Chapter V: Development

Costume, that sure guide of the historian and the antiquary, is perhaps nowhere more discernible than in literature, not merely in the dress of language and expression, but also in the visible exponents of that dress--writing and printing. Thus, a manuscript or a printed book may, by the character of its writing or printing alone, be ascribed to a determinate era. In other words, a careful investigation of the mode of construction will, in most cases, enable us to determine the approximate age of any book, from the early manuscript to the machine-printed volume of the present day.

In tracing the early development of printing, we are able to note those successive deviations from the form of its parent, Caligraphy, which were necessitated by the peculiarities of the new art. Commencing simply as a substitute for manuscript, it was naturally a close imitation thereof, and hence the first printers laboured under many inconveniences, which were shaken off as the capabilities of the new discovery became better understood. These changes often afford the only satisfactory evidence of the place and date of printing, as well as of the printer’s name. We propose, therefore, as an aid to chronological arrangement, to notice the points of similarity between the earliest printed books and manuscripts, especially with reference to the productions of Colard Mansion and William Caxton, and then to trace the novelties, purely typographical, introduced by the printers.

1. There was a selection of material. The scribe naturally wrote his choicest productions on fine vellum, carefully sorted in order to secure evenness in tone and quality; and with the same idea the early printers sorted out their paper before beginning to print. This is frequently seen when two or three copies of the same book are compared together. One is found to be printed entirely on thick, while another is wholly on thin paper--one has no defects, whereas another is made up of what the modern stationer calls “outsides.” The two copies of Caxton’s “Knyght of the Toure” preserved in the British Museum present a remarkable instance of this plan of selection.

2. It was a common practice with the scribes, when employing paper for their books, to use parchment for the inmost sheet of every section. The object of this was to give a firm hold to the thread of the binder, and thus strengthen the volume, but the alternation of paper and parchment did not present a pleasing appearance to the eye. Caxton adopted a modification of this plan, and instead thereof pasted a strip of vellum down the centre of the section. In books which have had the good fortune to escape the modern bookbinder, the observer may still see either the slips themselves or their traces in the brown stains left by the paste.

3. When commencing a book, the scribes had a custom of passing over the first leaf, and beginning on the third page, probably with the intention of protecting the first page from soiling during the execution and binding of the work. This practice was followed in the early works which issued from the presses of Flanders and of England, but unfortunately, in most of these books, on which an expensive modern binding has been placed, the blank leaf has been rejected as too coarse for a flyleaf, thus causing many volumes, although really perfect as regards the print, to be described by bibliographers as wanting the title-page.

4. The scribe necessarily wrote but one page at a time, and, curiously enough, in this the early printers also assimilated their practice. Whether from want of sufficient type to set up the requisite number of pages, or from the small size of the platen of the early presses, there is certain evidence of the first books from Caxton’s press having been printed page by page. Thus, in all the books printed with type No. 1, instances are found of pages on the same side of the sheet being out of parallel, which could not occur if two pages were printed together. A positive proof of the separate printing of the pages may, however, be seen in a copy of “The Recuyell of the Histories of Troye,” in the Bodleian Library; for there the ninth recto of the third quinternion has never been printed at all, while the complimentary page, which falls on the same side of the sheet, has been properly printed. A variation in the colour of the ink, though often very noticeable, is not a sure proof that the two pages so differing were printed separately, as that may have occurred through imperfect inking.

5. Many bibliographers, neglecting the study of manuscripts, and confining their examination of early books to the products of the printing press, have written and argued as if “signatures” were an invention of printers. This is an erroneous idea. It was as necessary for the scribe to mark the sequence of the sheets which he wrote as for the typographer to mark the order of those which he printed; because when the sheets, whether manuscript or printed, had to be bound, it was an absolute necessity for the binder to have every sheet signed, for the signatures were his only guide in the collation of the volume. There would seem to have been, for a long time, an antipathy to these useful little signposts, which, being needed only so long as the book remained unbound, were placed by the scribe as near as possible to the bottom of the leaf, that they might disappear under the plough of the binder. This is what has happened in the great majority of cases, but in every instance of the manuscript being preserved uncut they may still be seen.

It is interesting to notice the manner in which the early printers adopted and afterwards modified this custom of the scribes. As it was very inconvenient for them to print signatures of one or two letters away from the solid page, at the extreme margin of the sheet, and as the idea of disfiguring the text by making them a part of it was objectionable, they continued the old practice for some time, and actually signed every sheet by hand with pen and ink after it was printed. The uncut copy of “The Recuyell,” at Windsor Castle, is an example of a book with manuscript signatures at the extreme foot of every sheet. After some time, however, the prejudice was overcome, and the signatures were printed close up to the bottom line of the page. They were first introduced at Cologne in 1472 and adopted by Caxton in 1480.

6. The upper portion of the first written leaf of a manuscript was frequently left blank, for an illustration by the vignette-painter. Space was also left at the beginning of every chapter, and sometimes of every sentence, for an illuminated initial. For many years the early printers likewise followed this plan, every book they issued requiring the hand of the illuminator to complete it. This illumination was a distinct branch of trade, and the workmen employed in it did nothing but paint in the initials and paragraph marks. Through carelessness or ignorance a wrong initial was occasionally painted in, but as far as possible to prevent this, both scribes and printers inserted a small letter as a guide, which was usually covered over by the coloured capital.

7. When transcribing a book, it was seldom thought a matter of any importance to add the date of transcription and the writer’s name, though occasional instances of this are found. It was probably a like feeling which made the early printers follow a practice which has caused the modern bibliographer much doubt on many chronological points of the greatest interest. So needless was it thought to inform the reader when, where, or by whom a book was printed, that out of twenty-one works known to have issued from the press of Colard Mansion at Bruges, not more than five have a date affixed to them, and of nearly one hundred publications assigned to Caxton’s press, considerably more than two-thirds appear without any indication of the year of imprint.

8. The similarity, amounting almost to identity, between the printed characters of the early typographers and the written ones of their contemporaries, must also be noted. It was this similarity which probably first gave rise to the now admitted fable of Fust selling his bibles at Paris as manuscripts, his impeachment before the parliament as a sorcerer, and the necessity he was under of revealing his secret to save his life.

The first printer, when he set about forming his alphabet, could not have been troubled as to the shape he should give his letters. The form which would naturally occur to him would be that to which both he and the people to whom he hoped to sell his productions had been accustomed. It is not therefore at all wonderful, that the types used in the earliest printed books should closely resemble the written characters of the period, nor that this imitation should be extended to all the combinations of letters which were then in use by the scribes. Thus the bibles and psalters which appeared in Germany, among the first productions of the press, were printed in the characters used by the scribes for ecclesiastical service-books, while the general literature was printed in the common bastard-roman. There is nothing whatever to support the assertion frequently met with, that the first printers made their books purposely like the old manuscripts in order to deceive purchasers into paying a good price for them. This view truly is the genuine outcome of the nineteenth century.

When Sweynheym and Pannartz, emigrating from Germany, took up their abode in the famous monastery of Subiaco, near Rome, they cut the punches for their new types in imitation of the Roman letters indigenous to the country, without an idea that they could do anything else. In the dominions of the Duke of Burgundy, where the labours of the scribes had been most extensively encouraged, the same plan was pursued. Colard Mansion, the first printer at Bruges, was also a celebrated caligrapher, and the close resemblance between his printed books and the best manuscripts of his time is very marked. The same character of writing was also in use in England, and Caxton’s types accordingly bear the closest resemblance to the handwriting in the Mercers’ books, and to the volumes of that period in the Archives at Guildhall. Nevertheless Dibdin thus censures Caxton for not adopting Roman types:--“That perfect order and symmetry of press-work, so immediately striking in the pages of foreign books of this period, are in vain to be sought for among the volumes which have issued from Caxton’s press; and the uniform rejection of the Roman letter so successfully introduced by the Spiras, Jenson, and Sweynheym and Pannartz is, unquestionably, a blemish on our printer’s typographical reputation.”

9. The short spacing of the early printers also deserves remark.[2]--The uneven length of the lines, so noticeable in manuscripts, was a necessity, as the writer could not forecast the space between the words so as to make all the lines of an even length. But it certainly was no necessity with the printer; for although in this respect the time-honoured custom of the scribes was followed for a few years, the improved appearance which evenness gave to the work was soon observed, and thus a typographical step in advance was established. At Mentz and Cologne this occurred at a very early stage. The first Psalter, printed in 1457, and the Mazarine Bible of 1455 show, now and then, lines slightly deficient in length, as do some of the earliest productions of Ulric Zel; but this rudeness soon gave way to a systematic plan of spacing the lines to one even length. In the early specimens from the Bruges and Westminster presses, the practice of placing all the spare space at the end of the lines, instead of dividing it between the words, gives a very rude appearance to the page, and in these books it is carried to a greater extent than in the works of any German, Italian, or French printers. Colard Mansion abandoned this practice in 1479, and Caxton in 1480.

It will be apparent, from the foregoing remarks, that the books of our first printers bore no slight resemblance to manuscripts, and indeed, until quite recently, a copy of the Mazarine Bible, in the Library of Lambeth Palace, was so regarded;[3] but this resemblance was soon modified, in many particulars, to suit the requirements of typography.

The execution of manuscript capitals being both tedious and expensive led to the early introduction of large letters engraved on wood, which were either printed in black at the same impression as the other portion, or in red by a subsequent operation. Colard Mansion seems never to have adopted them, although several of his books are illustrated by large and numerous woodcuts. Caxton inserted illustrations engraved on wood in two or three books before 1484, the date of “Æsop,” in which woodcut initials first appear.

Title-pages, likewise, are purely typographical in their origin, the scribes having been content with heading their page with “Hic incipit” and the name of the treatise. Caxton followed the usage of the scribes in this particular; for, with one exception only, and at the very end of his career, where the title of the book is printed alone in the centre of the first page, his books appear without any title-page whatever.

Wynken de Worde adopted the use of title-pages immediately after the death of his master, but Machlinia of London, and the schoolmaster-printer of St. Alban’s, never used them.

These minute details may appear, at first sight, to be hardly worthy of record; but when we remember that two-thirds of Caxton’s books are without any date, and that, by careful examination of the workmanship, we can trace the printer gradually developing the changes from manuscript to typographical character, we appreciate the existence of a mass of technical evidence which, like the strata of the earth, or the mouldings of a cathedral arch, affords chronological data quite independent of any other source, and enables us, with a near approach to accuracy, to determine the age of any undated book. To this evidence may be added some other important signs which sometimes bear witness to the date when a book was printed. Such are the size of the printed page, its depth and width, the number of lines in a page, the number of sheets in a section, and, above all, the sequence in the use of various types. In Caxton’s books this sequence is very remarkable, as will be seen by the annexed table, where only books with fixed dates are entered, so that the reader may form his own judgment as to the chronological order of the above-mentioned peculiarities.

Some interesting facts may be gathered from this table.

1. The types used by Caxton bear a definite chronological relation to one another. Type No. 1 goes out of use, and is succeeded, in 1477, by No. 2. Type No. 3 is principally employed for headlines during the use of Nos. 2 and 4. In 1480 type No. 4 makes its appearance, but not till No. 2 is about to disappear. In 1483 type No. 4* supersedes its predecessor, and, in its turn, makes way for Nos. 5 and 6, which close the list. If the books were added which give the dates of their translation, which almost always coincide with those of their printing, the result would be the same.

2. All the books printed before 1480 were with lines of an uneven length, whilst all printed subsequently were spaced out evenly.

3. Signatures and even spacing of the lines were synchronous improvements, and both, when once adopted, were never afterwards abandoned. In the signatures themselves a curious fact may be noted--that whereas the custom of Caxton was generally to use letters and Roman numerals, as ~b j~, for his signatures, yet in the three years 1481 to 1483, and at no other period, he used Arabic numerals, thus ~b~ 1, or 2 1.

┌────────────────────────┬────────────────────────────────────────────────────┐ │ │ Date of Printing. │ │ │ │ No. of Type. │ │ │ │ │ Length of Line. │ │ │ │ │ Inches. │ │ │ │ │ │ Lines spaced out to │ │ │ │ │ │ the end or not. │ │ │ │ │ │ │ Length of Page. │ │ │ │ │ │ Lines. │ │ │ │ │ │ │ │ Sig.│ │ │ Title. │ │ │ │ │ │ │Init| ├────────────────────────┼─────────────────┼───────┼────┼─────┼────┼─────┼────┤ │ The Recuyell │ Before 1477 │ 1 │ 5 │ No │ 31 │none │none│ │ The Game of Chess, │ │ │ │ │ │ │ │ │ 1st ed. │ Do. │ 1 │ 5 │ No │ 31 │none │none│ │ The Life of Jason │ Do. │ 2 │ 5 │ No │ 29 │none │none│ │ Dictes, 1st ed. │ Nov. 18th, 1477 │ 2 │ 5 │ No │ 29 │none │none│ │ Moral Proverbs │ Feb. 20th, 1478 │ 2 │ In Metre.│ 28 │none │none│ │ Cordyale │ Mar. 24th, 1479 │ 2 & 3 │ 5 │ No │ 29 │none │none│ │ Chronicles, 1st ed. │ June 10th, 1480 │ 4 │ 4¾ │ Yes │ 40 │roman│none│ │ Reynard, 1st ed. │ June 6th, 1481 │ 2 │ 4¾ │ Yes │ 29 │arab.│none│ │ Tulle │ Aug. 12th, 1481 │ 2 │ 4¾ │ Yes │ 29 │arab.│none│ │ Godfrey │ Nov. 20th, 1481 │ 4 │ 4¾ │ Yes │ 40 │arab.│none│ │ Polycronicon │ July 2d, 1482 │ 4 │ 4¾ │ Yes │ 40 │arab.│none│ │ Chronicles, 2d ed. │ Oct. 8th, 1482 │ 4 │ 4¾ │ Yes │ 40 │arab.│none│ │ Pilgrimage of the Soul │ June 6th, 1483 │ 4 │ 4¾ │ Yes │ 40 │roman│none│ │ Liber Festivalis, │ │ │ │ │ │ │ │ │ 1st ed. │ June 30th, 1483 │ 4* │ 5 │ Yes │ 38 │roman│none│ │ Confessio Amantis │ Sept. 2d, 1483 │ 4 │ 2¾ │ Yes │ 46 │arab.│none│ │ Knight of the Tower │ Jan. 31st, 1484 │ 4* │ 5 │ Yes │ 38 │roman│none│ │ Æsop │ Mar. 26th, 1484 │ 4* │ 4¾ │ Yes │ 38 │roman│wood│ │ King Arthur │ July 31st, 1485 │ 4* │ 3¼ │ Yes │ 26 │roman│wood│ │ Charles the Great │ Dec. 1st, 1485 │ 4* │ 4¾ │ Yes │ 38 │roman│wood│ │ Paris and Vienne │ Dec. 19th, 1485 │ 4* │ 2⅜ │ Yes │ 39 │roman│wood│ │ Book of Good Manners │ May 11th, 1487 │ 5 │ 4¾ │ Yes │ 34 │roman│wood│ │ Directorium │ │ │ │ │ │ │ │ │ Sacerdotum, 2d ed. │ ---- 1489 │ 6 │ 4¾ │ Yes │ 31 │roman│wood│ │ Art and Craft │ June 15th, 1490 │ 6 │ 4¾ │ Yes │ 31 │roman│wood│ │ Eneydos │ June 22d, 1490 │ 6 │ 4¾ │ Yes │ 31 │roman│wood│ │ Fayts of Arms │ July 14th, 1490 │ 6 │ 4¾ │ Yes │ 31 │roman│wood│ └────────────────────────┴─────────────────┴───────┴────┴─────┴────┴─────┴────┘

KEY
Heading “Sig.” = “Signatures.”
Heading “Init” = “Initials.”
Under “Lines spaced out to the end or not”: No = “not”, Yes = “spaced out”
Under “Sig.”: roman = “rom. num.”, arab. = “arab. figs.”

We may further add that the use of the paragraph mark (ℂ) never appears before 1483; that the great device makes no appearance till 1487, the printed date to the third edition of the “Dictes” notwithstanding; and that initials in wood first appear in the “Æsop” in 1484.

By the application of these tests to the undated books we are enabled to assign each of them, with tolerable certainty, to a particular period.

FOOTNOTES:

[2] We may here observe, that bibliographers often misuse the word “justification” when referring to the practice of placing all the space at the end of lines. The printer’s term “justification” does not necessarily refer to the spacing out of the words in a line. Every line in a page must be “justified” or made of the normal length, and the last line in a paragraph, containing perhaps no more than one word, must be justified equally with the full-length line. Short lines are justified with quadrats, or pieces of metal, which fill up the line, but, being lower than the type, do not print. What is called “short,” or “bad,” or “imperfect justification,” is sure to reveal itself, to the dismay of the compositor, by allowing the faulty line to drop out when the “forme,” or mass of type, is lifted. The probable reason why Colard Mansion and Caxton did not space their lines to an even length is, that at that time they had not begun to use the _setting-rule_. This useful little slip of metal enables each letter as it is picked up by the compositor to be passed along on an even surface to its destination, instead of catching in every unevenness or burr of the previous line. Its absence would entail many obstructions to the spacing-out of lines, and render the plan of leaving all the spare space at the end, which was actually adopted by Caxton, at once more easy, expeditious, and free from accident.

[3] In 1856, an old established bookseller, in one of our largest cathedral towns, marked a copy of Caxton’s “Statutes of Hen. VII” as an old MS., _and sold it for 2s. 6d._! See also the remarks on Verard’s “Euryalus et Lucrece,” in the Catalogue of the Harleian MSS., Vol. III, No. 4392.