MANUSCRIPT, a term applied to any document written by the human hand (Lat. manû scriptum) with the aid of pen, pencil or other instrument which can be used with cursive facility, as distinguished from an inscription engraved with chisel or graver, worked laboriously. By usage the word has come to be employed in a special sense to indicate a written work of the ancient world or of the middle ages; collections of such “ancient manuscripts” being highly prized and being stored for preservation in public libraries. Down to the time of the invention of printing, and until the printed book had driven it out of the field, the manuscript was the vehicle for the conservation and dissemination of literature, and discharged all the functions of the modern book. In the present article a description is given of the development of the ancient manuscript, particularly among the Greeks and Romans, leading on to the medieval manuscripts of Europe, and bringing down the history of the latter to the invention of printing; the history of the printed volume is dealt with in the article Book (q.v.).
Materials.—The handbooks on palaeography describe in full the different materials which have been employed from remote time to receive writing, and may be referred to for minuter details. To dispose, in the first place, of the harder materials that have been put under requisition, we find metals both referred to by writers and actually represented by surviving examples. Thin leaves of gold or silver were recommended for the inscription of charms in particular. Leaden plates were in common use for incantations; the material was cheap and was supposed to be durable. On such plates were scratched the dirae or solemn devotions of obnoxious persons to the infernal deities; many examples have survived. As an instance of the use of soft substance afterwards hardened may be cited the practice by the Babylonians and Assyrians of writing, or rather of puncturing, their cuneiform characters on clay tablets while moist, which were afterwards dried in the heat of the sun or baked in the oven. Potsherds, or ostraka, were employed for all kinds of temporary purposes. Thousands of them have been found in Egypt inscribed with tax receipts and ephemeral drafts and memoranda, children’s dictation lessons, &c. Analogous to the clay documents of western Asia are the tablets coated with wax in vogue among the Greeks and Romans, offering a surface not to be inscribed with the pen but to be scratched with the sharp pointed stilus. These will be described more fully below. With them we class the wooden boards, generally whitened with a coating of paint or composition and adapted for the pen, which were common in Egypt, and were specially used for educational purposes. Such boards were also employed for official notices in Athens in the 4th century B.C.
Of the more pliant, and therefore generally more convenient, substances there were many, such as animal skins and vegetable growths. Practically we might confine our attention to three of them: papyrus, parchment or vellum, and paper, the employment of which, each in turn, as a writing material became almost universal. But there are also others which must be mentioned.
In a primitive state of society leaves of plants and trees strong enough for the purpose might be taken as a ready-made material to receive writing. Palm leaves are used for this purpose to the present day in parts of India; and the references in classical authors to leaves as early writing material among the Greeks and Romans cannot be dismissed as entirely fanciful.
The bark of trees, and particularly the inner bark of the lime-tree, φιλύρα tilia, was employed. The fact that the Latin word liber, bark, eventually meant also a book, would be sufficient proof that that material was once in common literary use, even if it were not referred to by writers.
Linen, too, was a writing material among the early Romans, as it was also among the Etruscans, and as it had been to some extent among the Egyptians.
Skins of animals, tanned, have doubtless served as a writing material from the very earliest period of the use of letters. The Egyptians occasionally employed this material. Instances of the use of leather in western Asia are recorded by ancient authors, and from Herodotus we learn that the Ionian Greeks applied to the rolls of the later-imported papyrus the title διφθέραι, skins, by which they had designated their writing material of leather. The Jews, also, to the present day hold to the ancient Eastern custom and inscribe the law upon skin rolls.
But generally these materials were superseded in the old world by the famous Egyptian writing material manufactured from the papyrus plant, which gradually passed beyond the boundaries of its native land and was imported at a remote period into other countries. Into Greece and into Rome it was introduced at so early a time that practically it was the vehicle for classical literature throughout its course. A description of the manufacture and use of this material will be found under Papyrus. Here it need only be noted that papyrus is associated in Greek and Roman literature with the roll form of the ancient manuscript, as will be more fully explained below, and that it was the supersession of this material by parchment or vellum which led to the change of shape to the book form.
The introduction of the new material, parchment or vellum, was not a revival of the use of animal skins as followed by the old world. The skins were now not tanned into leather, but were prepared by a new process to provide a material, thin, strong, flexible, and smooth of surface on both faces. This improved process was the secret of the success of the new material in ousting the time-honoured papyrus from its high position. The common story, as told by Pliny, that Eumenes II. of Pergamum (197–158 B.C.), seeking to extend the library of his capital, was opposed by the jealousy of the Ptolemies, who forbade the export of papyrus, hoping thus to check the growth of a rival library, and that he was thus compelled to have recourse to skins as a writing material, at all events points to Pergamum as the chief centre of trade in the material, περγαμηνή charta pergamena. The old terms διφθέραι, membranae, applied originally to the older leather, were transferred to the newly improved substance. In describing MSS. written on, this material, by common consent the term parchment has in modern times given place to that of vellum, properly applicable only to calfskin, but now generally used in reference to a medieval skin-book of any kind. Parchment is a title now usually reserved for the hard sheepskin or other skin material on which law deeds are engrossed. (See Parchment.)
Vellum had a long career as a writing material for the literature of the early centuries of our era and of the middle ages. But in its turn it eventually gave place to paper (q.v.). As early as the 13th century paper, an Asiatic invention, was making its way into Europe and was adopted in the Eastern Empire as a material for Greek literature side by side with vellum. It soon afterwards began to appear in the countries of southern Europe. In the course of the 14th century the use of it became fairly established, and in the middle of the century a number of paper manuscripts were produced along with those on vellum, particularly in Italy. Finally, in the 15th century paper became the common material for the manuscript book. The new paper, however, made no further change in the form of the manuscript. It possessed exactly the same qualities, as a writing material, as vellum: it could be inscribed on both sides; it could be made up into quires and bound in the codex form; and it had the further advantage of being easily manufactured in large quantities, and therefore of being comparatively cheap.
The Forms of the Manuscript Book.—In describing the development of the manuscript book in the ancient world, and subsequently in the middle ages, we have to deal with it in two forms. The common form of the book of the ancient world was the roll, composed of one continuous sheet of material and inscribed only on one side. This form had a long career. In Egyptian literature it can be traced back for thousands of years. In Greek literature it may be assumed to have been in vogue from the earliest times; actual examples have survived of the latter part of the 4th and beginning of the 3rd centuries B.C. As to its early use in Latin literature we cannot speak so definitely; but Rome followed the example of Greece in letters, and therefore no doubt also in the material shape of literary productions. Both in Greek and Latin literature the roll lasted down to the early centuries of the Christian era. It was superseded by the codex, the manuscript in book form (in the modern sense of the word book), composed of separate leaves stitched together into quires and made available to receive writing on both sides of the material. This form is still in vogue as the modern printed book, and probably will never be superseded. But the codex in this developed shape was only an evolution from the early waxen tablets of the Greeks and Romans, two or more of which, hinged together, formed the primitive codex which suggested the later form. Therefore it will be necessary to include the description of the tablets with that of the later codex.
The ordinary terms in use among the Greeks for a book (that
is, a roll) were βίβλος (another, form of βύβλος, papyrus) and
its diminutive βιβλίον, which included the idea of a
written book. The corresponding Latin terms were
The Roll.
liber and libellus; volumen was a rolled-up roll. A roll of material
uninscribed was χάρτης, charta, or τόμος (originally a cutting of
papyrus), applicable also to a roll containing a portion or
division of a large work which extended to more than one
roll. A work contained within the compass of a single roll was
a μονόβιβλος, or μονόβιβλον. The term τεῦχος seems also to
have meant a single roll, but it was also applied at a later time
to indicate a work contained in several rolls.
In writing the text of a work, the scribe might choose to make use of separate sheets of papyrus, κολλήματα, schedae, and then join them to one another consecutively so as to make up the roll; or he might purchase from the stationers a scapus, or ready-made roll of twenty sheets at most; and if this length were not sufficient, he might add other sheets or scapi, and thus make a roll of indefinite length. But proverbially a great book was a great evil, and, considering the inconvenience of unrolling a long roll, not only for perusal, but, still more so, for occasional reference, the practice of subdividing lengthy works into divisions of convenient size, adapted to the capacity of moderate-sized rolls, must have come into vogue at a very early period.
It was the practice to write on one side only of the papyrus; to write on both front and back of a roll would obviously be a clumsy and irritating method. Works intended for the market were never opisthograph. Of course the blank backs of written rolls which had become obsolete might be turned to account for personal or temporary purposes, as we learn not only from references in classical authors but also from actual examples. The most interesting extant case of an opisthograph papyrus is the copy of Aristotle’s Constitution of Athens in the British Museum, which is written on the back of a farmer’s accounts, of the end of the 1st century—but only for private use. It being the rule, then, to confine the writing to one side of the material, that is, to the inner surface of the made-up roll, that surface was more carefully prepared and smoothed than the other; and, further, the joints of the several sheets were so well made that they offered no obstacle to the action of the pen. Still further, care was taken that this, the recto surface of the material, should be that in which the shreds of papyrus of which it was composed lay horizontally, so that the pen might move freely along the fibres; the shreds of the verso side, on the other hand, being in vertical position. This point is of some importance, as, in cases where two different handwritings are found on the two sides of a papyrus, it may be usually assumed that the one on the recto surface is the earlier.
The text was written in columns, σελίδες, paginae, the width of which seems not to have been prescribed, but which for calligraphic effect were by preference made narrow, sufficient margins being left at head and foot. The average width of the columns in the best extant papyri ranges from two to three-and-a-half inches. The written lines were parallel with the length of the roll, so that the columns stood, so to say, with the height of the rolled-up roll, and were disclosed consecutively as the roll was unwound. Ruling with lead to guide the writing is mentioned by writers, but it does not appear that the practice was generally followed. The number of lines in the several columns of extant papyri is not constant, nor is the marginal boundary of the beginnings of the lines, for the accuracy of which a ruled vertical line would have proved useful, ordinarily kept even. No doubt in practice the horizontal fibres of the material were found to afford a sufficient guide for the lines of writing.
If the title of the work was to be given, the scribe appears to have written it ordinarily at the end of the text. But something more was needed. To be obliged to unroll a text to the end, in order to ascertain the name of the author, would be the height of inconvenience. Its title was therefore sometimes written at the head of the text. It appears also that at an early period it was inscribed on the outside of the roll, so as to be visible as the roll lay in a chest or on the shelf. But a more general practice was to attach to the top edge of the roll a label or ticket, σίλλυβος, or σίττυβος, titulus, index, which hung down if the roll lay on the shelf, or was conveniently read if the roll stood along with others in the ordinary cylindrical roll-box, κίστη, κιβωτός, cista, capsa. One such label made of papyrus has survived and is in the British Museum.
The scribe would not commence his text at the very beginning, nor would he carry it quite down to the end, of the roll. He would leave blank a sufficient length of material at either extremity, where the roll would naturally be most exposed to wear and tear by handling in unrolling and re-rolling; and, further, the extreme vertical edges might each be strengthened by the addition of a strip of papyrus so as to form a double thickness of material.
According to the particulars given by classical authors, the roll would be finished off somewhat elaborately; but the details described by them must be taken to apply to the more expensive productions of the book trade, corresponding with the full-bound volumes of our days. In practice, a large proportion of working copies and ordinary editions must have been dealt with more simply. Firstly, the roll should be rolled up round a central stick, of wood or bone, called the ὀμφαλός, umbilicus, to which the last sheet of the papyrus may or may not have been attached. But as a matter of fact no rolling-sticks have been found in company with extant papyri, and it has therefore been suggested that they were not attached to the material but were rolled in loose, and were therefore liable to drop out. In some instances, as in the rolls found at Herculaneum, a central core of papyrus instead of a stick was thought sufficient. The edges, frontes, of the roll, after it had been rolled up, were shorn and were rubbed smooth with pumice, and they were sometimes coloured. A valuable roll might be protected with a vellum wrapper, φαινόλης, paenula, stained with colour; and, further, it might be secured with ornamental thongs. The central stick might also be adorned with knobs or “horns,” plain or coloured. This seems to be the natural explanation of the κέρατα, or cornua, mentioned by the ancient writers. Finally, the title-label described above was attached to the completed roll, now ready for the book-market.
In the perusal of a work the reader held the roll upright and unrolled it gradually with the right hand; with the left hand he rolled up in the reverse direction what he had read. Thus, when he had finished, the roll had become reversed, the beginning of the text being now in the centre of the roll and the end of it being outside. The roll was “explicitus ad umbilicum,” or “ad sua cornua.” It had therefore now to be unrolled afresh and to be re-rolled into its normal shape—a troublesome process which the lazy man shirked, and which the careful man accomplished by making the revolutions with his two hands while he held the revolving material steady under his chin.
Although the codex or manuscript in book-form began to make its way in Greek and Roman literature as early as the 1st century of our era, the roll maintained its position as the recognized type of literary document down to the 3rd, and even into the 4th, century, when it was altogether superseded. We shall proceed to describe the codex after giving some account of the waxen, or, to speak more correctly, the waxed, tablet, its precursor in the book-form.
The ordinary waxen tablet in use among the Greeks and Romans was a small oblong slab of wood, beech, fir, and especially box, the surface of which on one or both sides, with the exception of the surrounding margins which were The Waxen Tablet. left intact in order to form a frame, was sunk to a slight depth and was therein coated with a thin layer of wax, usually black. The tablet thus presented the appearance of a child’s school-slate of the present day. Such tablets were single, double, triple, or of several pieces or leaves. In Greek they were called πίναξ, πινακίς, δέλτος, δελτίον.: in Latin cera, tabula, tabella, &c. Two or more put together and held together by rings or thongs acting as hinges formed a caudex or codex, literally a stock of wood, which a set of tablets might resemble, and from which they might actually be made by cleaving the wood. A codex of two leaves was called δίθυροι, δίπτυχα, diptycha; of three, τρίπτυχα, triptycha: and so on. The triptych appears to have been most generally used. A general term was also libellus.
Tablets served for the ordinary minor affairs of life: for memoranda, literary and other notes and drafts, school exercises, accounts, &c. The writing incised with the stilus could be easily obliterated by smoothing the wax, and the tabula rasa was thus rendered available for a fresh inscription. But tablets were also employed for official purposes, when documents had to be protected from unauthorized scrutiny or from injury. Thus they were the receptacles for wills, conveyances, and other legal transactions; and in such cases they were closed against inspection by being bound round with threads which were covered by the witnesses’ seals.
Small tablets, codicilli, pugillares, often of more valuable material, such as ivory, served for correspondence among other purposes; very small specimens are mentioned as vitelliani, for the exchange of love-letters.
A certain number of Greek waxen tablets have been recovered, chiefly from Egypt, but none of them is very early. They are generally of the 3rd century, and are mostly inscribed with school exercises. The largest and most perfect extant codex is one in the British Museum (Add. MS. 33,270), perhaps of the 3rd century, being made up of nine leaves, measuring nearly 9 by 7 in., and inscribed with documents in shorthand.
Of Latin tablets we are fortunate in having a fairly large number of examples. Exclusive of a few isolated specimens, they are the result of two important finds. Twenty-four tablets containing the records of a burial club, A.D. 131–167, were recovered between 1786 and 1855 from some ancient mining works in Dacia. In 1875 as many as 127 tablets, containing deeds connected with sales by auction and payment of taxes, A.D. 15–62, were found in the ruins of Pompeii. These specimens have afforded the means of ascertaining the mechanical arrangement of waxen tablets when adopted for legal instruments among the Romans. Most of them are triptychs, severally cloven from single blocks of wood. Subject to some variations, the triptych was usually arranged as follows. Of the six sides or pages of the codex, pages 1 and 6 (the outside pages) were of plain wood; pages 2, 3, 5 were waxed; and page 4, which had a groove cut across the middle was sometimes of plain wood, sometimes waxed. The authentic deed was inscribed with the stilus on the waxed pages 2 and 3; and the first two leaves were then bound round with three twisted threads which passed down the groove so as to close the deed from inspection. On page 4 the witnesses’ names were then inscribed (in ink if the page was plain; with the stilus if waxed), and their seals were impressed in the groove, thus securing the threads. In addition to the protection afforded to the seals from casual injury by their position in the groove, the third leaf acted as a cover to them. On page 5 an abstract or duplicate of the deed, as required by law, was inscribed. The arrangement of the Dacian tablets differed in this respect, that page 4 was waxed, and that the duplicate copy was begun on that page in the space on the left of the groove, that on the right being reserved for the names of the witnesses. In the case of one of the Pompeian tablets the threads and seals still remain.
The survival of the use of tablets to a late time should be noted. St Augustine refers to his tablets, and St Hilary of Arles also mentions their employment for the purpose of correspondence; there is a record of a letter written in tabellâ as late as A.D. 1148. They were very commonly used throughout the middle ages in all the west of Europe. Specimens inscribed with money accounts of the 13th and 14th centuries have survived in France, and similar documents of the 14th and 15th centuries are to be found in several of the municipal archives of Germany. Reference to their use in England occurs in literature, and specimens of the 14th or 15th century are said to have been dug up in Ireland. In Italy their employment is both recorded and proved by actual examples of the 13th and 14th centuries. With the beginning of the 16th century they seem to have practically come to an end, although a few survivals of the custom of writing on wax have lingered to modern times.
As already stated, the codex, or MS. in book-form, owed its existence to the substitution of vellum for papyrus as the common writing material for Greek and Roman literature. The fact that vellum was a tough material capable of being The Codex. inscribed on both sides, that writing, particularly if freshly written, could be easily washed off or erased from it, and that the material could thus be made available for second use, no doubt contributed largely to its ready adoption. In Rome in the 1st century B.C. it was used, like the waxen tablets for notes, drafts, memoranda, &c.; and vellum tablets began to take the place of the cerae. References are not wanting in the classical writers to its employment for such temporary purposes. To what extent it was at first pressed into the service of literature and used in the preparation of books for the market must remain uncertain. But in the first three centuries of our era it may be assumed that vellum codices were not numerous. The papyrus roll still held its position as the liber or book of literature. Yet we learn from the poems of Martial that in his day the works of some of the best classical authors were to be had on vellum. From the way in which, in his Apophoreta, he has contrasted as exchangeable gifts certain works written respectively on papyrus and on vellum, it has been argued that vellum at that time was a cheap material, inferior to papyrus, and only used for roughly written copies. Up to a certain point this may be true, but the fact that the earliest great vellum Greek codices of the Bible and of Latin classical authors, dating back to the 4th century, are composed of very finely prepared material would indicate a perfection of manufacture of long standing.
But, apart from the references of writers, we have the results of recent excavations in Egypt to enable us to form a more correct judgment on the early history of the vellum codex. There have been found a certain number of inscribed leaves and fragments of vellum of early date which without doubt originally formed part of codices or MSS. in book-form. It is true that they are not numerous, but from the character of the writing certain of them can be individually assigned to the 3rd, to the 2nd, and even to the 1st century. We may then take it for an established fact that the codex form of MS. was gradually thrusting its way into use in the first centuries of our era.
The convenience of the codex form for easy reference was also a special recommendation in its favour. There can be little doubt that such compilations as public registers must at once have been drawn up in the new form. The jurists also were quick to adopt it, and the very title “codex” has been attached to great legal compilations, such as those of Theodosius and Justinian. Again, the book-form was favoured by the early Christians. The Bible, the book which before all others became the great work of reference in their hands, could only be consulted with convenience and despatch in the new form. A single codex could hold the contents of a work which formerly must have been distributed through many volumes in roll-form. The term σωμάτιον, which was one of the names given to a codex, was expressive of its capacity. Turning again to discoveries in Egypt, it appears that in the early centuries the codex-form had become so usual among the Christians in that land that even the native material, papyrus, the recognized material for the roll, was now also made up by them into leaved books. The greater number of papyri of the 3rd century containing Christian writings, fragments of the Scriptures, the “Sayings of Our Lord,” and the like, are in book-form. On the other hand, the large majority of the non-Christian papyri of the same period keep to the old roll-form. Thus the codex becomes at once identified with the new religion, while the papyrus roll to the last is the chosen vehicle of pagan literature.
In the 4th century the struggle between the roll and the codex for supremacy in the literary field was finished, and the victory of the codex was achieved. Henceforward the roll-form remained in use for records and legal documents, and in certain instances for liturgies; and for such purposes it survives to the present day. But so completely was it superseded in literature by the codex that even when papyrus, the material once identified with the roll-form, was used as it sometimes was down to the 6th and 7th centuries and later, it was made up into the leaved codex, not only in Egypt but also in western Europe.
The shape which the codex usually assumed in the early centuries of the middle ages was the broad quarto. The quires or gatherings of which the book was formed generally consisted, in the earliest examples, of four sheets folded to make eight leaves (τετράς or τετράδιον, quaternio), Quires. although occasionally quinterns, or quires of five sheets (ten leaves), were adopted. Sexterns, or quires of six sheets (twelve leaves), came into use at a later period. In making up the quires, care was generally taken to lay the sheets of vellum in such a way that hair-side faced hair-side, and flesh-side faced flesh-side; so that, when the book was opened, the two pages before the reader had the same appearance, either the yellow tinge of the hair-side, or the fresh whiteness of the flesh-side. In Greek MSS. the arrangement of the sheets was afterwards reduced to a system; the first sheet was laid with the flesh-side downwards, so that that side began the quire; yet in so early an example as the Codex Alexandrinus the first page of a quire is the hair-side. In Latin MSS. also the hair-side appears generally to have formed the first page. When paper came into general use for codices in the 15th century, it was not an uncommon practice to give the paper quires additional strength by an admixture of vellum, a sheet of the latter material forming the outer leaves, and sometimes the middle leaves also, of the quire. The quire mark, or “signature,” was usually written at the foot of the last page, but in some early instances (e.g. the Codex Alexandrinus) it appears at the head of the first page of each quire. The numbering of the separate leaves in a quire, in the fashion followed by early printers, came in in the 14th century. Catch-words to connect the quires appear first in the 11th century and are not uncommon in the 12th century.
No exact system was followed in ruling the guiding lines on the pages of the codex. In the case of papyri it was enough to mark with the pencil the vertical marginal lines to bound the text, if indeed even this was considered needful (see above); the fibres of the papyrus were a sufficient guide Ruling. for the lines of writing. On vellum it became necessary to rule lines to keep the writing even. These lines were at first drawn with a blunt point, almost invariably on the hair (or outer) side of the skin, and strongly enough to be in relief on the flesh (or inner) side. Marginal lines were drawn to bound the text laterally; but the ruled lines which guided the writing were not infrequently drawn right across the sheet. Each sheet should be ruled separately; but two or more sheets were often laid and ruled together, the lines being drawn with so much force that the lower sheets also received the impressions. In rare instances lines are found ruled on both sides of the leaf, as in some parts of the Codex Alexandrinus. In this same MS. and in other early codices the ruling was not always drawn for every line of writing, but was occasionally spaced so that the writing ran between the ruled lines as well as on them. The lines were evenly spaced by means of guiding pricks made at measured intervals with a compass or rotary instrument down the margins; in some early MSS. these pricks run down the middle of the page. Ruling with the plummet or lead-point is found in the 11th century and came into ordinary use in the 12th century; coloured inks, e.g. red and violet, were used for ornamental ruling in the 15th century.
Mechanical Arrangement of Writing in MSS.—It has already been stated above that in the papyrus rolls the text was written in columns. They stood with convenient intervals between them and with fair margins at top and bottom. The length of the lines was to some extent governed Columns. by the nature of the text. If it was a poetical work, the metrical line was naturally the line of the column, unless, as sometimes was the case, the verse was written continuously as prose. For prose works a narrow column was preferred. It is noticeable that the columns in papyri have a tendency to lean to the right instead of being perpendicular—an indication that it was not the practice to rule marginal lines. In codices the columnar arrangement was also largely followed, and the number of columns in a page was commonly two. There are instances, however, of a larger number. The Codex Sinaiticus of the Bible has four columns to the page; and the Codex Vaticanus, three columns. And the tricolumnar arrangement occurs every now and then in later MSS.
In both Greek and Latin literary MSS. of early date the writing runs on continuously without separation of words. This practice however, may be regarded as rather artificial, as in papyri written in non-literary hands and in Latin deeds also, contemporary with these early literary Text without separation of Words. MSS., there is a tendency to separation. In a text thus continuously written occasional ambiguities necessarily occurred, and then a dot or apostrophe might be inserted between words to aid the reader. Following the system of separation of words which appears in ancient inscriptions, wherein the several words are marked off by single, double, or treble dots or points, the words of the fragmentary poem on the battle of Actium found at Herculaneum are separated by single points, probably to facilitate reading aloud; monosyllables or short prepositions and conjunctions, however, being left unseparated from the words immediately following them—a system which is found in practice at a later time. But such marks of separation are not to be confounded with similar marks of punctuation whereby sentences are marked off and the sense of the text is made clear. Throughout the career of the uncial codices down to the 6th century, continuity of text was maintained. In the 7th century there is some evidence of separation of words, but without system. In early Latin minuscule codices partial separation in an uncertain and hesitating manner went on to the time of the Carolingian reform. In early Irish and English MSS., however, separation is more consistently practised. In the 9th and 10th centuries long words tend to separation, but short words, prepositions and conjunctions, still cling to the following word. It was not till the 11th century that the smaller words at length stood apart, and systematic separation of words was established. In Greek minuscule codices of the 10th century a certain degree of separation takes place; yet a large proportion of words remain linked together, and they are even incorrectly divided. Indeed a correct system of distinct separation of words in Greek texts was never thoroughly established even as late as the 15th century.
But while distinction of words was disregarded in early literary texts, distinction of important pauses in the sense was recognized from the first. The papyrus of the Persae of Timotheus of Miletus, the oldest MS. of a Greek classic in existence, of the end of the 4th century B.C., is written Paragraphs. in independent paragraphs. This is a natural system, the simplicity of which has caused it to be the system of modern times. But, in addition, the Greek scribe also separated paragraphs by inserting a short horizontal stroke, παράγραφος, between them at the commencement of the lines of writing. It should be noted that this stroke indicated the close of a passage, and therefore belonged to the paragraph just concluded, and did not stand for an initial sign for the new paragraph which followed. The dividing stroke was also used to mark off the different speeches of a play. Besides the stroke, a wedge-shaped sign or tick might be used. But to make every paragraph stand distinctly by itself would have entailed a certain loss of space. If the concluding line were short, there would remain a long space unfilled. Therefore, when this occurred, it became customary to leave only a short space blank to mark the termination of the paragraph, and then to proceed with the new paragraph in the same line, the παράγραφος at the same time preventing possible ambiguity. The next step was to project the first letter of the first full line of the new paragraph slightly into the margin, as a still further distinction; and lastly to enlarge it. The enlargement of the letter gave it so much prominence that the dividing stroke could then be dispensed with, and in this form the new paragraph was henceforward indicated in Greek MSS., it being immaterial whether the enlarged letter was the initial or a medial letter of a word. As early as the 5th century there is evidence that the παράγραφος was losing its meaning with the scribes, for in the Codex Alexandrinus of the Bible it is not infrequently found in anomalous positions, particularly above the initial letters of different books, as if it were a mere ornament.
In Latin MSS. there was no such fixed system of marking off paragraphs as that just described. A new paragraph began with a new line, or a brief space in a line separated the conclusion of a paragraph from the beginning of the next one. It was only by the ultimate introduction of large letters, as the initial letters of the several sentences and paragraphs, and by the establishment of a system of punctuation, in the modern sense of the word, that a complete arrangement of the text was possible into sentences and paragraphs in accordance with its sense.
From the earliest times an elementary system of punctuation by points is found in papyri. Thus the papyrus of the Curse of Artemisia, at Vienna, which is at least as early as the 3rd century B.C., and in one or two other ancient examples, a double point, resembling the modern colon, Punctuation. separates sentences. But more commonly a single point, placed high in the line of writing, is employed. This single punctuation was reduced to a system by the Alexandrian grammarians, its invention being ascribed to Aristophanes of Byzantium, 260 B.C. The point placed high on a level with the top of the letters had the value of a full-stop; in the middle of the line of writing, of a comma; and low down on the line, of a semicolon. But these distinctions were not observed in the MSS. In the early vellum codices both the high and the middle point are found. In medieval MSS. other signs, coming nearer to our modern system, make their appearance. In Latin MSS. by the 7th century the high point has the value of the modern comma, the semicolon appears with its present value, and a point emphasized with additional signs, such as a second point or point and dash, marks a full-stop. In the Carolingian period the comma appears, as well as the inverted semicolon holding a position between our comma and semicolon.
Another detail which required the scribe’s attention in writing his text was the division of the last word in a line, when for want of room a portion of it had to be carried over into the next line. It was preferable, indeed, to avoid such division, and in the papyri as well as Division of Words at the End of a Line. in the codices letters might be reduced in size and huddled together at the end of the line with this view. In the early codices too it was a common practice to link letters together in monogrammatic form, such as the common verbal terminations ur, unt, and thus save space. But when the division of a word was necessary, it was subject to certain rules. According to the Greek practice the division was ordinarily made after a vowel, as ἔτυ|χον (even monosyllables might be so treated, as οὐ|κ). But in the case of double consonants the division fell after the first of them, as ἵπ|πος: and, when the first of two or more consonants was a liquid or nasal the division followed it, as ὀφθαλ|μός, μαν|θάνω. When a word was compounded with a preposition, the division usually followed the preposition, as προς|εῖπον, but not infrequently the normal practice of dividing after a vowel prevailed, as προ|σεῖπον. In Latin the true syllabic division was followed, but occasionally the scribes adopted the Greek system and divided after a vowel.
A modification of the practice of writing the text continuously was allowed in the case of certain works. Rhetorical texts, such as the orations of Demosthenes and Cicero, and the text of the Bible, might be broken up into short clauses or sense-lines, apparently with the view of assisting Colometry. reading aloud. Instances of MSS. so written are still extant. This system, to which the name of “colometry” has been given, is the arrangement by cola and commata referred to by St Jerome in his preface to Isaiah. It will be found more fully explained under the heading of Stichometry; where also is described the mechanical computation of the length of a text by measured lines, for the purpose of calculating the pay of the scribe.
The title of a MS., both in roll-form and in codex-form, was frequently written at the end of the text, but even at an early date it stood in some instances at the beginning; and the latter practice in course of time prevailed, although even in the 15th century the title was Titles and Colophons. sometimes reserved for the close of the MS. In this latter position it might stand alone or be accompanied by other particulars concerning the MS., such as the length of the work, the date of writing, the name of the scribe, &c., all combined in a final paragraph called the colophon. For distinction, title and colophon might be written in red, as might also the first few lines of the text. This method of rubrication was a very early practice, appearing even in ancient Egyptian papyri. Such rubrics and titles and colophons were at first written in the same character as the text; afterwards, when the admixture of different kinds of writing was allowed, capitals and uncials were used at discretion. Running titles or head-lines are found in some of the earliest Latin MSS. in the same characters as the text, but of a small size. Quotations were usually indicated Quotations. by ticks or arrow-heads in the margin, serving the purpose of the modern inverted commas. Sometimes the quoted words were arranged as a sub-paragraph or indented passage. In commentaries of later date, the quotations from the work commented upon were often written in a different style from the text of the commentary itself.
Accentuation, &c.—Accentuation was not systematically applied to Greek MSS. before the 7th century, but even in the literary papyri it appears occasionally. In the latter instances accents were applied specially to assist the reader, and they seem to have been used more frequently in texts which may have presented greater difficulties than usual. For example, they are found fairly plentifully in the papyrus of Bacchylides of the 1st century B.C. In the less well-written papyri they are fewer in number; and papyri written in non-literary hands are practically devoid of them. Accents have been frequently added to the ancient texts of Homer, as in the Harris and Bankes papyri, but apparently long after the date of the writing. They were not used in the early uncial MSS. Breathings also appear occasionally in the papyri. The rough and the smooth breathings are found in the form of the two halves of the H (├ ┤) in the Bacchylides papyrus; in other papyri they are in rectangular form, never rounded like an apostrophe; in fact rounded breathings do not come into general use until the 12th century. Other signs resembling accents are used occasionally in Greek MSS. For example, a short accent or horizontal stroke was employed to indicate a single-letter word, and an apostrophe was sometimes used to separate words in order to prevent ambiguity and was placed after words ending in κ, χ, ξ, ρ, and after proper names not having a Greek termination.
Accents were seldom employed by Latin scribes. In early Irish and English MSS., in particular, an acute accent is occasionally found over a monosyllabic word or one consisting of a single letter. In the 9th and 10th centuries a curious occasional practice obtained among the correctors of the texts of expressing the aspirate by the Greek half-eta symbol Ͱ, instead of writing the letter h in the ordinary way—perhaps only an affectation.
Corrections.—For obliteration or removing pen strokes from the surface of the material the sponge was used in ancient times. While the writing was still fresh, the scribe could easily wash off the ink by this means; and for a fragile material, such as papyrus, he could well use no other. On vellum he might use sponge or knife. But after a MS. had left his hands it would undergo revision at the hands of a corrector, who had to deal with the text in a different manner. He could no longer conveniently apply the sponge. On hard material he might still use the knife to erase letters or words or sentences. But he could also use his pen for such purposes. Thus we find that a very early system of indicating erasure was the placing of dots or minute strokes above the letters to be thus “expunged.” The same marks were also (and generally at later periods) placed under the letters; in rare instances they stood inside them. It need scarcely be said that letters were also struck out with strokes of the pen or altered into others, and that letters and words were interlined. A long sentence, however, which could not be admitted between the lines, was entered in the margin, and its place in the text indicated by corresponding reference marks, such as hd., hs. = hic deest, hoc supra or hic scribas, &c.
Abbreviations and Contractions.—The practice of shortening words in writing has played an important part in the history of the ancient and the medieval manuscript. Two reasons have disposed men to follow this practice: firstly, the desire to avoid the labour of writing over and over again words or portions of words of common occurrence which can be readily understood in a shortened form as when written in full; and, secondly, the necessity of saving space at a time when it was an object to make the most of the writing material to hand. To meet the former requirement, a simple and limited method alone was needed; to satisfy the second, a more elaborate system was necessary. The most natural method of reducing the length of a word is to suppress as much as possible of its termination, consistently with intelligibility, that is, by simple abbreviation. But if space of any appreciable value is to be saved in a page of writing, a system is necessary for eliminating letters from the body of the word as well as curtailing the termination, that is, a system of contraction as well as abbreviation; and, in addition, the employment of arbitrary signs, analogous to shorthand, will serve still further to condense the text. An elaborate system of contraction of this nature was naturally only fully developed after very long practice. Both in Greek and in Latin MSS. from the 9th to the 15th century such a system was in full force.
Different kinds of literature were, according to their nature, more or less abbreviated and contracted. From early times such curtailment was more freely employed in works written in technical language, such as works on law or grammar or mathematics, wherein particular words are more liable to repetition, than in MSS. of general literature. The oldest system of abbreviation is that in which a single letter (nearly always the initial letter) or at most two or three letters represent the whole word. This system we know was in common use among both Greek and Latin writers, and ancient inscriptions afford plentiful examples. It is well adapted for the brief expression of the common words and phrases in works of a technical nature (as for example such a phrase as C D E R N E = cujus de ea re notio est); but for general literature it is of little use, and practically has been restricted to express proper names and numerals.
When abbreviations were employed only with the view of speed in writing, it is obvious that they would occur more frequently in the ephemeral documents of daily life than in carefully written literary works intended for the book-market. Abbreviation in Greek MSS. Hence they are not to be found in Greek papyri of the latter class. On the other hand in literary papyri written in non-literary script they naturally occur just as they would in contemporary common documents. As early as the 3rd and 2nd centuries B.C. the ordinary method of abbreviation was to omit the termination or latter portion of the word and to mark the omission by a short horizontal stroke or dash; or the letter which immediately preceded the omission was written above the line as a key to the reading, as τελ for τέλος. Such a system obviously might be extended indefinitely at the discretion of the writer. But in addition, at quite an early period, symbols and monogrammatic forms for particular words must have been developed, for they are found in common use in cursive papyri. A notable instance of their employment in a full degree occurs in the papyrus of Aristotle’s Constitution of Athens, of the 1st century.
Like the well-written literary papyri, the early vellum uncial codices of the Bible, being inscribed with calligraphic formality, avoided in principle the use of abbreviations. But by the 4th to the 6th century, the period when they were chiefly produced, the contraction or abbreviation of certain words and terminations had, it seems, become so fixed by usage that the contracted forms were adopted in the texts. They are ΘC = θεός, ΙC = ίησοῦς, ΧC = χριστός, ΠΝΑ = πνεῦμα, CΗΡ = σωτήρ, ΚC = κύριος, CΤΡΟC = σταυρός, ΠΗΡ = πατήρ, ΜΗΡ = μήτηρ, ΥC = υἱός, ΑΝΟC = ἄνθρωπος, ΟΥΝΟC = οὐρανός, Κ = καί, Τ = ται, Μ = μου, μοι, &c. Final Ν, especially at the end of a line, was dropped, and its place occupied by the horizontal stroke, as ΤΟ .
But while this limited system was used in biblical, and also in liturgical MSS., in profane literature a greater licence was recognized. For example, in a fragment of a mathematical work at Milan, of the 7th century, we find instances of abbreviation by dropping terminations, just as in the earlier papyri, and, in addition, contracted particles and prepositions are numerous. Technical works, in fact, inherited the system instituted in the early papyri written in non-literary or cursive hands; and this system, undergoing continual development, had a larger scope when the cursive writing was cast into a literary form and became the literary minuscule script of the middle ages. From the 9th century onwards a fully developed system of abbreviation and contraction was practised in Greek MSS., comprising the early system of the papyri, the special contractions of the early biblical MSS., and also a large number of special symbols, derived in great measure from tachygraphical signs.
In the early Greek minuscule MSS. contractions are not very frequent in the texts; but in the marginal glosses, where it was an object to save space, they are found in great numbers as early as the 10th century. The MS. of Nonnus, of A.D. 972, in the British Museum (Wattenb. and Von Vels., Exempla, 7) is an instance of a text contracted to a degree that almost amounts to tachygraphy. In the 12th, 13th and 14th centuries texts were fully contracted; and as the writing became more cursive contraction-marks were more carelessly applied, until, in the 15th century, they degenerated into mere flourishes.
As far back as material is available for comparison, it appears that abbreviations and contractions in Latin MSS. followed the same lines as those in Greek MSS. We have no very early papyri written in Latin as we have in Greek to show us what Abbreviations in Latin MSS. the practice of Roman writers was in the 3rd and 2nd and early 1st centuries B.C.; but there can be little doubt that in that remote time there was followed in Latin writing a system of abbreviation similar to that in Greek, that is, by curtailment of terminations, and that in ephemeral documents written in cursive characters such abbreviation was allowed more freely than in carefully written literary works. The early system of representing words by their initial letters has already been referred to. It was in common use, as we know, in the inscriptions on coins and monuments, and to some extent in the texts of Roman writers. But the ambiguity which must have always accompanied such a system of single-letter abbreviations, or sigla, naturally induced an improvement by expressing a word by two or more of its letters. Hence was developed the more regular syllabic system of the Romans, by which the leading letters of the several syllables were written, as EG = ergo, HR = heres, ST = satis. At a later time Christian writers secured greater exactness by expressing the final letter of a contracted word, as ds = deus, dō = deo, scs = sanctus. Further, certain marks and signs, many derived from shorthand symbols, came into use to indicate inflections and terminations; or the terminating letter or a leading letter to indicate the termination might be written above the line, as Qo = quo, Vm = verum, No = noster, Si = sint. This practice became capable of greater development later on. Among the special signs are c = est, ƚ = vel, n = non, p’ = pre, ꝑ = per, ꝓ = pro, 9 = termination us. The letter q with distinctive strokes applied in different positions represented the often recurring relative and other short words, as quod, quia.
In Latin Biblical uncial MSS. the same restrictions on abbreviations were exercised as in the Greek. The sacred names and titles DS = deus, DMS, DNS = dominus, SCS = sanctus, SPS = spiritus, and others appear in the oldest codices. The contracted terminations Q· = que, B· = bus, and the omission of final m, or (more rarely) final n, are common to all Latin MSS. of the earliest period. There is a peculiarity about the contracted form of our Saviour’s name that it is always written by the Latin scribes in letters imitating the Greek IHC, XPC, ihc, xpc, and ihs, xps.
The full development of the medieval system of abbreviation and
contraction was effected at the time when the Carolingian schools
were compelling the reform of the handwriting of western Europe.
Then came a freer practice of abbreviation by suppression of terminations
and the latter portions of words, the omission of which was
indicated by the ordinary signs, the horizontal or oblique stroke or the
apostrophe; then came also a freer practice of contraction by omitting
letters and syllables from the middle as well as the end of words,
as oio, omnino, prb, presbyter; and then from the practice of writing
above the line a leading letter of an omitted syllable, as inta = intra,
tr = tur, conventional signs, with special significations, were also
gradually developed. Such growths are well illustrated in the change
undergone by the semicolon, which was attached to the end of a word
to indicate the omission of the termination, as b; = bus, q; = que,
deb; = debet, and which in course of time became converted into
a z, a form which survives in our ordinary abbreviation, viz. (i.e.
vi; = videlicet). The different forms of contraction were common
to all the nations of western Europe. The Spanish scribes, however,
attached different values to certain of them. For example, in
Visigothic MSS., qm, which elsewhere represented quoniam, may be
read as quum; and ꝓ, which elsewhere = pro, is here = per. Nor
must the use of arbitrary symbols for special words be forgotten.
These are generally adaptations of the shorthand signs known as
Tironian notes. Such are 
= autem, ÷ = est, ℈ = ejus, 
= enim,
⁊ = et, v̇ and u̇ = ut, which were employed particularly in early MSS. of
English and Irish origin.
By the 11th century the system of Latin contractions had been reduced to exact rules; and from this time onwards it was universally practised. It reached its culminating point in the 13th century, the period of increasing demand for MSS., when it became more than ever necessary to economize space. After this date the exact formation of the signs of contractions was less strictly observed, and the system deteriorated together with the decline of handwriting. In conclusion, it may be noticed that in MSS. written in the vernacular tongues contractions are more rarely used than in Latin texts. A system suited to the inflexions and terminations of this language could not be readily adapted to other languages so different in grammatical structure.
Palimpsests, &c.—Palimpsest MSS., that is, MSS. written upon material from which older writing has been previously removed by washing or scraping, are described in a separate article (Palimpsest). The ornamentation of MSS. is fully dealt with under the headings Illuminated MSS., and Miniatures.
Writing Implements.—In conclusion, a few words may be added respecting the writing implements employed in the production of MSS. The reed, κάλαμος, calamus, was adapted for tracing characters either on papyrus or vellum. By the ancient Egyptians, and also probably by the early Greek scribes in Egypt, it was used with a soft brush-like point, rather as a paint-brush than as a pen. The Greek and Roman scribes used the reed cut to a point and slit like the quill-pen; and it survived as a writing implement into the middle ages. For scratching letters on the waxen tablet the sharp pointed bodkin, στῦλος, γραφεῖον, stilus, graphium, was necessary, made of iron, bronze, ivory, or other suitable material, with a knobbed or flattened butt-end wherewith corrections could be made by smoothening the wax surface (hence vertere stilum, to correct). Although there is no very early record of the use of quills as pens, it is obvious that, well adapted as they are for the purpose and to be had everywhere, they must have been in request even in ancient times as they afterwards were in the middle ages. Bronze pens, fashioned exactly on the model of the quill-pen, that is in form of a tube ending in a slit nib (sometimes even with a nib at each end), of late Roman manufacture, are still in existence. A score of them are to be found scattered among public and private museums. The ruler for guiding ruled lines was the κανών, canon, regula; the pencil was the μόλυβδος, plumbum, the plummet; the pricker for marking the spacing out of the ruled lines was the διαβάτης, circinus, punctorium; the pen-knife, γλύφανον, σμίλη, scalprum; the erasing-knife, rasorium, novacula.
Inks.—Inks of various colours were employed from early times. The ink of the early papyri is a deep glossy black; in the Byzantine period it deteriorates. In the middle ages black ink is generally of excellent quality; it tends to deteriorate from the 14th century. But its quality varies in different countries at different periods. Red ink, besides being used for titles and colophons, also served for contrast, as, for example, in glosses. In the Carolingian period entire MSS. were occasionally written in red ink. Other coloured inks—green, violet and yellow—are also found, at an early date. Gold and silver writing fluids were used in the texts of the ancient purple vellum MSS., and writing in gold was reintroduced under Charlemagne for codices of ordinary white vellum. It was introduced into English MSS. in the 10th century.
Authorities.—H. Geraud, Essai sur les livres dans l’antiquité (1840); E. Egger, Histoire du livre depuis ses origines jusqu’à nos jours (1880); T. Birt, Das antike Buchwesen (1882) and Die Buchrolle in der Kunst (1907); W. Wattenbach, Das Schriftwesen im Mittelalter (1896); K. Dziatzko, Untersuchungen über ausgewählte Kapitel des antiken Buchwesens (1900); J. W. Clark, The Care of Books (1901); W. Schubart, Das Buch bei den Griechen und Römern (1907); and generally the authorities quoted in the article Palaeography. See also Textual Criticism. (E. M. T.)