische Akademie der Wissenschaften (ed.), Kants gesammelte Schriften. Neuedition der Abtheilung I (Werke). Berlin/Boston: De Gruyter, 2023 -, 9 volumes

This is not so much a review as a report, with relevant background, on the new Kant edition of the Berlin-Brandenburg Academy of Sciences and Humanities, of which the first volume has now been published (Kant 2023), and its potential for innovation. After reconstructing the publication history of Kant’s complete editions, the report considers the impact of Kant’s Academy edition since 1900, which has been marked by important and increasing criticism regarding its accountability. It then delves into the editorial guidelines of Kant’s new Academy edition. Finally, given that the Academy edition has been extensively used for about five quarter-centuries, and thus it is not unwarranted to think of readers of the new Academy edition well into the second half of the twenty-first century, the report offers a look forward into the use of digital datasets for approaching Kant.

1. Pre-history

An historical-critical edition – and Kant’s Akademie-Ausgabe (in the following AA) is most clearly one – requires a careful consideration of all editions that preceded, whose variants the editors of the AA-Neuedition have accounted for in the Apparat at the foot of each page. ¹ While Kant was still alive, several editions appeared of his kleinere Schriften. The most authoritative among these were the four volumes of the Vermischte Schriften edited by Johann Heinrich Tieftrunk (1760–1837), the first three of which appeared in 1799 and the fourth in 1807 (Kant 1799–1807; see also, as edited by Friedrich Theodor Rink (1760–1841), Kant 1800). In 1838, thirty-four years after Kant’s death, not one but two complete editions of his printed writings were offered to scholars, both organized according to the individual parts of philosophy, with the texts in chronological order in each volume: almost simultaneously, there appeared the edition of Karl Rosenkranz (1805–1879) and Friedrich Wilhelm Schubert (1799–1868) in twelve volumes (Kant 1838–1842) and the edition of Gustav Hartenstein (1808–1890) in ten volumes (Kant 1838–1839), the latter of whom subsequently published a second edition in eight volumes, no longer arranged according to content, but strictly chronologically (Kant 1867–1868).

In 1868, Julius Hermann von Kirchmann (1802–1884) founded the Philosophische Bibliothek. His editions of Kant’s writings (Kant 1868–1878) closely follow Hartenstein’s first edition. Chronology also prevails within each volume. Karl Kehrbach (1846–1905) brought out the first selected edition (Kant 1877–1884), which included only the most important of Kant’s writings in seven volumes in the series Universal-Bibliothek. ² The Cassirer-Werkeausgabe (published by Bruno Cassirer [1872–1941]) was subsequently published in eleven volumes (Kant 1912–1922) in strict chronological order. It was edited by Hermann Cohen (1842–1918); Otto Schöndörffer (1860–1926); Benzion Kellermann (1869–1923); Albert Görland (1869–1942); Otto Buek (1873–1966); Ernst Cassirer (1874–1945), and Artur Buchenau (1879–1946). In 1922–1923, it was published in a second, largely unchanged edition. This edition also claims to be complete; however, some of Kant’s larger writings are missing. Finally, the last twentieth-century Kant edition of Wilhelm Weischedel (1905–1975) was intended as a study edition. It first appeared in 1956 as a six-volume thin-print edition, and since 1968 has been available as a ten-volume paperback and a twelve-volume study edition (Kant 1956, 1968a, 1968b). However, neither is this edition complete, as it is missing some of Kant’s both smaller and larger writings (Rosie and Schepelmann [hereinafter R/S] 2023: xxvi–xxvii).

2. History

The idea of establishing the AA dates back to 1894, when at the request of Eduard Zeller (1814–1908); Wilhelm Dilthey (1833–1911), and Max Heinze (1835–1909), the Königlich-Preußische Akademie der Wissenschaften zu Berlin (hereafter referred to with the current acronym BBAW, for Berlin-Brandenburgische Akademie der Wissenschaften) considered appointing a scholarly commission to plan and supervise the complete edition of Kant’s Druckschriften, Briefwechsel, Nachlaß, and Vorlesungen (Kant 1900–2020). ³ The proposal was approved in 1896, and work started immediately (R/S 2023: xxviii). If it is true that Kant came after the monumental Academy edition of Aristotle, it is also true the AA was intended as a model for Wilhelm von Humboldt’s Gesammelte Schriften and the Leibniz-Edition of the Sämtliche Schriften und Briefen. ⁴

1. The AA is divided into four sections:

2. I. Works (vols. 1–9)

3. II. Correspondence (vols. 10–13)

4. III. Manuscripts (vols. 14–23)

5. IV. Lecture transcripts (vols. 24–29)

Section I (Werke, in the sense of printed works, i.e., Druckschriften) was entrusted to individual specialists, on the whole to fifteen editors: Max Heinze (1835–1909); Kurd Lasswitz (1848–1910); Paul Natorp (1845–1924); Wilhelm Windelband (1848–1915); Benno Erdmann (1851–1921); Alois Höfler (1853–1922); Johannes Rahts (1856–1938); Karl Vorländer (1860–1928); Oswald Külpe (1862–1915); Erich Adickes (1866–1928); Heinrich Maier (1867–1933); Georg Wobbermin (1869–1943); Paul Gedan (1871–1932); Paul Menzer (1873–1960), and Max Frischeisen-Köhler (1878–1923).

Volumes 1–8 (1902–1912 in the first edition) were replaced by the second edition, known as the Neudruck (Berlin: Reimer, De Gruyter, 1910–1920). Only the text of the Metaphysik der Sitten (vol. 6) underwent changes. The last volume of Section I (vol. 9), which contains the Logik (ed. 1800 by Gottlob Banjamin Jäsche [1762–1832]), the Physische Geographie (ed. 1802 by Friedrich Theodor Rink), and the Pädagogik (likewise ed. Rink, 1802), was first published in 1923. From 1968 onwards, the volumes of Section I were offered in a paperback edition, the Textausgabe (Kant 1968c; see R/S 2023: xxvi–xxix). The Textausgabe is a reprint of part of the first and part of the second edition of Section I. However, the texts in these paperback editions were by no means unaltered photomechanical reprints, as stated on the frontispiece – some deviations demonstrably took place (R/S 2023: xxviii). After 1977, in addition to the texts in vols. 1–9, the Textausgabe included two volumes of Anmerkungen, with the Editorische Berichte and Lesarten. However, it is necessary to note a point of misleading information, synecdochally due to a pars pro toto fallacy: the 1968 reprint gives 1902 as the beginning of the first edition, which is not true, since the first edition began in 1900 with volume 10, the first of the Briefwechsel. To be sure, volume 1, the first of the Druckschriften, appeared in 1902. In any case, the information provided has proved seriously misleading (Stark 2000b: 3).

Three volumes (10–12) of Section II (Briefwechsel) were edited by Rudolf Reicke (1825–1905). As a considerable number of undocumented letters were discovered during the publication of the first edition of the correspondence (1900–1905), the second edition of volumes 10–12 was published in 1922 at the same time as the first publication of their Sacherläuterungen (vol. 13) edited by Paul Menzer and Rose Burger (biographical data unknown), with the cooperation of Johannes Reicke (1861–1931) and Athur Warda (1871–1929). The publication of the volumes of Section III (Nachlaß), commissioned to Adickes, began in 1911. Its completion was delayed by the death of Adickes in 1928. Due to the loss of many handwritten originals during World War II, much of the content of Section III had to remain unedited. Volume 23, the final volume of Section III, was published in 1955. However, it was still only a formal end.

The overall concept of Section IV (Vorlesungen) was largely undefined at the beginning. The basis of the edition was mostly handwritten student transcripts, the reliability of which first had to be checked. For various reasons, editorial work was abandoned in 1920 without a single volume being published. In the mid-1950s, the then Deutsche Akademie der Wissenschaften zu Berlin revived the original concept and commissioned Gerhard Lehmann (1900–1987) to edit Section IV. In 1961, a partial volume with the edition of the Vorlesungen über Philosophische Enzyklopädie and some of the Vorlesungen über Logik was published by Lehmann outside the AA with the Akademie Verlag. After the construction of the Berlin Wall in 1962–1963, the Niedersächsische Akademie der Wissenschaften zu Göttingen agreed to continue work on Section IV in consultation with the Deutsche Akademie der Wissenschaften zu Berlin. Volume 24 (Vorlesungen über Logik) was published in 1966 as the first volume of Section IV. In 1987, the Kant Commission of the Göttingen Academy decided to relocate the Kant unit from Berlin to the Philipps Universität in Marburg. Volume 25 (Vorlesungen über Anthropologie) was completed and published by Reinhard Brandt (b. 1937) and Werner Stark (b. 1953), in time for the 20th World Congress of Philosophy in Boston (1998).

Since 2002, the complete edition of Kant’s writings has been an academy project of the BBAW. The transfer of responsibility to the BBAW also fulfilled a wish of the Niedersächsische Akademie der Wissenschaften zu Göttingen. Since 2002, the Kant Arbeitsstelle Potsdam has focused on three projects: new editions of the Kritik der reinen Vernunft, Kritik der practischen Vernunft, and Kritik der Urtheilskraft (Section I); the new transcription, electronic publication, and historical-critical new edition of Kant’s last work, known under the title Opus postumum (Section III); and the edition of Kant’s Vorlesungen über physische Geographie (Section IV). The first of two partial volumes of the Geographie-Nachschriften has been available since 2009 (volume 26.1); the second partial volume (volume 26.2) was published in the fall of 2020.

In addition, Kant’s handwritten annotations in his own copy of Alexander Gottlieb Baumgarten’s Metaphysica (3rd ed., Halle: Renger, 1750) were discovered in 2018 and edited for the first time outside the AA (Kant 2019). They are planned to be part of the edition of volume 23, the final volume of the Nachlaß Section III, which will also contain the manuscript holdings that are not of the Opus postumum and the Lose Blätter that have been newly discovered since 1955 and published outside the AA. Finally, the edition of all Mitschriften of Kant’s lectures taken by Johann Gottfried Herder will appear as a separate endeavour.

3. Impact

In the aftermath of World War II, the impact of the AA was assessed by Paul Menzer (1873–1960), Heinz Heimsoeth (1886–1975), and Gerhard Lehmann. ⁵ In 1957, Heimsoeth claimed the mission was achieved:

With the publication of the 23rd volume (Kant’s manuscript Nachlaß) Volume 10. Vorarbeiten und Nachträge, Berlin 1955 (W. de Gruyter & Co.), the enormous work is now truly complete in the three sections of the printed works, the correspondence, and the manuscripts – after more than sixty years of incessant planning, searching, collecting and organizing efforts. (Heimsoeth 1957: 337)

However, the celebrations were premature, as anyone might understand when one considers that Wilhelm Weischedel began his new edition precisely in those years with the aim of preserving the ‘sound and rhythm of Kant’s language’ (Klang und Rhythmus der Sprache Kants) and by laying out again ‘the original printings …, except for the few that could no longer be obtained’ (Weischedel, in Kant 1956, 6: 819). Weischedel was very clear:

Like almost all previous complete editions, it [the Weischedel edition] attempts to find a viable middle way between the documentary reproduction of the texts on the one hand and their unrestricted modernization on the other; while it refrains from all fundamental interventions with regard to Kant’s language (phonetic order, inflection, word formation, etc.), it tries, with a few exceptions to be identified below, to adapt his orthography to the current spelling as it is fixed by the Duden. Regarding punctuation, it endeavours to carefully modernize the usage found in the originals themselves. (Weischedel, in Kant 1956, 6: 819–20)

Weischedel felt he had a good cause to pursue, because when the AA was envisaged in 1894 its main goal was to provide reliable and complete texts for scholars and students. In compliance, the Kant-Kommission asked the editors to eliminate most of the variant spelling and syntactic patterns. Since Kant’s spelling habits – so argued the editors of the first volume of the Druckschriften, which appeared in 1902 – are neither systematic nor consistent, the Kant-Kommission thought it appropriate not to disturb most readers with obsolete forms (Kant 1900–2020, 1: 513). Therefore, Kant’s works from 1747 onward were rewritten using the spelling and punctuation of the Kritik der reinen Vernunft, with the result that Kant’s polygraphy was lost. Weischedel touched exactly upon this. Generally, he meant that ‘the linguistic changes made in most editions are not infrequently interpretations and should therefore be avoided in a conscientious presentation of the Kantian text’ (Weischedel, in Kant 1956, 6: 820). ⁶ Specifically, he denounced what had happened in the AA:

There is something absurd about the fact that, for example, the Thoughts on the True Estimation of Living Forces (1747), which Kant himself did not have published a second time, appear in the Academy edition in the Kantian style of the 1790s. (Weischedel, in Kant 1956, 6: 820) ⁷

Widersinnig, absurd, for ‘by refraining from such changes, the linguistic differences between the works belonging to the various creative periods of Kant are largely preserved’ (Weischedel, in Kant 1956, 6: 820). Concerning punctuation, while the editors of the AA did not hesitate to revise it by referring to the practice of Setzerschlendrian, the sloppiness of the typesetter, Weischedel recalls Kant’s habit ‘of inserting the punctuation marks only after completion of the work copy (and often only in the fair copy)’ and notes that ‘in any case, the use of punctuation is much more due to Kant than the editors of the AA have assumed’ (Weischedel, in Kant 1956, 6: 821). Contrarily to Weischedel, the editors of the AA had argued in justification of their approach that numerous peculiarities of the original printings were not due to Kant himself, but to the ‘arbitrariness and fickleness of the individual typesetter’ (Kant 1900–2020, 1: 512), i.e., the Setzerschlendrian that they assumed took place in the print shop (Kant 1900–2020, 1: 513). Lehmann defended the achievements of the AA:

The principle of Weischedel’s edition was originally that of absurdity. In the first published volume (W II, 1956), the Kritik der reinen Vernunft, he [Weischedel] succeeded in undoing all the improvements, conjectures and corrections that had been worked on for a century and a half. (Lehmann 1967: 591; cf. Tuschling 1968)

The Stolperstein, the stumbling block, was the new edition of the first introduction to the Kritik der Urtheilskraft, which had been first edited by Jakob Sigismund Beck (1761–1840) and was a rare case of an available manuscript of Kant’s that was transcribed by Johann Gottfried Karl Christian Kiesewetter (1766–1819) and annotated by Kant. Against the editorial practices of Lehmann and Burkhard Tuschling (1937–2012) went Norbert Hinske (b. 1931) (Hinske 1968, 1990, 1994). And twenty years later, it was Werner Stark who raised definitive doubts about the accountability of the AA (Stark 1988). In 1998, a conference was held at the University of Marburg on the state of the AA and its horizons (Sturm 1999; Stark 2000a). Stark was particularly careful to document that the entire project rested on the shoulders of individual, barely paid editors such as Adickes, Menzer, and Lehmann (Stark 2000b: 2–3).

The question remains: how is it that the AA, despite its flaws, enjoys the greatest popularity? Three reasons come to mind. First, completeness: the AA remains the most complete of all Kant editions. Second, the spread of the affordable eleven-volume box set of the reprint (Kant 1968c). Third, the habit that has developed since the seventies of the last century of citing Kant in the same way as one cites Aristotle, namely by citing volume, page, and line of the editio princeps.

4. New edition

Thanks to the initiative of Volker Gerhardt, the Deutsche Forschungsgemeinschaft has provided funding for the endeavour of the Neuedition of Section I. A total of seventy-four texts are being reedited according to the current state of edition philology and research, under the responsibility of twenty-two external editors, national and international Kant scholars. In preparation for Kant’s tercentenary jubilee (Bundesinstitut für Kultur 2016; Gerhardt et al. 2022) and on time for the 25th World Congress of Philosophy in Rome (2024), the first volume has appeared with De Gruyter, edited by the BBAW (Kant 2023). The editors of the individual Kantian texts from 1747 to 1756 are Antonio Lamarra (b. 1952), Helmut Pulte (b. 1956), Riccardo Pozzo (b. 1959), Hansmichael Hohenegger (b. 1960), Thomas Sturm (b. 1967), Konstantin Pollok (b. 1969), and Fabian Burt (b. 1990), under the coordination of Maja Schepelmann (b. 1971) and Martin Rosie (b. 1985).

We owe the latter a careful coordination of the editorial principles and guidelines (R/S 2023: xvi–xxix). Generally, in the AA-Neuedition the principle holds that the edited texts are specially created versions of the original text witnesses that are as close as possible to the original. An edited text of course always requires editing and, if necessary, intervention by the editors (R/S 2023: xvi).

Most importantly, the AA-Neuedition presents the edited text based on either the individual edition or, failing that, a specific original edition determined by text-critical analysis (e.g., the De igne). The corresponding Editorische Berichte account for all editorial decisions together with all available information on the publication history of each text and its translations, while the Sacherläuterungen provide access to the index of names, the references to Kant’s literal quotations of external authors, and the cross-references within Kant’s writings (e.g., from the Werke, i.e., the Druckschriften, to the Nachlaß and the Vorlesungsnachschriften) (R/S 2023: xxv–xxvi).

Other textual witnesses of the same manuscript, in a few cases also surviving manuscripts, are listed in the form of relevant deviations in the Apparat at the bottom of each page (which corresponds to the Lesarten of the AA, which were included in the Editorische Berichte). More detailed information is available at https://kant-digital.bbaw.de, e.g., the complete lists of variants proposed in all previous editions (R/S 2023: xvi). So far, in one case, the work of the editors has brought about the discovery of an unknown manuscript (Hohenegger et al. 2022). ⁸

Due to World War II, it still happens that the textual witness is lost or otherwise unavailable. In Volume One, this is the case for Kant’s De igne, for which the textual form was reconstructed based on information from the older editions that still had the textual witness available (Rosenkranz and Schubert (Kant 1838–1842); Hartenstein (Kant 1838–1839), and Lasswitz in AA). The Editorischer Bericht of De igne provides detailed information about the reconstruction of the text on a basis that is very close to Penelope’s celebrated canvas.

Except for the Latin texts and the contributions to the Teutscher Merkur, Kant’s writings were originally printed in Fraktur. In the AA-Neuedition, all of Kant’s German-language writings and the translations of the Latin writings have been set in Stempel Garamond, which replaces Fraktur, while the Latin writings as well as all single Latin or French words in his German-language writings are set in Segoe UI. Lastly, an important innovation of the AA-Neuedition is that the editors of Kant’s Latin writings have revised and added the translations into German that are as close as possible to Kant’s Wortschatz (R/S 2023: xxiii). The four Latin dissertations – De igne, Nova dilucidatio, Monadologia, and De forma – are each accompanied by a German translation in a synoptic text presentation. Three of these are based on the late eighteenth-century translations printed in the Tieftrunk edition (1799 and 1807 of Kant 1799–1807), while one, that of De igne, is based on Otto Buek’s early twentieth-century translation (Kant 1907: 1–215).

The guiding principle of the centrality of the original text results in the greatest restraint regarding possible interventions in all cases in which one could intervene in the Kantian text or in which interventions have been made in the history of the edition. Conspicuous features and peculiarities of the original textual witnesses are not assigned the status of errors in the AA-Neuedition. This means that the characteristic of polygraphy (e.g., DEus, Deus, deus: different spellings of the same word), even within the same script or the same chapter, on the same page or in the same sentence, is not regarded as an error to be corrected, but as a characteristic of eighteenth-century German texts. Peculiarities concerning either the inflection of a word or the syntax are not corrected, especially if it can be recognized or assumed that they are due to dialect or colloquial language (of the persons involved in the sentence). Instead of an intervention, reading suggestions are made in the Apparat where necessary for understanding. In the case of different spellings of fractions, the rendering is standardized (R/S 2023: xvii).

The editors have intervened only where the original text is demonstrably incorrect in terms of structure, context, or the linguistic conventions of the time and author. Emendations are implicit improvements and are not recorded in the Apparat. They are made only in unproblematic cases of clear printing errors, such as incomplete, slipped, reversed, twisted, or illegible letters and characters. The passages classified as errors are listed in a separate list of errors, which is published in the editorial report or online, depending on its scope. Incidents categorized as errors in this sense do not include missing punctuation and all other cases in which more than one type of intervention is conceivable (R/S 2023: xviii).

Interventions that require justification beyond a reference to linguistic or written accuracy are conjectures. These are documented by citing the original wording and, where appropriate, its position in the Apparat. They may either follow another original or later edition or represent genuine editorial intervention in the context of the new edition. Reasons for a conjecture may be given in the Editorischer Bericht if necessary; they are not given in the Apparat (R/S 2023: xvi).

The Apparat lists the preceding editions and, where applicable, other textual witnesses in all cases in which the respective deviation (Varianz) affects the meaning of the text or influences the syntax, grammar, expression, or rhythm of a sentence (R/S 2023: xviii). Conjectures of the editors of the AA-Neuedition are recorded, as are the conjectures of the editorial history (but always only the historical first of identical ones). Linguistic modernizations of later editions in terms of expression, spelling, or punctuation are generally not considered. The historical context of the interventions into a Kantian text is documented in the form of conjecture tables and made accessible online at https://kant-digital.bbaw.de. These tables also contain additional information on various Kant editions and their peculiarities or errors as well as cross-edition correlations (R/S 2023: xvi).

When dealing with Kant’s original texts, it is important to be prepared for some linguistic peculiarities of Kant and his time. For example, Kant’s writings contain many elliptical sentences in which parts of speech are omitted for the sake of brevity or elegance. Hence, the editors have preserved Kant’s sound and rhythm with respect to:

Construction. In early writings and in letters, the ellipsis of the personal pronoun (in the first-person singular or plural) is also quite common. References are often constructed according to meaning (σύνϵσις synesis, constructio ad sensum), for example by forming a reference word that refers to a collective singular, for example ‘mankind’, in the plural, thus indirectly including ‘the people’ (R/S 2023: xix).

Grammar. Sometimes a sentence is started differently grammatically or regarding the word field than it is ended, for example sometimes a certain verb is preferred at the beginning, which is then replaced in the continuation of the sentence or at its end by another verb with the same meaning. This is an expression of the proximity to the spoken language, proximity in which many texts of the time, including Kant’s, are still found (R/S 2023: xix).

Spelling. At the time of Kant’s writings, spelling was not standardized as it is today. Kant’s manuscripts and early editions have – as mentioned above – many variant spellings. The later canonization of German spelling and grammar led to a regimentation of writing that moved away from this closeness to language and liveliness (R/S 2023: xix).

The editors have respected these peculiarities, even if they extend to anacolutha, e.g., in the by no means rare cases of nominativus pendens. In the text of De igne, there is a case of this kind. Such peculiarities in the formulation have caused difficulties for some editors (e.g., Hartenstein, in Kant 1838–1839: 8: xviiif.) and have sometimes been wrongly corrected, e.g., Lasswitz changed the nominative plural of materiae durae, which was retained in the original and in all later editions, to materias duras (AA 1: 375). However, in the sentence below we have a nominativus pendens and not an oversight, as the second comparative word, aqua, also occurs in the nominative case:

Quae in corporibus duris compressiones vulgo vocantur, dilatationis verius s. extensionis nomine nuncupandae sunt; quippe materiae durae multo minus, quam aqua, in arctiora spatia vi comprimente adigi posse, per se liquet. [Was man bei harten Körpern gemeinhin eine Zusammendrückung nennt, sollte eher eine Ausweitung oder Ausdehnung genannt werden, denn es leuchtet von selbst ein, daß harte Stoffe viel weniger wie das Wasser durch Druck auf einen engern Raum zusammengepreßt werden können.] (Kant 2023: 338–9)

Furthermore, in Kant’s texts, Latin words or loan words from Latin or Greek are not written with the v and u according to today’s usual rule based on their position in the word (v in the initial sound, u in the internal sound), but according to the rule that was widespread at the time, namely that both letters may appear anywhere, but are to be read consonantly before vowels and vocally before consonants (R/S 2023: xx).

Finally, Kant’s punctuation often does not correspond to the usual way of placing caesuras in syntactic contexts today. The proximity to the spoken word is evident: in the eighteenth century, commas were sometimes used like pauses for breath, while semicolons were used like intensified pauses or paragraphs; colons indicate a (multi-step) derivation; question marks are sometimes written in lowercase and semicolons in upper case. All these special features have been retained in the new edition. Punctuation has only been added or deleted if a veritable typographical error is to be assumed, or if it seems essential to improve the readability of complicated sentence structures (R/S 2023: xx).

5. A look forward

Given that the AA has been in widespread use for about five quarter-centuries, it is not unreasonable to think of readers of the AA-Neuedition well into the second half of the twenty-first century (see Pozzo 2020). The question is, for example: how will a graduate student read Kant in 2070? He/she will do close reading, of course, but also distant reading, which means approaching a research question by looking at digital datasets first qualitatively, then quantitatively, and finally again qualitatively (Moretti 2013).

An edition is a collection of records. Words are the matter. As noted above, the texts of the AA-Neuedition have been tagged very precisely with a markup that captures the formatting of the individual pericopes, their articulation in levels, foreign words, proper names, and cross-references to Kantian and non-Kantian texts (R/S 2023: xxix). It is to be assumed these texts are bound to be made available in due time for artificial intelligence interaction on a dedicated platform.

Kantian Wortschatz has been the subject of considerable research (Schmid 2005 [1795]; Mellin 1970–1971 [1797]; Eisler 1994 [1926]; Willaschek et al. 2015). During Kant’s lifetime, it was already clear that understanding and evaluating Kant’s writings required a complete mastery of Kantian terminology and a precise explanation of words and arguments (cf. Vaihinger 1970 [1881]). Annotating Kant has been undertaken with increasing regularity over more than 50 years alongside the progress of computational linguistics. The start was given by the Allgemeiner Kantindex, which gives Kant’s words in non-inflected form (Martin 1967; Roser et al. 1992). A leap forward was achieved by Tullio Gregory and Norbert Hinske, respectively, with the Lessico Intellettuale Europeo (which since its inception used a markup language very similar to TEI and now uses TEI) (Gregory et al. 1967–2022) and the Kant-Index (built on TUSTEP) (Hinske 1982–2019), which granted access to Kant’s writings in lemmatized form with metadata and semantic annotations that are interoperable with regard to multilingualism (i.e., Kant’s use of Greek, Latin, German, and French).

Terminology, as a system of concepts, is the cornerstone of any philosophical theory. Today, we can empirically calculate the meaning and complex relations of meaning, including polysemy, hyponymy, relational meaning, and change of meaning (over time) of concepts and terms in Kant’s terminology and that of the philosophers who influenced Kant, by extending, integrating, and comparing philosophical research and theorizing. Most importantly, since the dating of some of Kant’s writings is partly unknown or disputed (Hinske 1982; Stark 1993; Capozzi 2002), correct dating of whole texts but also of fragments even of individual pericopes is a prerequisite for detecting semantic changes in Kant’s conceptual treasury in the longitudinal direction.

In recent decades, Kant research has brought to light many of the sources and influences of Kant’s thought, which includes both traditional research on the history of philosophy (Zammito 1992; Banham et al. 2012; Hohenegger 2020) and the use of quantitative analysis (Gregory et al. 1967–2022; Hinske 1982–2019; Roelcke 1989). It is to be expected that the amazing array of sources listed in the AA-Neuedition will push forward research on Kant’s engagement with the history of philosophy.

Considering current trends toward a data-driven history of philosophy as a branch of the digital humanities, the idea is that the history of philosophy will profit from finding ways to radically improve the manner in which we curate, archive, annotate, access, and translate corpora. If it is true that tools for this kind of analysis have not yet been implemented, it is also true, as suggested by Moretti’s (2013) notion of distant reading, that experimenting with meaningful patterns in philosophical corpora is a step towards making new machine learning technologies usable for tackling problems in the humanities.

The AA-Neuedition lays the groundwork for studying and discussing Kant’s philosophy in a global context, a long-term effort that relies on the synergies between philosophy, computational linguistics, machine learning, and translation studies. Research into Kant’s translations has expanded in recent years and offers a good overview of Kant’s translations in Latin alphabets (Schlüter and Hohenegger 2020). However, Kant’s translations in non-Roman scripts remain a little-explored area (Pozzo et al. 2022). Think of the potential of interconnecting Kant’s writings in German and Latin with their current complete translations in, e.g., Chinese (Li Qiuling, Kant 2003–2019), English (Guyer and Wood, Kant 1992–2006), and Russian (Tuschling and Motroshilowa, Kant 1994–2020). All things considered, one can say that the AA-Neuedition provides unrivalled material for historical-philosophical investigations.