How to Make an Index

Before commencing his work, the indexer must think out the plan and the kind of index he is to produce; he will then consider how he is to draw out the references.

Whatever system is adopted, it is well to bear in mind that the indexer should obtain some knowledge of the book he is about to index before he sets to work. The following remarks by Lord Thring may be applied to other subjects than law:

"A complete knowledge of the whole law is required before he begins to make the index, for until he can look down on the entire field of law before him, he cannot possibly judge of the proper arrangement of the headings or of the relative importance of the various provisions."

During his work the indexer must constantly ask himself what it is for which the consulter is likely to seek. The author frequently uses periphrases to escape from the repetition of the same fact in the same form, but these periphrases will give little information when inserted as headings in an index; and it is in this point of selecting the best catchword that the good indexer will show his superiority over the commonplace worker.

This paramount characteristic of the good indexer is by no means an easy one to acquire. When the indexer is absorbed in the work upon which he is working, he takes for granted much with which the consulter coming fresh to the subject is not familiar. The want of this characteristic is most marked in the case of the bad indexer.

In printing references to the entries in an index it is important to make a distinction between the volume and the page; this is done best by printing the number of the volumes in Roman letters and the page in Arabic numerals. When, however, the volumes are numerous, the Roman letters become cumbersome, and mistakes are apt to occur, so that one is forced to use Arabic numerals; and in order to distinguish between volume and page, the numbers of the volumes must be printed in solid black type.

When a book is often reprinted in different forms it would be well to refer to chapters and paragraphs, so that the same index would do for all editions. The paragraphs in Dr. Jessopp's edition of North's Lives of the Norths are numbered, but they are not numbered throughout. The references are very confusing and require a key. Thus, P stands for Preface; F for Life of the Lord Keeper; D, Life of Dudley; J, Life of Dr. John; R, Autobiography of Roger, and also Notes; R L, Letters from Lady North; R I, Letters from Roger North; and S, Supplementary. In the Letters the references are to pages and not to paragraphs. With such a complicated system, one is tempted to leave the index severely alone. This is the more annoying in that the index is not a long one, and the pages might have been inserted without any great trouble.

Much confusion has been caused by reprinting an index for one edition in a later one without alteration. An instance may be given by citing the reprint of Whitelock's Memorials, published at the University Press, Oxford, in 1853. The original edition is in one volume folio (1682, reprinted 1732), and the new edition is in four volumes octavo. But to save expense the old index was printed to the new book. The difficulty was in part got over by giving the pages of the 1732 edition in the margin; but as may be imagined, it is a most troublesome business to find anything by this means. Moreover, the old index is not a good one, but thoroughly bad, with all the old misprints retained in the new edition. As a specimen of the extreme inaccuracy of the compilation, it may be mentioned that under one heading of thirty-four entries Mr. Edward Peacock detected seven blunders. Although Mr. Peacock had no statistics of the other entries, his experience led him to believe that if any heading were taken at random, about one in four of the entries would be found to be misprinted.

In the case of a large index it is necessary to take into consideration the greatly increased work connected with arrangement. The amount of this may be said to increase in geometrical rather than in arithmetical progression. When the indexer comes to the last page of a great book he rejoices to have finished his work; but he will find by experience, when he calculates the arrangement of his materials, that he has scarcely done more than half of what is before him.

If cards or separate slips are used, these will only need to be arranged for the press; but if sheets of paper have been, written upon, these will have to be cut up. There is little to be said about this, but it is worth giving the hint that much time is saved if shears or large scissors are used, so that the whole width of paper may be severed in two cuts.

In the case of a small index there is little difficulty with material, for it can be arranged at once into first letters, and when the table is cleared of the slips these can be placed in the pages of an ordinary book to keep them distinct, and can then be sorted in perfect alphabet and pasted down. In the case of a large index it will be necessary to place the slips in a safer place. Large envelopes are useful receptacles for first letters; and when the slips are placed in them, the indexer will feel at ease and sure that none will be lost.

It is well to go through the whole of the envelopes of first letters and sort the slips into second and third letters before the pasting is commenced, so that you may know that the order is correct, or make such alterations as are necessary before it is too late. The final perfect alphabetical arrangement can be made when the slips are placed on the table ready to be pasted.

The sorting of slips into alphabetical order seems a simple matter which scarcely needs any particular directions; still such have been made.

The late Mr. Charles F. Blackburn, who had had a considerable experience, gave some instruction for sorting slips in his Hints on Catalogue Titles (1884). He wrote:

"Having never seen in print any directions for putting titles into alphabetical order, I venture to describe the system I have been accustomed to use. First sort the entire heap into six heaps, which will lie before you thus:

Then take the heap A—D and sort it into its component letters, after which each letter can be brought into shape by use of the plan first applied to the whole alphabet. It is best to go on with the second process until you have the whole alphabet in separate letters, because if you brought A, for example, into its component parts and put them into alphabetical order, you might not impossibly find some A's among the later letters—one of the inevitable accidents of sorting quickly. With this hint or two the young cataloguer will easily find his way; and various devices for doing this or that more handily are sure to suggest themselves in the course of practice. The great thing is to be started."

The latter part of this extract is good advice, but I think it is a mistake to make two operations of the sorting in first letters, for it can be done quite easily in one.

The following suggestion made by Mr. Blackburn is a good one, and is likely to save the very possible mixture of some of the heaps:

"In my own practice I have got into a way of letting the slips fall on the table at an angle of forty-five degrees. Then, if the accumulation of titles should cause the heaps to slide, they will run into one another distinct, so that they can be separated instantly without sorting afresh."

I have never myself found any difficulty in sorting out into first letters at one time, and it soon becomes easy to place the slips in their proper heaps without any thought. Mr. F. B. Perkins, of the Boston Public Library, however, in his paper on "Book Indexes" gives some good directions which are worth quoting here:

"Next alphabet them by initial letters. This process is usually best done by using a diagram or imaginary frame of five rows of five letters each, on which to put the titles at this first handling. The following arrangement of printers' dashes will show what I mean. (The letters placed at the left hand of the first row and right hand of the last indicate well enough where the rest belong.)

A – – – – – U

B – – – – – V

C – – – – – W

D – – – – – X

E – – – – – YZ."20

When the alphabetical arrangement is completed so far as the indexer considers it necessary for his purpose, it is time to think of the pasting down of the slips. This can be done in several ways, and the operator will doubtless choose that which suits him best. As already remarked, men will always find out the way most agreeable to themselves, and it is unwise to insist on others following our way in preference to their own.

The human mind is capable of interesting itself in almost anything it may undertake; but indexing cannot be other than hard work, and it is unfair to make it harder by fixing unnecessary limits. The worker is always happier at his work if he is allowed to do it in his own way.

The first thing to settle is as to the paper upon which the index is to be pasted. A very large-sized paper is inconvenient, and foolscap or quarto is the best for constant handling,—all the pages should be of exactly the same size. Sometimes it is necessary to have a small margin, but generally the width of the paper used for the index should be followed. There is no greater mistake than to study economy in the use of paper for pasting on. Some persons have facilities for the use of wastepaper that has been printed on on one side, and, not having been used, is in good order and of equal size. Some persons cut up newspapers, but this is a practice not to be recommended, not only on account of the print, but because the paper is generally so abominably bad and tearable. If the wastepaper referred to above is not within reach, it is well to buy a good printing-paper, which can be cut into the size required. There are, however, many cheap papers already machine-cut into the size required, which can easily be obtained.

Some with the love of saving strong upon them cut up newspapers into lengths of about four inches wide, and paste the slips upon these, with the result that all the ragged ends give continual trouble, and are apt to be torn away. Of all savings, this is the most ill-advised.

Although the "copy" is to be printed from at once, and will soon become useless, it is a great comfort to have material that is convenient to handle while it is required. Some thought may also be given to the compositor, whose life will be made a burden to him if you send him "copy" with all the ends loose. It is also well to keep the pages as flat as possible, so that a heap of these do not wobble about, but keep together smooth and tidy.

Sometimes it may be desirable to paste only on half the paper, so as to have room for additional entries. If this is done, the side must be altered periodically, or the pages will slip about and give endless trouble.

When the index is in course of arrangement the greatest care must be taken that none of the slips are lost, for such a loss is almost irreparable—first because you do not know when a slip goes astray; and even if you do know of your loss it is almost impossible to remedy it, as you have no clue to the place from which the slip came.

There will always be anxiety to the indexer while his work is being cut up and sorted. A breeze from a window when a door is opened may blow some of his slips away. Too many of the slips should not be allowed on the table at one time, and the indexer will feel the greatest comfort when he knows that his slips are safely reposing in their several envelopes. All queries should also be kept in envelopes, and each envelope should be inscribed with a proper description of its contents. When the slips are pasted down they are safe—that is if they have been affixed securely to the paper.

Having made these general observations, we may now proceed to consider how to paste. It seems a very simple matter, that requires no directions; but even here a few remarks may not be out of place.

When your paper is ready in a pile of about fifty pages, each page numbered in its proper sequence, you can proceed to work. For the purpose of laying down slips on uniform pages at one time, paste is the only satisfactory material. Gum will only be used by the inexperienced. It cannot be used satisfactorily on large surfaces, like paste, and when it oozes up between the slips it is stickier and does more damage in fixing the pages together than paste does. You might as well fix paperhangings on your walls with gum.

As to paste, if you have a long job on hand it is better to have it made at home, of a good consistency, but not too thick. It ought to run freely from the brush. A good cook will make good paste, but if you are specially particular you can make it yourself. If you require it to last for any time, you must add a little alum; but when you have a big index before you, you will use a bowl of paste in an evening, and there is therefore no question as to keeping.

"Stickphast" is a very good material; it sticks well and keeps well, and it is an excellent adjunct to the writing-table, but it is not suitable for pasting down a long index. It is too dear, it is too thick, and it is too lumpy. If the paste is made at home, it need not be lumpy; and lumps, when you are pasting, are irritating to the last degree.

The paper and the paste being ready, with a fair-sized brush to spread the paste, we come to consider how best to proceed with the work in hand. You require a good-sized table,—a large board on tressels in an empty room is the best, but a dining-table will serve. At the extreme right of the table you place the batch of paper upon which you are about to paste, and then sort your slips in perfect order, ranging them in columns from right to left. The object of thus going backwards is to save you from passing over several columns as you take the slips off the table, and, instead, going straight on. You can push your batch of paper on as the various columns successively disappear. More slips should not be set out than you can paste at one sitting, as it is not well to leave the slips loose on the table. Of course, you can paste from the left side if you wish, and then the columns will range from left to right; but this is not so convenient for continued arrangement of the columns of slips as you require them.

There are more ways than one in placing the paste upon the paper; the most usual way is to paste down the two sides of the paper just the width of the slips, and some add a stroke down the middle. Another way is to put a plentiful supply of paste on a page or board, and then to place the back of each slip upon this. If you place your fingers on the two ends and press them towards the middle, the slip will be ready to be placed in its proper position, having taken up just sufficient paste. A still different plan is to paste the board or paper as in the previous case, and then place the face of the whole page on this. You then take it off, and, placing the dry side on the batch of paper, proceed to affix the slips to it. The advantage of the two last processes is that the paper is not so wet as in the first-mentioned plan, and in consequence the paper does not curl so much, but lies flatter. In the first place the sheets must be set out separately on the floor to dry, so that they may not stick together, but this is not so necessary in the two latter processes.

Some indexers strongly object to pasting. This was the case with Mr. E. H. Malcolm, who wrote thus to Notes and Queries:

"I long ago discovered the cause of imperfections in my own work. It was the 'cutting into slips' and 'laying down' processes. The fact is you cannot be sure of preserving the cuttings or slips, if very numerous; they are almost certain to get mixed or lost, or elude you somehow. My remedy is this. I now take cheap notepaper and write one entry only on each leaf. Having compiled my index thus from A to Z, I arrange my slips and manipulate them as I would a pack of cards, although shuffling only for the purpose of getting the arrangement of the letters right. Thus I save myself all the labour and trouble of pasting or laying down the slips in analytical order. I do not mind a little extra expenditure of paper by only entering one item on every slip, for I am compensated for the appearance of bulk by finding that I have secured order and arrangement free from the consequences of a finical arrangement of the slips and a dirty and tiresome labour of pasting down."21

As already pointed out in these pages, Mr. Malcolm is quite right respecting slips for a growing index; but when it comes to sending the "copy" to the printer the case is different. Here there is more safety in the pasted down slips, which are less likely to be lost than the loose ones even when numbered.

As you proceed in your work you may wish to know how far your index agrees with other indexes in its proportion of letters, and to calculate what proportion of the whole you have already done.

Some calculations as to the relative extent of the different letters have been made. Thus B is the largest letter in an index of proper names, but loses its pre-eminence in an index of subjects; and S takes high rank in both classes.

Mr. F. A. Curtis,22 of the Eagle Insurance Office, made in 1858 a calculation of the relative proportions of the different letters of the alphabet in respect to proper names. He described his object in a letter entitled, "On the Best Method of Constructing an Index." He wrote that, having had occasion to construct an index of the lives assured in the "Eagle" Company, he had drawn up a few observations upon the subject. "The requirements of an index and the proportions of its several parts are the two principal questions to be considered. Under the first head it may be observed that the index of a company upon a large scale should afford as much abstract information as possible. Those who refer to it do so with different views, for the objects of their inquiry must necessarily vary with their respective duties. It is therefore desirable that the index should be constructed with a view to provide for the wants of each person, so far, at least, as to enable him to obtain information in the most direct way; and it will be proper to insert in the index particulars some of which do not usually find a place in such a book. Let it be supposed that an individual signing his name 'J. Smith' inquires about the bonus, premium, or assignment, etc., of his policy, without stating either number, date, or amount. This is not an unusual case, and it will serve to illustrate my meaning by showing the nature of the difficulties which have to be encountered. J. may stand for John, James, Joseph, etc. There will probably be many of each kind in connection with the like surname, and it would be very difficult to discover, without a tedious investigation, to which policy J. Smith refers, unless the individuality of each person recorded in the index under that name be distinctly shown. The 'locality' of the assurance might be adopted as a mark of distinction; and we should in many instances be able to fix upon the right name by simply comparing the address of the writer with the place where the policy was effected."

This is a most valuable suggestion to all indexers. Many persons, to save trouble at the time, write initials instead of full Christian names. It should be a rule always to write these in full. When the index comes to be printed, the Christian names can be contracted if it is necessary to save space. The most important matter in the arrangement of an index is to avoid the confusion of two persons as one, and the possibility of making this blunder is greatly increased by the use of initials instead of full names. In the British Museum Catalogue it has been found necessary in many cases to add particulars to distinguish between men with the same names.

Mr. Curtis goes on to say:

"With regard to the second part of this subject—i.e. the proportions of the several parts of the index—I may observe that the most useful mode of division appears to me to be that which is adopted by many offices—namely, to classify the surname under its first letter, and to subdivide according to the first vowel thereafter, adopting the first subdivision for such names as 'Ash,' 'Epps,' etc., which have no succeeding vowel."

This, however, is a very unnatural arrangement, and has been, I believe, very generally given up. It is therefore unnecessary to refer further to Mr. Curtis's calculations of the proportions of the vowels in the subdivisions. Calculations can be made for the subdivision of the complete alphabet with a better result. Of course, in the case of initial vowels the following consonants have most to be considered, and in initial consonants the following vowels. Mr. Curtis's calculations respecting the first letters of surnames are of much value. He used the commercial lists of the Post Office London Directory, and compared them with Liverpool, Hull, Manchester, Sheffield, Birmingham, and Bristol directories, and with three lists of different assurance companies; and after making his calculations from nearly 233,000 surnames, he found the total average very similar in its result. Mr. William Davis made similar calculations from the Clergy List, which came out much the same. These he contributed to Notes and Queries,23 and subsequently he made a further calculation from French names.24

I have united these results in one table as follows:

It will be noticed that B is strongest in all three, and C is fairly equal. S is smaller in French names, but probably would be much larger in German names. H and W are also much smaller in French, while D and L are much larger. The preponderance of the latter letters is of course caused by the large number of names beginning with De and La.

Indexes are not confined to proper names, and therefore it is necessary to add some calculations as to the proportions of the several letters in indexes of subjects. The following table is formed from three large indexes, each different in character. I. represents Gough's Index to the Publications of the Parker Society, which may be taken as a very good standard index. The subjects are very varied, and there are no specially long headings; it also contains proper names as well as subjects. II. represents an index of subjects in Civil Engineering which contains a good number of large headings. III. represents the index to the Minutes of a public board, and also contains a considerable proportion of large headings. It will be seen that the numbers vary so considerably as to be of very little practical value. The percentages are, I think, interesting, but they show conclusively that indexes will vary so considerably that in order to obtain a satisfactory percentage a separate calculation will have to be made in each case. Large headings will vitiate any average; in fact, I have lately had to do with an index in which R was the largest letter, on account of such extensive headings as Railways and Roads.

One striking point in the averages is that B is found to be displaced from the pre-eminent position it occupies in the percentages of proper names.

When the whole index is pasted down it is not yet ready for the printer, as it will require to be marked for the instruction of the compositor. The printer will have general instructions as to the kind of type to be used and the plan to be adopted, but it will be necessary to mark out those words that are not to be repeated and to insert lines indicating repetition. There are also sure to be little alterations in wording, necessitated by the coming together of the slips, which could not be foreseen when the slips were first written out.

In a large work it is probable that your employers are importunate for "copy," and you will be urged to send this to the printer as you have it ready. If possible, it should be kept to the end, so that you may look over it as a whole, and so see that the same subjects are not in more places than one. You will probably have to make modifications in your plan as you go along, and this may cause difficulties which you will now be able to set right.