From the WAB archives: A selection from the Bergen Wittgenstein archives' Working Papers and audio-visual materials

1. Introduction

In his catalogue of the Nachlass¹, G.H. von Wright writes, "A few pages [of TS 211] are missing, but can be identified in 213. Photocopies of the missing pages have been inserted."

In the course of transcription work at the Wittgenstein Archives, however, it became apparent that these replacements need detailed documentation. They have been taken from various versions of TS 213 and are not marked as copies, such that the unwary reader is likely to take them in good faith to be authentic².

In addition to the pages shared with TS 211, TS 213 also contains two pages removed from TS 210. These have also been replaced in their source by Xerox copies.

The aim of this survey has been a better documentation of the replaced pages in TS 210 and TS211.

2. Background

Wittgenstein compiled TS 213 from three previous typescripts. He produced a rough draft for the new work by rearranging fragments from TS 208, TS 210 and TS 211. The resultant collection of cuttings has survived largely intact, and is listed by von Wright as TS 212. From this collage, and making only negligible alterations to the text, Wittgenstein then dictated a clean copy of the work, ensuring, as usual, that the typist produced at least one carbon copy. It is this final clean version (and often only the top copy thereof) which is now referred to as the Big Typescript ( TS 213).

All of TS 208, TS 210 and TS 211 once existed at least in duplicate ─ as top copy and one or more carbon copies. The top copies of TSS 210 and 211 have survived more or less intact. The carbon copies, on the other hand, served as raw material for Wittgenstein's cut-and-paste technique of revising his texts.³

In preparing the collage TS 212, most of the material from TS 210 and TS 211 was cut into fragments of less than a page in length. But occasionally Wittgenstein was satisfied with larger spans of the original text and was able to incorporate in his draft whole pages from the source typescripts. Later, during the stage of dictating from this rough version, the intact pages were transferred yet again from the collage into the final, clean copy. Thus TS 213 ─ the ultimate product of this process ─ now contains pages which were once physically part of TS 210 and TS 211.

In total there are 75 such adoptive pages in TS 213. Of these 73 are taken from TS 211 and 2 from TS 210. In their new context these pages are easily recognisable due to their double pagination. Central in the upper margin ─ and invariably undeleted ─ is the number the page had in its position in the source typescript, while the number that marks its position in TS 213 has been added in the upper right-hand corner.

Since Wittgenstein was producing not just a top, but also a carbon copy of TS 213, he needed more than one copy of the source pages that were to be transferred intact. This need was met by plundering the otherwise untouched top copies of TS 210 and TS 211. Top pages from the source TSS took their places in the top copy of TS 213 while carbon copies from the sources were inserted into the carbon copy of the new work.

In this way conspicuous gaps appeared in the top copies of TS 210 and TS 211, ─ and it is these gaps which von Wright mentions as being filled with photocopies. For all but three pages of TS 211 ─ which, according to the Nachlass reproductions available at the Wittgenstein Archives, are still missing ─ these photocopies have been taken from TS 213 and inserted into the earlier TSS as proxies, thus leaving the originals where Wittgenstein last placed them in TS 213.

If the Nachlass is to be restored at all, then this is probably the only defensible strategy. It is at least more defensible than the alternative, of returning the shared pages to their original homes and putting the Xerox copies instead into TS 213. What confuses matters, however, is that the Xerox copies were made, sometimes from the top copy of TS 213, which is rich in later annotation, sometimes from the carbon copy, and sometimes even from a transcription ─ which in all probability was made long after Wittgenstein's death, and therefore has no real claim to be regarded as part of the Nachlass.

3. Sources

This survey is based on the evidence of two facsimile sources: the Cornell edition (1968) and a more recent microfilm of the holdings at Trinity College Library⁴.

The Trinity microfilm (TMF) consists of 11 rolls of microfilm. All items involved in this study are to be found on rolls 1 and 2. Roll 1 contains TS 210, TS 211 pp. 1-686, and a carbon copy of TS 213 pp. 0-528. Roll 2 contains the remainder of TS 211 (with an overlap of two pages, hence pp. 685-771) and the ’definitive’ version of TS 213 (complete and, in its earlier parts, heavily annotated).

TS 210 is not contained in the Cornell edition. The version of TS 211 captured in Cornell (volumes 88A-C) is evidently the same as that on the TMF. Thus we have to contend with only one version of each of these items.

The case of TS 213 is more complex.

The Cornell paper edition of TS 213 fills four volumes, Cornell 89A-D. Volumes 89A to 89C reproduce pages 0-528 of the ’definitive’ version of this typescript, i.e. the same as is found on roll 2 of the Trinity microfilm. Cornell volume 89D, however, does not continue with the same version. Instead it reproduces a transcription of the last 240 pages (pp. 529-768).

So far, it has not been possible to establish who made this transcription⁵. Indeed, it seems to have passed largely unnoticed that it is one. And perhaps not surprisingly. The potential for confusion with the authentic version is considerable. For the transcriber of these pages was evidently at pains to reproduce the original in every relevant ─ as well as a fair amount of irrelevant ─ detail. For example, the vast majority of line- and pagebreaks occur in exactly the same places as in the original, and even the double pagination of the pages adopted from TSS 210 and 211 has been reproduced.

The evidence that the pages in question are indeed a transcription is nonetheless sufficient. Of relevance for the pages studied by this survey is the fact that the typefaces of the original and the transcription are different. This difference is most apparent from a comparison of the numerals⁶.

4. Method

The method of this survey has been to compare closely the different copies of the double-paginated pages in these items. Study of the originals would doubtless have made this work much easier. Even so, evidence for the distinctions is visible and fairly consistent in the sources available.

The most important aspects of comparison can be summarised as follows:

─ On carbon copies the type generally appears fuzzier than it does on the corresponding top copies.

─ Both top and carbon copies are occassionally annotated. Where this is the case there tends to be an easily distinguishable difference of tone between the typewriter pigment and the handwriting medium (ink or lead pencil). There are Xerox copies of annotated pages, but no Xerox copies with new (authentic) annotation (presumably all Xerox copies were made after Wittgenstein's death).

─ Both top and carbon copies can exhibit noticeable variations in pigment density resulting from the varying pressure with which the keys were struck. This is not the case with Xerox copies, which generally show less differentiation of pigment tone. On Xerox copies the pigment of type and handwriting usually appears equally dense. Alternatively, what appears grey on a top or carbon copy frequently seems granular on the Xerox, and there are often granular shadings away from the centre of the page due to uneven light exposure.

─ Xerox copies are generally recognisable from marks resulting from the mechanics of their production. Equally spaced along the top edge of most Xeroxed pages are a number of small "skid-marks", presumably made by the wheels that draw the paper through the machine.

─ Many of the Xerox copies exhibit narrow shadowy stripes down the vertical edges, possibly the result of the image frame being larger than the depicted page (i.e. the stripes are a glimpse of darker background).

In most cases these features have been sufficient to allow a fairly safe classification of the relevant pages according to the distinctions top/carbon copy, original/Xerox copy. All four permutations of these distinctions occur among the results (original top copy (O), carbon copy of original (C), Xerox copy of original top (OX), and Xerox copy of carbon copy of original (CX)).

5. Table: Summary of the findings

O = original
T = transcription
C = carbon copy of original
X = xerox copy
a, b, c ... = annotated