University of Virginia Library

Some Notes on Letter Editions: With Special Reference to German Writers
by
Siegfried Scheibe [*]

The editing of correspondence has increased considerably in the German-speaking countries over the past few decades. Numbers of editions of letters written by and to historic personages were begun, opening up major new materials and sources, especially from the 18th to the middle of the 20th century. In the GDR alone, just to mention that country, letter editions of the following personages are being published: Christoph Martin Wieland, Georg Forster, Friedrich Schiller, Johann Gottfried Herder, Adelbert von Chamisso, Heinrich Heine, Alexander von Humboldt, Karl Marx and Friedrich Engels. The special form of a calendar edition of all letters to Johann Wolfgang Goethe should also be mentioned in this context. However, the aforementioned editions are confined to historical and critical letter editions which always comprise both the letters written by a given author and those addressed to him. In addition, numerous collections of letters (partly complete, partly selective) of German and foreign writers, artists, scientists and politicians were published in the GDR in the form of so-called study editions or reading editions on which I do not want to elaborate here. Scholars take a great interest in direct statements made by such persons in letters available to us: they provide important information about the living conditions of these persons and problems relating to their work, comments on other persons from among the author's acquaintances, and, last but not least, information about the period in which the author lived and worked. Such information is hardly available otherwise in such a condensed, although always subjective, form. The host of recent letter editions, e.g. those reflecting the 18th century, will largely add to and differentiate the picture of that period in scholarly research. The interest of the general public in such letter editions, however, is more difficult to estimate. At least the publishers complain

about the fact that letter editions which are intended for a wide audience do not sell easily.

But I do not want to dwell upon the economic problems that are always associated with letter editions, and historical and critical editions in general. I will deal with some concrete problems we encountered in the editing of letters by and to Georg Forster and Christoph Martin Wieland. They are certainly of significance for the editing of correspondence beyond the two editions and may stimulate the discussion on aspects of letter editing, which has been not so pronounced so far.

Dealing with a "letter", we must know first of all what it is, as the postal definition of a letter is not sufficient nor applicable for our purposes. It seems more appropriate to begin with a purely formal definition of the term. A letter could be described as a text which begins with a salutation and ends with a signature. Anywhere in this text there will be a date and the place where the letter was written. This formal definition of a letter, however, is not sufficient to differentiate it from a work composed in the form of a letter. A "literary letter" may also have a salutation and a signature or even a date and a real letter may lack these particulars fully or in part. Hence, the definition of a "letter" must also include criteria relating to its contents such as the following: the letter is addressed to a person who is not in the same place (room) with the letter-writer. The letter contains a message, information or orders which could be given to the addressee orally in about the same way if he were present. And the message, information or orders contained in a letter are only meant for the addressee (and possibly a narrow circle of other persons such as dependants and friends) and the writer did not intend to publish them in this form. So the decisive criteria relating to a letter's contents are its message character on the one hand, and the intimate character of the information on the other.

From this we can derive the following summary definition of a letter:

This definition, for one thing, covers a major part of what we are generally inclined to classify under the term "letter". These things pose no problems when being included in a letter edition, namely (apart from actual letters) notes, postcards, telegrams and other things of this kind, i.e., texts which contain a message meant for an addressee. But in reality we encounter a number of other messages, pieces of information or orders which, as far as their form and structure is concerned, comply with the definition of a letter but which at least give rise to some doubt as to whether they should be classified as a "letter" and thus be included in a letter edition. Some cases in point will illustrate this. Until far into the 19th century, IOUs, bills of exchange, and other business documents were commonly written in the form of letters, which makes them conform to the above-mentioned definition of a letter with regard to contents and form. However, these are usually "letters" we would not send like that today but take a printed form to be filled in instead. If such records were available in the compilation of material relating to a 20th-century author it would not occur to anyone to regard such filled-in forms, e.g. for financial matters, as letters and include them in a letter edition.

Other borderline cases are inscriptions in books and entries in autograph books, etc. Just like a letter they also contain a message to the addressee (although autograph book entries are often made in the addressee's presence) and their basic structure is also that of a letter. The message they contain, unlike a message in a letter, does not constitute a piece of information which is valid only for a certain period but is supposed to preserve and confirm the memory of a given person or visitor. On the other hand, there are also actual letters which were written for the sole purpose of lasting remembrance and thus hardly differ from book inscriptions or autograph-book entries as regards their contents.

Therefore, when preparing a letter edition, the editor has to make the important preliminary decision of whether or not to include such messages of a letter-like nature. All letter-like records mentioned so far contain statements about the author, his living conditions, his encounters and conversations with other persons, all of which may be relevant to his biography. So it is indeed desirable to present them in an edition in order to make them accessible to research. However, whether to include them in the letter edition itself cannot be said in general but has to be decided largely pragmatically. The decision is relatively easy in the case of a complete edition of all the works of an author planned to comprise volumes with biographical records, records of conversations and other material of this kind: here the records mentioned above may be clearly related to biographical ones and should be printed as such. If, however, a pure letter edition is planned or a complete edition of works and letters where such biographical records would not appear, then it will be harder to decide to exclude this material. Moreover, it is often very difficult to differentiate exactly between such records and actual letters, for various types of records may always mingle in various ways. For instance, among Georg Forster's there are some "pure" letters

which contain an IOU or a bill in their text. This gives rise to the question why these financial documents that appear within a letter are included in the edition, while separate such documents are not. Even a restriction to the calendar form (i.e. the paraphrasing of such a "letter") does not leave the editor in a better position, as the calendar version will hardly be shorter than the "letter" itself. It is therefore more appropriate (and makes an interpretation less ambiguous) if the original form of the text is reproduced.

But it is not only problems like these that we are often faced with when preparing letter editions. Many of the authors we have to deal with were not "free-lance writers" all their life but often had an occupation where letter-writing was one of their functions. These letters are not private statements by these persons but official notices of an authority, institution, etc. On the other hand, they were also addressed to a private person or another authority, which makes them comply with our above definition of a letter, i.e. they contain messages, information or orders for a specific addressee. And their contents and gist are largely determined by their author, whose individual style can be identified even if he had to adapt it to the set officialese. Such correspondence is mostly known as "official letters". Just like personal statements by the author they are relevant to his biography, often making him appear in a way different from his literary work. That is why the inclusion of such letters in a letter edition should not be precluded from the start. (In the case of Goethe, for example, it is indeed relevant that in the criminal case of a woman child-murderer, as an official person he demanded that she should be sentenced while as a dramatist he pronounced a moral acquittal on Gretchen in Faust.) German letter editions used to do without such "official letters". Meanwhile, there is at least a trend towards including such letters, as the scholarly community wishes to get an impression of the author that is as comprehensive as possible. Goethe again provides an interesting example for developments when it comes to including official letters: there were no such letters in an edition of Goethe's letters published at the turn of the century. In 1949, a series to publish his Amtliche Schriften ("Official Works") was begun. At first the editors believed that they could restrict the series to actual works or "writings", i.e. comments on certain facts with which Goethe dealt in his capacity as a minister in Weimar. Soon they realized that letters should be included as well and eventually they saw that there was an interaction between official and purely private messages and that full understanding was possible only in their overall context. Goethe did not only give official comments on most official affairs. He often used private correspondence as well in order to give his opinion on an official matter, discuss a decision to be made

in advance, or provoke a certain attitude towards this matter in other people. And he also used letters with largely official contents for adding personal messages. So the boundaries between Goethe's official and private spheres are fluid. But this is not only true of Goethe if we think of what we do when writing private or official letters. The more intimate the terms on which we are with the addressee, the more statements on private and official occasions mingle, which frequently renders a clear distinction between the two types of messages impossible. Even if there is an alternation between distinctive official and private correspondence to the same person, one letter will often not be understood without knowing the other. If an edition is restricted to private letters, at least the relevant annotations should quote major parts of the official letters in order to explain allusions or statements. However, where this is frequently necessary, the question arises whether a letter edition should not be designed from the start so that it may include both private and official letters, as it was done in the edition of Georg Forster's letters.

Another problem case is formed by letters an author wrote in his capacity of "correspondent" of a newspaper or "corresponding member" of an institution. Today these terms are merely regarded as a kind of title without an obligation to correspond, whereas in the 18th century these words were applied in accordance with their meaning: they refer to persons who sent letters with information to a newspaper or institution. In these letters they reported news from the various scientific, cultural or political fields they dealt with or took a particular interest in. This occurs with both Georg Forster and his father, Johann Reinhold Forster, and especially in the period when they lived and worked in England. We hold letters written by them to the Berlin publisher Johann Karl Philipp Spener whose publishing house was responsible for the Haude und Spenersche Zeitung, one of the major newspapers in Prussia. These letters embody normal personal statements side by side with information and news about recent research and research findings in England, about scientific literature and fiction there, about the theatre, about political affairs in England, etc., which were expressly meant to be published in Spener's newspapers. (We find the same in letters by Johann Reinhold Forster to Christian Gottlob Heyne, a professor in Göttingen and editor of the Göttingische gelehrte Anzeigen, a journal which did not only publish reviews but also information about recent research both at his own academy and at academies abroad.) As long as such letters are available to us in their original form, i.e. as a mixture of private statements and official information, their inclusion in a letter edition will be beyond dispute. The problem becomes more intractable if there is no record of the letter concerned, while we find a notice in a

newspaper or journal which might read: "Our correspondent in London reports the following news", mostly followed by direct quotations. It is irrelevant whether the name of the correspondent appears or whether it can be deduced from the topic and way of presentation. What must be decided is what to do with this text: whether it constitutes a letter to be included in the edition or not. It may be argued in this case that the text in the form available does not comply with our definition of a letter, as it is information meant for a wide circle of readers and written for the purpose of being printed. But this can be countered by saying that this kind of text was not composed directly in the form of a work (an essay, for instance) but that the facts were originally told in a normal letter where personal and general information was integrated. In this respect, this is first of all always personal information, which the publisher of the newspaper or journal never published in full but from which he selected those items that he deemed to be of general interest based on the situation in his country. Since there is enough documentary evidence for the fact that such correspondent news was indeed part of normal letters which no doubt would be reproduced in a letter edition, the news by a given correspondent as quoted in newspapers and journals must also be included in a letter edition.

While in all problem cases outlined so far we find a trend towards including records of the kinds mentioned in a letter edition, this may probably be ruled out from the start in the case of what is known as an "open letter". This special form of a letter was definitely written for publication and, as a rule, became known to the addressee only as published. An "open letter" may be considered as a specific type of essay which, just like the "literary letters" mentioned earlier in this paper, was couched in the outer appearance of a letter without being one. I have so far no evidence that an essay published as an "open letter" has ever been sent directly to the addressee as a letter. If, however, this should occur, then this particular letter, which will be on hand as a manuscript or typescript, definitely belongs in a letter edition, while an "open letter" published in this form equally definitely belongs in the edition of the works of the author concerned where all his other "open letters" should be reproduced, too.

However, a distinction must be made between this and a form of polemic letter publication which was quite common particularly in the 18th century in connection with a controversy among various writers or scholars. In this case, the rivals initially exchanged private letters which were not meant for publication. Only as the controversy escalated the one, and sometimes also the other, resolved to go public, give a generalized account of the facts and substantiate them by publishing his own

letters or those from his rival. Hence, such publications contain letters which definitely belong in a letter edition of the given person, while the summary essays (including the letters reproduced in them) must be regarded as a work and thus belong in an edition of the works even if they are written in the form of an "open letter".

I suppose that our definition and the additional remarks have shown what should be included in a letter edition as a "letter". This, however, does not solve all the problems associated with the inclusion of letters in an edition because we must now differentiate between the individual letters of an author. While so far we asked: what is a letter?, we must now answer the question: what is one letter?

Here we can start out from the formal definition given at the beginning: a letter starts with a salutation and ends with a signature. Does this rule also apply if a postscript is added? The function of postscripts is to embody additions to a given text of a letter. Hence they are supplements to the actual text of the letter which were usually written after the conclusion of the letter, i.e., after the writer had signed his name. In this respect, postscripts do not alter our formal definition of a letter, the body of which still ends with the signature. However, being supplements to letters, postscripts form a homogeneous whole together with the text of the letter and have to be published along with it.

The case is different when a letter to addressee B is enclosed in a letter to addressee A with the request that A should hand over the enclosure to B. Here both the letter to A and that to B must be regarded as independent letters and have to appear separately in the edition. However, it should be indicated for both of them, in the description of the original record or in the annotations, that there is a certain relation between the two, i.e., that the one was an enclosure in the other.

The decision becomes more difficult if the letter to B (who might be a relation of A's) immediately follows the letter to A, i.e. if it is written on the same sheet of paper. Here too, the decision whether this is one letter for two persons or two individual letters depends on the formal structure of the text. If the letter to B also contains a salutation and a signature (after the letter to A had been closed with a signature) it should be considered an independent letter and reproduced accordingly. If, however, the letter to A merely ends with a remark such as "And now some words to your wife", after which the text is immediately continued, then the letter should be regarded and printed as a homogeneous whole and the heading inserted by the editor should read "To A and B."

The same applies where a letter is meant for two different persons who are possibly even alternately addressed in the text ("To A", "To B", "To both"). Here the formal unity of the letter, limited by the complementary

close and signature, which are meant for either addressee, would compel the editor to assume one homogeneous letter and his heading should again be "To A and B".

Finally, a special case is formed by those letters of an author where another person added a postscript or, in the reverse case, letters of another person where our author added a postscript. The decision on which parts of a given letter should be printed in the text section of an edition depends on the author to be edited: only his text (or, in correspondence editions, also the text addressed to him) is of primary relevance to the edition. Hence, the postscript added by another person would be completely disregarded when the letter is printed in this section of the edition and likewise only the postscript added by our author to another person's letter would appear whereas the letter itself would not. (In this case the editor should mark it "Postscript to a letter from A to B"). This regulation in this strict form only applies to the text section of a letter edition, for it can be assumed on principle that the author knew the other letter to which he added a postscript and that he also took notice of the other person's postscript to his own letter. Since these other texts may be relevant to the progress of the correspondence with the addressee it is always advisable to print them in full in the annotations or in connection with the description of the record. This is not only important for possible interpretations of relations between our author and the person who wrote the body of the letter or the postscript. It would also be a chance of opening a letter of this person up to the public. (It should go without saying that such complete reproductions of other texts are clearly indicated to the user in an edition's table of contents.) An edition, particularly a letter edition, should always attempt to present as much additional material as possible and reasonable which may give a more distinct picture of the author in the context of his personal circumstances and his period. (Note that I mean primary sources and not comments or interpretations on the part of the editor.)

This also applies to the inclusion of letters written by a person close to the author (a relation, a friend, etc.) instead of the author himself. In connection with Georg Forster's Vilnius period, for instance, there are a number of letters written by his wife to her father, where she describes their life, pointing out that her husband has no time to write himself and conveying to her father requests and orders from her husband. Such letters are substitutes for letters not written and contain important personal and general statements which the author would have made himself if he had had the time to write. These facts are lost to research (or could be found only by chance or with difficulty) if substitute letters are not published within the framework of the edition. Although they should

not be included in the series of letters of the author, i.e. not in the main section of the edition of this author's letters, they should be reproduced in an appendix to the given volume (with their contents also being reflected in the indexes) so that this important additional material can be used for purposes of scientific research.

So far we have dealt with letters the text of which is directly available to us. In this context I ignore all problems arising from the kind of the records available, i.e. whether we have original letters, transcripts made by the author or others, more or less complete and reliable printed copies or only quotations of individual passages in other persons' letters or in autograph catalogues, etc. Nor do I wish to dwell upon the different treatment of these materials in editing and problems arising therefrom. This would be another subject. What remains to be discussed, however, is the question what we editors should do with letters of which there are no records. Because for every author we can assume that at least as many letters have been lost as have been preserved, if not that the number of letters lost exceeds that of surviving letters.

In modern German editions it has become common practice to include in the edition also a certain part of lost letters, namely those for which there is direct evidence. On the one hand, this may be welcomed because it increases the number of known letters considerably and it may shed light on persons, as correspondents of the author, whose relations with the author have not been known so far. However, the inclusion of such inferred letters also brings problems, as the user is tempted to assume that now the edition covers all letters the author ever wrote or that were addressed to him. This would be a misinterpretation and the editor has to counter it in advance.

But there is another important problem. Since the publication of Briefwechsel Wielands ("Wieland's Correspondence")—where this method was applied for the first time after 1945—it has become the practice in German letter editions that both recorded and inferred letters are put on one level and treated equally. Both types are printed in the text volume in chronological order and numbered equally. This, however, has major disadvantages: the date of an inferred letter cannot be definitely ascertained in most cases; the editor has to confine himself to giving a terminus ante quem instead. Now if a letter which was merely inferred so far is found—and this may happen relatively often despite all efforts to cover the whole material in preparing such an edition—there will at any rate be a discrepancy between the place in which the letter now is proved to have been written and the place were it was so far supposed to have been written on the basis of a vague date. Hence, a letter has to be inserted in the addenda to a letter volume already published

in a place not marked before, while there is now a gap in another place which was marked so far. This cannot be avoided even if the editor takes pains to completely reconstruct the correspondence between the author and all his correspondents. It does not bring absolute certainty of the letters actually written either. But also every letter which was hitherto unknown (i.e. neither available as a record nor inferred) or which suddenly appears on the autograph market may entail the inference of a hitherto unknown letter which again has to be inserted subsequently in the order of the text volume already published. These two examples alone make it clear that recorded and inferred letters indeed cannot be considered equal nor treated equally in an edition. That is why their presentation within a homogeneous chronology is rather hazardous and does not provide the hoped-for exact survey of the author's correspondence. Therefore, a method applied to the recording of inferred letters in a number of Slavonic editions seems more appropriate: the two types of known letters are separated and treated differently in accordance with their differing significance. Only the recorded letters are printed in the text section of the edition whereas all inferred letters are referred to an appendix to the respective volume where their identifiable content is described. Thus, changes in the chronology, which may occur due to the finding of a letter which was hitherto merely inferred or due to the inference of other letters, do not have such serious effects on the arrangement and reproduction of recorded letters as in the German editions. Therefore, new or planned letter editions are recommended to apply this method which takes account of the differing significance of the two types of known letters. This method, however, does not affect the fact that it is useful for the user to be exactly informed in a letter edition also about the letters which can only be inferred.

It may be important to name the criteria which make the inference of letters possible but also show the limits set to a safe procedure. Letters should be inferred in the following circumstances:

Particularly in the case of uncertain indications ("about two years ago") it is important and indispensable to know exactly both the author's working method and his general behaviour in order to really infer a letter. As soon as an author is suspected to mention an alleged letter only to justify his long silence no letter should be inferred. However, the disputable fact has to be explained in the annotations to the passage concerned. (We encounter such cases for instance for Bertolt Brecht but also for other authors.)

On the other hand, letters should never be inferred in the following uncertain circumstances:

In all these cases, the possibility of the existence of a letter which cannot be proved definitely should be indicated, if necessary, only in the annotations to the letter on hand.

In connection with the inference of letters we have come across the problem of the truth contained in statements in letters. A separate study would be necessary to make a thorough inquiry into this important subject. What I want to do here in closing is warn against using the record of a letter alone as evidence for the truth of a given statement, because the writer of a letter gives his opinions always subjectively from his own view of things and often does not strive for objectivity. On the other hand, he normally adapts himself to the addressee of his letter, taking account of the latter's potential sensitive points, state of tension with certain people, aesthetic, moral and political conceptions, and filtering his statements accordingly. Where quotations from letters (or from autobiographical writings) are used to prove a fact, not only the writer of this text but also the addressee should be taken into consideration. The truth of a statement can be proved with a degree of certainty not until a similar statement is found in letters to other persons or additionally substantiated by other evidence. To cite one example: Georg Forster's statements about Goethe in his letters to Friedrich Heinrich Jacobi are always cautious and critical. If we only had these commentaries on Goethe by Forster and did not know that in those years Jacobi was at enmity with Goethe (for which Forster showed consideration in his letters to him), we would get a wrong impression of Forster's attitude towards Goethe. Our judgment is clarified as soon as we refer to Forster's statements about Goethe in letters to other persons, thus receiving diversified information on Forster's attitude. Since the presence or absence of letters (and thus, the judgements contained in them) is always dependent on chance, the user of a letter edition should use all those records with a pinch of salt, and critically. Important as information and opinions contained in letters may be, they are never objective reflections of real facts and must not be regarded as such. The editor should raise doubts as to the credibility of a statement in the annotations and, if necessary, point to different remarks which were made to another correspondent and qualify certain statements. However, he should not allow his edition to become an omniumgatherum of contradictory statements by various persons but should rely on the knowledge and ability to differentiate in the users of his edition.

Addendum

• Some papers on letter editing in German language:
• Böhm, Hans: Neue Weimarer Ausgabe. Bemerkungen zur Neubearbeitung der Briefe und Tagebücher Goethes. (Abteilung III und IV der Weimarer Ausgabe.) Goethe. Neue Folge des Jahrbuchs der Goethe-Gesellschaft 29 (1967), S. 104-138
• Fetzer, Günther: Das Briefwerk Hugo von Hofmannsthals. Modelle für die Edition umfangreicher Korrespondenzen. (Marbach am Neckar 1980.) (Deutsches Literaturarchiv. Verzeichnisse, Berichte, Informationen 6)
• Hahn, Karl-Heinz: Briefe an Goethe. Erklärungen zu einer geplanten Regestausgabe der an Goethe gerichteten Briefe. Weimarer Beiträge 6 (1960), Sonderheft, S. 1125-1146
• Hahn, Karl-Heinz; Reuter, Hans-Heinrich: Fünfte Abteilung der Weimarer Ausgabe. Die Briefe an Goethe. Regestausgabe. Goethe. Neue Folge des Jahrbuchs der Goethe-Gesellschaft 29 (1967), S. 65-103
• Probleme der Brief-Edition. Kolloquium der Deutschen Forschungsgemeinschaft Schloβ Tutzing am Starnberger See 8.-11. September 1975. Referate und Diskussionsbeiträge hrsg. v. Wolfgang Frühwald, Hans-Joachim Mähl und Walter Müller-Seidel. (Boppard 1977.) (Deutsche Forschungsgemeinschaft. Kommission für germanistische Forschung. Mitteilung II)
• Reuter, Hans-Heinrich: Die Regestausgabe sämtlicher an Goethe gerichteter Briefe. Zugleich Thesen über die prinzipiellen Möglichkeiten und die Methoden der Darbietung eines Briefnachlasses in Regestform. Euphorion 62 (1968), S. 150-159

Notes