AN UNUSUAL METHOD OF MAKING A BOOK

A SYMPOSIUM

Oliver Stallybrass submitted an article with this title. It was considered suitable as a basis for a symposium to which a number of members were invited to contribute. Mr. Stallybrass's article—with an —is printed first.

OLIVER STALLYBRASS The first step was to allocate a precise Without having made an exhaustive search of the alphabet to each recto or right- of the literature, I have the impression that hand page. This I did by counting the num book indexes are almost invariably compiled ber of recto pages for each letter, dividing by means of cards or slips of paper; and that the index of Chambers's encyclopaedia into theuse of a thumb-indexed notebook, which the appropriate number of roughly equal is not mentioned in Collison's standard sections for each corresponding letter, and work1, would strike most indexers as heretical, inscribing accordingly the top right-hand indeed as utterly and obviously retrograde. corner of each recto in the notebook—keep Needless to say, this would be truein the ing the captions mutually exclusive but never longer than three initial letters, and ignoring case of any index which was going to remain in that form, while continuing to grow; but the more improbable combinations; thus the as a stepping-stone to a typed, and ultimately nine pages for B were B-Bap, Bar-Baz printed, index the thumb-indexed notebook (ignoring Baq), Bea-Bem, Ben-Bez, Bi-Bl seems to me, after a single experiment in its (there being no references to the Bhagavad- use, to have, for many types of book, over Gita), Bo, Bra-Bri, Bro-Bukand Bul-By. whelming advantages. The index of Chambers's encyclopaedia The book in question, which I indexed for (chosen simply as being the largest in my Seeker & Warburg, was a 270-page translated possession) is not of course tailor-made to biography, by Philippe Jullian, of the fin-de- fit an index consisting largely of French siecle writer, wit, dandy and aesthete Robert names; Painter's Marcel Proust, which de Montesquiou (the original of Proust's shares so many dramatis personal with the Baron de Charlus). The French edition had Montesquiou biography, might have spread no index, but my allowance of ten double- the load a little more evenly from page to column pages, together with my preliminary page. A framework, however, modelled on reading, underlining and annotating, sug a large general index can serve again and gested that a 200-page, 7in. x 4£in., stiff- again—and has indeed already served backed notebook costing 4/3d. would be again for Pasternak's Letters to Georgian adequate; and so it proved. (I would, as it friends—thus reducing the time needed for turned out, have needed about 22/- worth initial preparation from about an hour on of 5in. x 3in. cards.) this occasion to five or ten minutes each subsequent time. In any case, only one For the prime and tremendous advantage recto (Corn-Con) remained virgin; and only of the notebook over the card index is in 10 per cent of the versos were brought sheer mechanical speed of operation, partic into requisition—the indexer's equivalent ularly at the entry stage. Moreover, whereas of those authors' * scriggles' which, in card indexes get more unwieldy the larger E. M. Forster's words, 'surge up from the they grow, notebook entries take progres margin, they extend tentacles, they inter sively less time as more and more headings breed'2—because no more interpolations are written in, and even the need for two were otherwise possible. The need for some seconds' thought on the placing of a head emergency parking ground, as it were, was ing is eliminated. Even with plenty of of course one of the two reasons why I began guide-cards, it takes astonishingly much by using only the recto pages; the other was longer to find the right card, extract it, insert that it is considerably easier, mechanically, a marker (or partly pull out the next card when thumbing to and fro, to have to write to serve the same purpose), write the entry, on rectos only. and wonder perhaps whether to keep the card out for possible further use soon (a Within each recto I divided the space by frequent minor dilemma) before re-filing it guess-work; and whether by luck or good and removing the marker, than it does to management there was nearly always room flick to the page and write the entry, often— where I needed it. In placing each new and one soon learns the short cuts, e.g. that heading I took into account, of course, not Huysmans is one page back from I—almost only the number of further headings likely in one motion. The notebook is also quicker to precede or follow it immediately, but the than the file of cards, for obvious reasons, likely number of entries under any heading. at the typing stage. This placed a higher premium than usual on mental notes made during the reading Other advantages of the notebook are compactness, portability, and the absence of stage—e.g. ' remember Desbordes-Valmore such perils as upsetting the file or—more and Deslandes, close together and each with serious—mislaying cards. These advantages quite a few entries'—and on an ability are particularly important when it comes to (which might surprise my correspondents) the checking of names that one's personal to write small but clearly when the occasion library has failed to unearth, and more than demands. In not one case was it necessary compensate for the danger of jumping a to number a group of misplaced headings heading or two when typing from a note before typing. book; the latter error will come to light when the typing is checked against the To anticipate an obvious question: I re original (at which stage a helper is even sorted to cards for entries under Montesquiou more essential than with cards), whereas a himself and for cognominal members of his card or wodge of cards left in the Reading family, which together accounted for a fifth Room of the British Museum or London of the index. The use of different-coloured Library may well sink without trace. cards, however (red for Montesquiou's rela tives, white for his characteristics, yellow for The notebook method has its limitations, his books), combined with modification of of course, and is clearly better suited to an the one-card-one-heading principle (thus index consisting predominantly of names one white card covered, just as a page in the than to one where headings are more liable notebook might have done, alphabetically- to be changed or combined at a late stage, arranged characteristics from ' intelligence ' where sub-headings and sub-sub-headings to 'pugnacity'), made the finding of headings are going to abound, or where for any and sub-headings on the few (23) cards reason a considerable amount of editing is almost as quick as in the notebook itself. likely to be necessary. (For the next book I indexed, a symposium on Voting in cities Incapable of copying a three-figure number published by Macmillan, it was clear that correctly should be making an index at all. many headings would need to be sub-divided at least once, and so I reverted unhesitatingly Addendum to cards.) In general it has the disadvantages At the proof stage I have a sad little post of inflexibility—which means that the in- script to insert: the publishers of Robert dexer should know what he is doing before de Montesquieu, I have just learned, returned he starts. (One possible way of getting the the press copy of the page proofs without best of both worlds might be to use a loose- waiting for my (punctually delivered) leaf thumb-index notebook; though I guess index—which was accompanied by a list of that the loss of mechanical ease of reference some fifty names given wrongly or in would outweigh the gain in flexibility.) consistently in the text There are thus One fundamental objection to the note numerous discrepancies between the uncor- book method must be anticipated and rected text and the index. This, of course, answered: how, since one cannot shuffle has nothing to do with the subject of my and re-shuffle into page order of first article; but having drawn attention to what unchecked entry, and back into alpha could be mistaken (since not all the names betical order, does one carry out the page- concerned are household words) for a by-page checking of entries against the text thoroughly careless index, I feel constrained which Collison3 takes for granted? At the to add this cautionary tale. risk of scandalising fellow-indexers and scaring off publishers and authors, I will 1 Robert L. Collison, Indexes and indexing, 2nd boldly confess that this incredibly cumber edition, London, Benn, 1959. some (and therefore costly) procedure is one 2 The Library, series 5, vol. 13 (1958), p. 142. that I have always skipped: nobody who is 3 Op. cit, p. 87.

From M. D. ANDERSON I have used the notebook method several times for short indexes, and find it satisfac Mr. Stallybrass's method of making an tory, though not perhaps as time-saving as index in a notebook resembles in many ways Mr. Stallybrass would suggest. It would also Dr. Holmstrom's * shingled sheets' method, be applicable to longer indexes involving described in The Indexer, Vol. 2, pp. 26-30, perhaps only names, or other very short and Vol. 4, p. 129. Dr. Holmstrom uses, not entries with few or no sub-headings. Where a notebook, but sheets of paper stapled to there are only a few long entries, as in Mr. gether in bunches of ten, with each succes Stallybrass's index to the 270-page bio sive sheet projecting a quarter of an inch graphy, his use of index cards for these alone lower than the one on top; the caption for extends the scope of the method. But when each page is written on the bottom right- most of the entries in an index have sub hand corner. Mr. Stallybrass takes the headings, and when many entries overflow groups of letters for the captions from an on to two, three, four, or even more cards, encyclopaedia, Dr. Holmstrom from an exist and much editing is required, then the note ing large index on the same subject, or from book method is not appropriate. a dictionary of the subject. Both methods have the same advantages: (1) speed of Mr. Stallybrass himself acknowledges the operation; (2) portability; (3) economy of limitations of his method; within these lim material; (4) no risk of losing an entry. itations it can have a useful place in indexing Both provide for reserve space—on the verso procedure, and both Mr. Stallybrass's and pages of the notebook, or the left hand Dr. Holmstrom's versions should find their halves of the sheets of paper. way into books on the subject. I agree with Mr. Stallybrass that Mr. compensates for this. I use cards for an in Collison's instruction to check back all index dex where it is evident that there are to be entries against the text is a counsel of a very great number of subject entries with perfection, and too time-consuming to be out many page entries to each. attempted in ordinary circumstances. But I In the matter of sub-entries, I usually find find that the editing of a complex index does that by the time one starts the actual enter involve the incidental checking of many page ing, one has a pretty good idea of those numbers, for in the process of sorting the names or subjects which are going to require numbers of a long entry under suitable sub these, and I provide separate foolscap sheets headings, it is often necessary to refer back for this purpose. If I have made a mistake, to the text to be sure that the grouping is as and find afterwards that there are entries logical as possible. which need sub-dividing and which I had not allowed for, it is a simple matter to look From DELIGHT ANSLEY, an American these up again (the entries are never very member numerous) and divide them up as necessary. The best way to make an index is the way This also serves as a check-up. that the indexer finds most convenient. If Incidentally, there is one point in Mr. the result is a good index, nobody will care Stallybrass's method about which I am not whether it was made with a notebook, a quite clear. He says that he sub-divides each computer, or knots in a piece of string. letter roughly according to Chambers's en A notebook might be practical for indexing cyclopaedia (B, for instance, having nine a short biography such as the author des sub-divisions). Does he mean that, on com cribes. He understands that his method has ing across a B entry, he first thumbs to the limitations in dealing with larger books or B section of the notebook and then turns more complicated subject entries. An obvious over to the correct page? Or has he in some difficulty is that unless one could predict the way divided the graduated part of the sheet amount of space which would be needed for so that he can thumb directly on to the every section of the alphabet, it would be subdivision of the letter in question? My hard to keep the alphabetization correct, method, because it only entails dividing each especially with sub-entries. letter into two parts, enables me to do this, thus saving further time in the first instance. For his last sentence the author deserves any medal that the Society of Indexers can I should like also to bear out Mr. give him. 'What oft was thought, but Stallybrass's points about the convenience of ne'er so well expressed.' compactness and, in particular, portability: also his argument against a loose-leaf thumb- From IVAN BUTLER indexed notebook. I tried this once and found that the loss on the swings—including I was very interested in Mr. Stallybrass's temper at the thing's clumsiness—more than article, as I have used a method based on a outweighed any roundabout gain. thumb-indexed notebook myself for well over a hundred indexes of varying sizes, and [Mr. Butler, in a to his letter have found no reason to change, except in accompanying the above, says: The article special circumstances, to using cards or slips. brought to mind one point. I heard the My own method is different in detail from other day of an indexer who entered each that of Mr. Stallybrass, entailing the use of sub- or sub-sub-entry on a separate card, a foolscap size thumb-indexed book, and or slip. Is this usual? It would save a good necessitating one extra item of procedure in deal of time in hunting out particular cards, respect of alphabetizing. Even so, I have I imagine, but the mass of material at the found, as he has, that the time saved in the end (and the expense, if cards are used) actual writing down of entries more than must be pretty formidable 1—Ed.] From G. V. CAREY not in Indexes and indexing, but in his other book on the subject (R. L. Collison, In I find nothing 'heretical' in the method dexing books. Benn, 1962, p. 17). However, propounded by Mr. Stallybrass. The chief given all the reservations and qualifications emotion that it has stirred in me is exasper that Mr. Stallybrass mentions, the method is ation at my own dull-wittedness in never not perhaps without its uses. It would not having thought of trying it myself. I can be suitable for an index which was either see, as he does, that there must necessarily extensive or meant to cumulate; it would not be reservations: in particular, that the book do for an index which had a great many to be indexed must be neither very long nor sub-headings; it would not be very practical very recondite; also that the indexer should in cases where a great deal of editing has to have had time to read the whole book first— be done. Mr. Stallybrass very disarmingly desirable of course in all circumstances, but covers all these points. Two further criti by no means always practicable. cisms which may be brought and which he I can see also, on the other hand, the does not mention are (1) revision of the possibility of an advantage in the notebook index would not be easy, e.g. for revised method additional to the advantages men editions of the book, and (2) it is not clear tioned by Mr. Stallybrass. By pure chance, how one copes with what librarians call it suggests itself to me from my immediate 'tracings'—those internal notes and refer circumstances — temporarily homeless and ences that the indexer makes to remind him wintering in a hotel. Many an indexer no of related headings, e.g. if a concept is in doubt finds himself, for one reason or an dexed under two separate headings, then the other, separated from his familiar facili indexer often needs to remind himself of this ties and * apparatus' for, perhaps, some fact. weeks on end, and may feel himself reluc tantly obliged, in consequence, to refuse I must confess that Mr. Stallybrass's ad work that he would otherwise have wel mission about his failure to check his index comed. It is possible surely that, for some against the text was mildly shocking. With types of assignment at any rate, a thumb- all the meticulousness and accuracy in the indexed notebook might obviate such refusal. world I am sure that the checking operation My concluding reaction is: 'Would that is still required. What I have often done in I could myself be granted the opportunity of order to avoid filing and re-filing is to check putting the Stallybrass Method to the test.' the text from alphabetically filed slips. (I To which the obvious reaction of any reader like to keep entries in alphabetical order who knows me must be: 'He's old enough from the beginning and therefore I usually to know better.' finish the actual indexing with the slips in this order.) This means darting about from page to page of the text but in many cases this takes no longer than repeated filing. From G. DIXON On the whole then, bearing in mind its limitations, the method is one which the My own experience of thumb-indexed notebooks for indexing has been limited to indexer ought not to ignore completely, but one or two very small index/catalogues for I do think that the qualities of flexibility and such things as gramophone records. On the hospitality are more important in indexing face of it the method does appear to be too and cataloguing than Mr. Stallybrass will constricting and inflexible and, I would allow. After all, he admits that he had to imagine, is usually so regarded. Incidentally, resort to cards himself for one-fifth of his Mr. Collison does condemn this method— index!

8 From NEIL R. FISK works out at one second per letter, and is, I No I At all events, not for indexing tech suppose, just barely possible including turn nical matter or collections of abstracts, and ing over the pages, but this could not include it is such work only that I have been con thumb-indexing 26 pages. Five minutes cerned with; probably, also, not for any would be utterly impossible. Much better large book. than 'guesswork' to divide the space is In the index to a technical book the entries the alphabet: top quarter Boa-Bof, next as well as the headings must be in alphabeti Bog-Bol, Bom-Bos, and Bot-Boz. cal order. This restriction—not frequently I note without comment that even in the imposed upon their work by the indexers example, a smallish index (internal evidence of biographies, memoirs, histories, etc.— suggests that the Montesquiou index con makes the notebook method unsuitable. tained fewer than 1,300 lines set in 7-point It is almost impossible to ' jump a heading type) ' largely of French names' to a by no or two' when typing from cards, and ' one means big book, Mr. Stallybrass ' resorted to soon learns the short cuts' with cards, too. cards ... for a fifth of the index'. Far from obvious to me is it that typing The claim that' the tremendous advantage from a notebook is quicker than from cards. [over cards] is in sheer mechanical speed' All the cards will be in perfect order, but can, I suggest, not be supported except by 10 per cent, of the rectos of the notebook test, and I would back myself with blank will have various entries on versos to be cards, against Mr. Stallybrass with an un watched out for. How much quicker, and touched notebook, to finish compiling first. why? If clearly handwritten ' copy' were accept Some printers can still be persuaded to able I should take 30 per cent to 60 per cent, take clearly handwritten index 'copy', and longer, but Mr. Stallybrass must add 100 cards can be mounted to provide c pages', per cent, for typing and checking typescript. eliminating the typewriter. No printer I entirely agree about the ' incredibly . . . would set type from a handwritten book costly procedure' commonly employed for with 10 per cent, of the pages written on checking entries. Only a spot-check need both sides. be run, but this is all the more necessary if The risk of mislaying cards is admitted. a typescript intervenes, because, although we The cardinal rule is therefore tripartite: not may be infallible copiers, very few people more than one card out at a time, never are infallible proof-readers. more than two out, never go away from the 'Nobody who is incapable of copying a drawer even 'for a minute* leaving even three-figure number correctly should be one card out. making an index at all.' Again I agree with As to cost, 'cards' cut to 5 x 3 inches Mr. Stallybrass, but I would add that no from typing paper by a printer would cost body who leaves in the British Museum ' a less than one-third of Mr. Stallybrass's 22/- wodge of cards ' containing the results of his and abolish at one stroke his complaint professional work is fit to be out by himself. about bulk. Plenty of 'used' paper, un marked on one side, is thrown away every day in offices: this can be guillotined to From GEOFFREY C. JONES supply an unlimited number of excellent Some years ago I used this method of in ' cards' without charge, or even 1 per cent, dexing successfully on several small indexes. of waste, let alone 45 per cent. However, as soon as I tried it on a larger ' Five or ten minutes' for the initial pre index, I was less pleased with the results. paration of the notebook after once using The use of a large book meant much time the' framework * seems impossibly optimistic. spent in breaking down letters to sub Lettering 100 rectos each with (average) six divisions and thumbing through many pages. letters (counting the hyphen) in 600 seconds Secondly I found it difficult to estimate in advance sufficiently accurately which were copy to be typed, thus involving extra time going to be the long sections and just how and/or expense, although it is true that, if long they would be. Over-estimation re the typing is done at home, some of the sulted in wasted space and too many pages editing can also be carried out in the pro to thumb through and under-estimation in a cess. horrible muddle needing to be re-written on a separate sheet of paper. From A. B. LYONS Finally, probably because I had a run of larger indexes, I discarded the system and As to the 'method' revealed, it seems to have not used it since. I would not now go me to demonstrate what I have always felt— back to it simply because I prefer to use a that indexers should be prepared at any uniform system applicable to any book, time to use their respective loaves and should irrespective of its size or complexity. How not consider themselves bound by text books ever, there is no doubt that the method can on indexing or by any conventions. be successful for a short index. How short I have not myself ever used Mr. Stallybrass's probably depends on the indexer's patience, method but I should like to tell you of one temperament, and the size and neatness of I adopted during the war. I was engaged on his handwriting. indexing the Annual digest of public inter national law cases and I was travelling by train between London and the north of From G. NORMAN KNIGHT England (in the blackout, may I add). I took with me the galley proofs and tore off the six I do not think that there is anything to eight inches of blank paper at the foot of very new or 'unusual' about Mr. Oliver each. I clamped these pieces together with Stallybrass's method of indexing by means a bull-dog clip and at the top right-hand of a thumb-indexed notebook. I used note corner of each I wrote the letters of the books for this purpose in the case of some alphabet. I knew where the 'heavy' of my earliest indexes—over forty-five years entries—' diplomatic immunity', ' treaties', ago. Nor can there be any objection to this and so on—would fall. Subject to that, I method for a comparatively small index, divided each page into quarters or sixths, especially after Mr. Stallybrass's careful and marking them Ba, Be, Bi, Bo, Bu, By, etc. elaborate system of preparation. But for one And then I set to work. that is analytically compiled in depth and which will occupy forty or more columns it When the job was finished, I handed the might prove quite impracticable. sheets to an expert secretary whom I had married a few years previously, and, with the While it may not be referred to in Indexes aid of my * squiggless and one or two of her and indexing, Mr. Collison does devote a own, she very soon reduced the whole thing longish paragraph to thumb-indexed note to a beautiful set of typed pages ready for books on page 17 of his later Indexing hooks the printer. (1962). He does not recommend them, ex No checking back of entries against the cept for a very small index or for preliminary text, I am afraid. But then, I don't think work—especially while travelling—on larger it was—or should be—really needed. indexes, but he adds: ' Some highly experi enced indexers have made a success of this From ROBERT J. PALMER, an American method'. member My own preference for cards is due to the fact that, if carefully and legibly inscribed I once started to make an index in the and punctiliously edited, they can be used way suggested by Mr. Stallybrass, did a as ' copy * for the printer, whereas the note couple of chapters, but then decided to finish book index will in every case require a fair on 5 x 3 cards. Perhaps this was not a fair

10 trial of the method, but indexing with cards Also, in order to work the cards speedily seemed to fit my temperament and working and efficiently it is necessary that the surface habits better. on which the cards rest be not too slippery. However, there is one situation in which I The open metal trays that are sold com would choose the method of Mr. Stallybrass. mercially for this purpose are generally too That would be if I had to compile an index highly polished, and the cards dance around while travelling and planned to work in air on the bottom like proverbial donkeys on planes, trains, or hotel rooms. ice. The surface on which the cards rest But in such a case I would prefer a small should be somewhat rough. A wooden or loose-leaf notebook to a bound notebook. cardboard box is best. One could thumb through a loose-leaf note book just as quickly, and the possibility of in serting blank sheets would ensure flexibility. From Wg. Cdr. R. F. PEMBERTON It would even be possible to create a rather I have read Mr. Oliver Stallybrass*s article, complex index. For example, in indexing a 'An Unusual Method of Making a Book biography in which one wanted to make a Index', with much interest. He makes out rather detailed analysis of the biographee, a very good case for his system; and he one could prepare loose-leaf sheets for the should, in my opinion, be able easily to sell biographee with such sub-headings as refrigerators to Esquimaux (particularly as, * childhood of', * education of', * works of', I believe, they use the things to keep their etc. Sub-subs could be entered directly on food warm). to such sheets. I have heard of the system before—from Incidentally, Mr. Collison mentions the an author who indexed one of his own loose-leaf notebook on page 41 of his books, a historical work. Even so, I do not Indexes and indexing. feel inclined to try it myself. I use slips of Under less itinerant conditions, the use of fairly stiff paper (which are better than 5x3 cards would be the method of choice cards because they are more flexible and do in the ' present state of the art \ However, not stick together), and I am perfectly happy the secret of indexing at speed with with them. The great merit of the slip (or 5x3 cards is not to stuff the box too full. card) system is that it is infinitely flexible Mr. Stallybrass complains of the difficulty of and adaptable; and a final run through finding cards in a file, of extracting them, before typing ensures that the slips or cards and of inserting markers. The point is that are in the right (alphabetical) order. An the indexer should always leave a fair other merit is the fact that some printers amount of play in the box of cards with will accept typed slips or cards instead of which he is working. A box designed to typed quarto sheets; I had this experience hold a thousand cards should never in the with two quite large indexes recently. working stage contain more than six or seven hundred, even if it becomes necessary for My slips cost about £1 10s. Od. for 2,000 the indexer to start a new box. When cards (including purchase tax and postage); and I are relatively loose, the indexer can thumb use most of them four times each (top and rapidly through them, and the place from bottom of front and back), using different which he takes out a card automatically re coloured inks to avoid confusion. Only mains open. There is no need to insert a those slips which carry long multiple entries marker. One can work speedily and effici (with sub-headings) are not available for ently with 1,800 cards in three boxes, but it top and bottom use; but these have clean is impossible to do this with 1,800 cards in backs which can usually serve twice. two boxes. On the other hand, of course, To revert to the Stallybrass System, while the cards should not be so loose that they I appreciate its advantages for small and fall down under one another. simple indexes, I feel that it would not be

11 suitable for a large or complicated one. The With Mr. Stallybrass's last paragraph one slip or card system works very well in all cannot but agree. If the Society of Indexers circumstances and is the only one suitable should ever have occasion to call upon its for large and complicated indexes. As I members to work to rule, no doubt observ said, it is infinitely flexible and adaptable. ance of the rule of page-by-page checking And Mr. Stallybrass himself rather gives the of entries against the text would be enough game away by admitting that he reverts to of itself to cause the requisite frustration and cards or slips for the principal and * cog- delay. Meanwhile, like many of the rail- nominal ' entries in indexes of biographical waymen's rules, it would seem in normal works, which introduces another complica circumstances best to keep it in the rule- tion: the use of two systems for one index. book and exercise a blind eye. Incidentally he might use a small supple mentary notebook for these entries. From JOHN L. THORNTON

The method described is probably one of the most primitive employed, and is still From F. H. C TATHAM used by some indexers, chiefly amateur/ I do not feel that Mr. Stallybrass need fear author/indexers. I believe it was mentioned denunciation for heresy, since I know at by Sir Steven Runciman in a talk to the least one experienced indexer who adopts a Society in 1958 or 1959 as being the method similar method of making a book index, also he employed. It can be useful in compact, based on a notebook, and no doubt there are straightforward indexes, but can seldom others. In any case, it seems to me that if save time in the long run. Marking up he finds that it works well with him, he the pages with letters of the alphabet, would be wise to stick to it. It is probably scrambling through these for each entry, and only mental laziness and conservatism which finally typing them, must take more time prevent me from experimenting with the than writing each entry on a slip, without method myself. attempting to find headings requiring added entries, but leaving this for the editing after I think however that Mr. Stallybrass is final alphabetization, and then submitting probably wrong, when he compares the the slips to the printer. All good printers speed of operation with notebooks and card accept legibly written slips, adequately indexes, in assuming that those who use marked and numbered. This saves time taken cards normally take out an individual card in typing them out and then checking the from the file, make an entry and put it back result. Mistakes are niinimised, the work is again. One normally adopts some such expedited and the cost to the publisher/ method as the following: look through the author is lessened, because the cost of typing or section of the book which one should be extra to the indexing fee. intends to index that evening. Note the I agree with Mr. Stallybrass that no worth entries which are likely to arise, extract them while indexer should find it necessary to from the file before starting to index and check every entry page-by-page. Further keep them on the desk in front of one. At more, few indexers, particularly in highly- the end of the evening sort these cards back specialised subjects, find it possible to read and refile them. This method avoids the through the text of a book before attempting constant pulling out and putting back of the to index it—another useless rule so often same card. Very occasionally one writes out laid down I a new card and finds that it duplicates one already in the index. But the time saved The chance of losing cards and slips by makes this risk of occasional time wasted removing them from the sequence is com worth chancing. pletely avoided if queries are copied on to

12 separate sheets. But carelessness should be he always be so lucky; I am sure we would minimal among indexers, whose work de like to hear how he makes out in future mands constant concentration and the indexes. patience of angels. Indexers must number among their ranks a large number of fallen angels, and if I were to be numbered among them, so far it would not be for mislaid A. R. HEWTTT, a Vice-President of the slips I Society, writes after reading the proofs of the foregoing contributions:

I have read Mr. Stallybrass's Unusual From R. WELLSTOOD, an American Method with interest. I agree with those of member the other commentators who suggest that if he finds the use of a notebook the best My first comment on reading the article method for his purpose so much the better by Oliver Stallybrass is, in my experience, but I cannot see that it saves much time in that very few indexers have an opportunity the long run. This, and other short cuts to read a whole book before they have to (e.g. typing on gummed slips and stick prepare the index. Normally the indexer is deluged with galley proofs and the printer ing down afterwards) have little to commend wants the index back yesterday or even them over the 'old fashioned' method of sooner. Rarely does one get a complete set writing slips or cards and alphabetizing them of proofs, thus having an opportunity to read at the end and not necessarily progressively the whole book before starting the index. when working under pressure. My wife (who is an expert assistant sorter) and I As far as having to pull out cards continu must have developed some kind of second ally is concerned, as one goes along one gets sight, as it were, in rapid alphabetizing be to know the ones that are most likely to be cause neither of us find it the time consum in greatest demand, and these are usually ing and difficult task that some others seem kept separate and close to hand so that the to do; there is, I think a tendency to make complete index does not have to be searched for one particular card. heavy weather of it. Mr. Stallybrass's method might be of some While working as a librarian I inherited use in the case of simple indexing, but it a master book index in loose-leaf form. This would be quite impossible in the specialist was very handy for quick reference, but the field, especially where many sub- and sub- amount of re-typing that had to be done sub-headings are essential. when a number of new books came in was hardly worth the trouble. I really felt the He, and some of the others, mention sub work after a trip to England and the pur sequent typing of the index and I have chase of 100-odd new books to fill in gaps always wondered why indexers allow them in the library. selves to be bamboozled into doing this chore, usually without additional reward. A notebook would be fine for, as the In some forty years of indexing (a period author says, someone who remembers all that has to be indexed, but certainly would during which I have indexed some 120 or never work from scratch. more works) I have never agreed to do more than provide written slips. Further, I would Also, as Mr. Stallybrass says, he was not accept a commission which required a lucky and everything came out right. May typed index.

13