VERSCHLIMMBESSERUNG:

CORRECTING THE CORRECTIONS IN TRANSLATIONS OF KANT

The present article, which analyses the translation-histories of passages taken from Kant’s Kritik der reinen Vernunft and Prolegomena, was prompted by the author’s long-standing dissatisfaction with the more recent translations of an important passage in the Preface of the latter work, and by discussions of translation issues appearing in Interpreting Kant. (Gram 1982.) Its method, though, was inspired by several articles (especially those appearing as chapters 2, 3, 4, 5, and 6) in The Sign of Three. (Eco and Sebeok 1978.) The problem of the meaning of the texts was taken as a puzzle to be solved in the manner of a Holmes or Peirce. The similarity between these case studies and the two Sherlock Holmes tales The Man with the Twisted Lip and Silver Blaze proves to be illuminating.

In his article ‘Is Sensation the Matter of Appearances?’ (Aquila 1982: 11-29), Richard E. Aquila takes up the issue of ‘how accurately one ought to reproduce elements of vagueness or ambiguity in translation’. (Aquila 1982: 11.) After discussing semantic ambiguity, he investigates ‘ambiguous’ pronoun references. He refers to the following passage at A42|43=B59/60, of Kant’s Kritik der reinen Vernunft:

Aquila claims that ‘the pronouns in question’ (which he does not specify, but which must be the sie in lines 5 and 7 above, and Jene in line 4) refer back to unsere Art of line 1.

Aquila evidently believes that the translations are irreconcilable with text, and therefore, in some sense, ‘impossible’. The poor quality of the texts is well documented, however, and the translators have all had to deal with them. Although Aquila conflates the three translations in his article, they need to be examined separately. (We have indicated the translations of the ‘pronouns in question’ by using bold italics.)

[Meiklejohn, 1855]:

[Müller, 1881, 1896]:

[Kemp Smith, 1928, 1933]:

The first problem is the pronoun Jene in line 4. In German this pronoun is generally used to refer to the first of a pair of nouns and is usually translated as ‘the former’ when paired with diese, ‘the latter’. (There are two pairs of these pronouns in successive sentences.) It appears to refer to ‘Raum und Zeit...die reinen Formen derselben’, which is clearly a plural form. Yet there is no plural verb form that is associated with this pronoun. The only verb that can be related to it (heißet) follows a conjunction, (und), and another pronoun, (sie), and is singular: (A literal translation would be: ‘Space and time alone can we recognize a priori...and it is called pure intuition’.) This problem is not nearly so difficult as it seems. Patterns of verb-noun agreement in German are somewhat different than in English, and this is a perfect example of such a difference. (See: Curme 1905: 501-507; Lockwood 1968: 203-205; Priebsch and Collinson 1958: 342-343.) Müller simply translates this as ‘are called pure intuition’. Müller, then, is following the well-recognized principle of constructio ad sensum. It needs no special justification. The following points, though, may be relevant:

1) Kant repeatedly refers to ‘space’ and ‘time’ as ‘pure intuitions’ or ‘pure forms of intuition’. It is inconceivable that he would intend anything else but to equate them. Because of attraction, the verb can sometimes agree with the predicate; (the proximity of ‘reine Anschauung’ must have influenced Kant to use a singular verb form, ‘heißet’ ‘is called’).

2) The very next sentence begins also with Jene, which clearly must refer to ‘Raum und Zeit...die reinen Formen derselben’, because of: a) the meaning of the sentence; b) the lack of an intervening sentence which could supply a new noun to refer to; and: c) the presence of a corresponding plural verb form, anhängen, ‘inhere’. Since the second Jene clearly refers to ‘Raum und Zeit’, etc., and since both pronouns must refer to the same noun or nouns, then the first Jene must also refer to ‘Raum und Zeit’.

3) Therefore the sie of line 5 must also refer to ‘Raum und Zeit’, etc., because of the way the sentence reads, i.e., they are linked by the conjunction und (‘and’).

The only pronoun left is sie in line 7. Following Müller, again, it seems necessary that this sie (feminine) must refer to ‘unserm Erkenntnis’ (neuter) despite the apparent incongruity of the gender of the pronoun and the noun. Aquila seems puzzled by Müller’s choice here: ‘feminine, nonetheless Müller.....’. (Aquila 1982: 16.)

Among them, Karl Vorländer, (1899) and Benno Erdmann, (1889) emend ‘unserm Erkenntnis’ (neuter) to ‘unsrer E.’ (feminine) in this passage. The Akademie edition (edited also by Benno Erdmann, 1904), curiously, does just the reverse, and, following the text of the third edition (1790), changes the sie to es, thus connoting its relation to the neuter Erkenntnis. The following statement in the entry for Erkenntnis in the massive Grimm dictionary lends further support to these emendations, and to Müller’s comments:

The first translation to be considered was published in 1819 by John Richardson:

At this, the earliest stage in the history of this passage in English, the translator has clearly ascribed die Erinnerung to Hume, not to Kant. Richardson’s translation of Erinnerung as ‘hint’ contrasts sharply with 20th-century practice.

The problems apparently began in 1891 with the translation of Ernest Belfort Bax:

Carus opted for the following treatment in his 1902 version:

In 1977, Ellington, apparently following Beck, translated the passage in the same manner:

Beck and Ellington may have picked up this notion of ‘remembering’ from Norman Kemp Smith, who translated the passage in his Commentary. (Smith 1923: xxix-xxx.) Smith uses both ‘teaching’ and ‘recollection’:

Kant’s statement, frequently quoted, has been mistranslated by Kemp Smith, Blakney, Beck, and Ellington. They have, it appears, confused two quite distinct meanings of Erinnerung by misconstruing the genitive (des David Hume) following die Erinnerung. The use of the genitive here indicates that this was something that originated with Hume; it is called the ‘Genitive of Origin’. (Curme 1905: 512.). These translators have apparently mistaken it for the ‘Objective Genitive, which denotes the object toward which the activity is directed.’ (Curme 1905: 512-513.). This latter, however, according to Curme,

It is difficult to be certain to which specific utterance of Hume Kant is referring here, but the remarks on the relationship between Hume and Kant by Sir William Hamilton, the noted early 19th-century scholar, though more a paraphrase than a translation, may shed light on the passage:

Further instance of Kant’s usage of Erinnerung in the sense of ‘admonition’ can be found in the first Critique (at A98) where he issues a Vorläufige Erinnerung to the reader, before plunging into the Deduction of the Categories, because the Deduction is ‘attended with so many difficulties’ (‘mit so viel Schwierigkeiten verbunden...’). Here, Erinnerung must be taken to mean something like ‘admonition’, ‘cautionary note’, or ‘preliminary remark’ (Müller’s translation). Another clue comes from the (obsolete) compound Vorerinnerung (‘premonition’ or ‘preamble’), which does not make much sense if Erinnerung can mean only ‘recollection’, for we cannot (at least not yet) ‘remember the future’.

Consultations with Adelung and Grimm, with several 19th-century bilingual dictionaries such as the Muret-Sanders, Felix Flügel, [Flügel]-Schmidt-Tanger and others, and with Spalding have confirmed the accuracy of the translation of Erinnerung as ‘admonition’ or ‘critical remark’; e.g:

More recent (20th-century) German-English dictionaries do not list this group of meanings as prominently, if at all. Part of the reason for the shift in the translations of this passage from ‘hint’ (1819) and ‘warning voice’ (1881) to ‘remembering’ (1977) may be the use of these more modern dictionaries by the translators.

Intra-textual clues are provided in the present case as well. Earlier in the Preface, (KPA: 257) Kant refers to Hume’s ‘attack’ (Angriff) upon metaphyics, and in the paragraph immediately following the one under consideration, we find:

Because of its proximity, the word Einwurf (which Richardson translates as ‘scruple’, and Beck and Ellington translate as ‘objection’), must be understood as amplifying Erinnerung. Later, (KPA: 262) Kant refers to Hume’s ‘doubt’ (Zweifel). One can see then, how neatly ‘warning voice’ fits into a text wherein we have Hume’s ‘attack’, ‘objection’, and ‘doubt’. These statements, considered together, lead inescapably to the conclusion that Kant is referring to one thing, viz., Hume’s ‘earnest presentation of reasons in opposition to something’.

1) Kant here is referring to some kind of first encounter with Hume, not a ‘recollection’. It is difficult to see how Kant could have been more strongly shaken by his own memory of Hume than by his first encounter with Hume’s criticisms.

2) Nor does it seems likely, that a mere eleven years (1772-1783) would, even for Kant, qualify as ‘many years ago’. A time in the ‘60’s or perhaps even the ‘50’s is quite likely. So one would do well to accept this constraint, and isolate the search for the ‘awakening’ in the period before 1770, although much speculation by recent scholars has centered on the Winter of 1772. In particular, Kant’s letter to Marcus Herz of February 21, 1772 is often thought to be supportive of the thesis that Kant became (in Wolff’s words) ‘reacquainted with Hume’s attack on causality’ through a translation, published in that year, of James Beattie’s Essay on the Nature and Immutability of Truth (1770), in which Beattie quotes from Hume’s Treatise, with which Kant was (supposedly) previously unacquainted. (Ewing 1969: 15; Wolff 1962: 22.). As Isaiah Berlin points out, however, Kant could easily have learned of Hume’s position from his friend J.G. Hamann, who did know English, but ‘was a life-long student of Hume’, and ‘preferred him to his old Königsberg friend’. (Berlin 1977: 95-116.)

The purpose of this digression is to investigate an apparent case of historical/translational circularity: it seems that the mistranslation of Erinnerung in this passage has been given support by the presupposition that it refers to that which it is used to prove! There is no factual evidence whatsoever to support it. Kemp Smith himself admits: ‘The evidence in support of this contention is entirely circumstantial.’ (Smith 1923: xxx.) And Wolff also confesses: ‘There is no direct evidence as to what produced this "recollection" (Erinnerung) of Hume, but the most probable cause was [the 1772 translation of Beattie’s book].’ (Wolff 1962: 25.). Wolff appears to have been influenced by Kemp Smith’s speculation, and quotes from the Beck translation in support of this thesis.

When revising older translations, one must be careful to avoid disimprovement, or, as its is called in German, ‘Verschlimmbesserung’: the creation of a worse situation by ‘correcting in errors’. The difficulty involved in evaluating and revising previous translations of such texts, however, cannot be overstated.^_ It is, therefore, easy to understand how and why passages like these have confounded so many.

Beck’s renunciation has not reached all quarters, because his original 1950 translation, and the views expressed in his Early German Philosophy, continue to influence Kant scholarship. A late example of this is the article by Frederick C. Beiser on Kant’s intellectual development which appears in a recent collection. (Guyer 1992.). Beiser assigns four phases to Kant’s development, the last of which results from his ‘recollection of Hume’, which Beiser places after 1772, again citing the appearance in German translation of Beattie’s Essay. If, however, this ‘recollection’ of Hume is merely a mistranslation, Beiser’s position becomes more tenuous. It is worth noting indeed that Beiser mentions the

On the other hand, Manfred Kühn’s article Kant’s Conception of ‘Hume’s Problem’ (Kühn 1983) though older by ten years, is informed by an awareness of Beck’s renunciation, and the conclusions he draws (though based on different data) coincide with the present investigation. Kühn argues for an earlier date for the ‘awakening’ (1771) than does Beiser, though Kühn, too, notes that Kant

Kühn cites the appearance of a German translation of the last chapter of Book I of Hume’s Treatise in the July 5th & 12th Königsberger Zeitung. Kühn argues that this finally awakened Kant.

The question remains regarding how far back Kant was referring to in 1783. Both Beiser and Kühn mention, quite explicitly, Kant’s early acquaintance with Hume’s criticisms. Denying the historicity of the 1772 ‘recollection’ itself naturally opens up the possibility of an earlier date. The author, therefore, leans toward a date earlier—perhaps much earlier—than the early 1770’s. The skeptical attitude already displayed in the Dreams of a Spirit Seer (1766) is otherwise difficult to explain.

Erdmann, Benno, ed. (1889). Hamburg and Leipzig: Leopold Voss.

Michael A. Scarpitti (b. 1949) is an independent scholar of Philosophy whose principal interests include the English and German mind of the eighteenth century, as well as Kantian exegesis and translation theory. He is a member of the North American Kant Society.

Susann Möller (b. 1951) holds a Ph.D. in German. She has studied Philosophy and pursued her interests in Kant under Ingeborg Heidemann in Bonn. In the United States, she has taught the History of Ideas and Comparative Literature, and currently focuses on Cultural Linguistics.