Writing Systems and Manuscripts
by Rochelle I. Altman
[Dr. Rochelle Altman is a specialist in ancient phonetic-based writing systems. Her publications include _Psalms from the Paris Psalter: Psalm 22(23)_, 1993; _The Diplomatic E-Edition of Beowulf_ (MS emulation — 4 editions) 1994; and _DIPLOMAT_: Computer Application for creating diplomatic editions, 1995. She is also co-coordinator of the IOUDAIOS-L discussion list.–JRD]
Writing systems are systems in the precise dictionary meaning of the word: “A set or assemblage of things connected, associated, or interdependent, so as to form a complex unity; a whole composed of parts in an orderly arrangement according to some scheme or plan.” The interconnectedness of a writing system means that we must look at a document as a whole. Within any given writing system, be it Eastern or Western, Ancient or Modern, each component holds meaning beyond that of the written word. As the primary purpose of writing is to control the power of the word, a writing system is an integral part of a given culture. A writing system gives us social, political, and religious information about a culture at any period in its written history.
_The Law of Parsimony_ states that a system or organism uses only what is needed — no more. The Law of Parsimony works in languages and writing systems as much as it does in the hard sciences. A writing system consists of a _necessary_ number of sub-systems, none of which can be left out. Unlike the sub-systems of, for example, a modern air- plane or house system, we can list the parts of a writing system in one short paragraph. These sub-systems are: a finite symbol-set system, a prescribed script system, a limits system, a mensural system, a size system, a punctuation system, a spelling system, a comprehension system, a format system, and a content system.
Perhaps it would be easier to understand the sub-systems of a writing system if we look at the sub-systems of a modern house. A heating or plumbing or electrical system is composed of an interdependent series of sub-components that make up the whole. Each whole sub-system is designed to function as a self-contained unit; however, it is also designed to function within the complete house-system. If the designer ignores the roofing system, no matter how modern and functional the other sub-systems are, the house will collapse — because the roofing sub-system secures the wall system. In turn, the wall system holds the plumbing, electrical, and heating systems in their assigned places. Writing systems are of the same order; each sub-system is designed as a self-contained unit intended to function in its appropriate place within the complete system. No matter how well designed a script system, if we ignore the limits or comprehension or symbol systems, the writing system will not be readable. As with a house system, we must take into account every part of a writing system.
Similarly, for any system, the components within a necessary sub-system need not be the same for each design; there is always a choice. Just as one designer will specify electric heating for a house system and another gas, a designer of a writing system will choose from among limit systems or format systems or script systems.
There are always choices to be made. It is these choices that give us information about the society within which the writing system functions. If, for example, one group uses totally bilinear limits and another group trilinear ones, this tells us a great deal about the two societies.
Limits, the framework of a writing system, are designed to distinguish semantic units. Limits follow naturally from the decision to concatenate graphic symbols to create clusters of expressions. Whether written horizontally or vertically and read left-to-right, right-to-left, top-to-bottom, or bottom- to-top, all writing systems assign limits.
Nor does the direction of writing (and reading) matter; vertical or horizontal, all writing systems are primarily bilinear (two lines). The outer framework of Western writing systems consists of two horizontal lines: an upper and a lower limit. This structural foundation confines the writing zone (the writing space) to two possible formats:
1) a central area with vertical and horizontal movement within the zone,
2) an entirely filled zone.
Our modern capitals use bilinear limits. Bilinear scripts and fonts *fill* the entire space between the upper and lower limits. In palaeographic terminology, these lines are referred to as the headline (upper limit) and the baseline (lower limit). Because they fill the writing zone, bilinear scripts are tightly constrained on the horizontal plane and cannot move on the vertical plane at all.
Because the symbols fill the entire space, bilinear limits are necessarily static. The main purpose of a totally bilinear limit system is to confine and constrain the written word. It intentionally “freezes” the words into an unchanging form to preserve the magical power of the word and to control people and things. Consequently, magical-mystical oriented societies choose strictly bilinear limits. The Egyptian system was totally bilinear, as was the Etruscan, the Official Roman Imperial, and the Official Neo-Babylonian.
Trilinear (three lines) scripts are written *within* the two limits. They have three positions, upper limit, central space, and lower limit. These scripts have two upper limits. The outer upper limit and the upper edge of the central space. The symbols hang from these upper limits, that is, symbols are written from the upper lines downwards. In these systems, the symbols move both vertically and horizontally within the central zone. The amount of movement within trilinear limits is constrained by the proportion of the central zone to the upper and lower limits designed into the system. In turn, these constraints create a central writing zone, an internal area between the upper and lower limits wherein the bow (the round part of a ‘p’ or ‘b’) overlaps.
Trilinear limit systems are practical and dynamic. Speech itself is dynamic; duration, stress, and phone (quality) change as words are vocalized. The main purpose of trilinear systems is to record speech as-spoken; the speaker/writer’s voice. Trilinear limits are designed to allow the symbols to move up and down and from side to side within the three limits in imitation of the patterns of speech. In antiquity, the “voice of authority” was a literal, concrete reality, not a metaphor. Modern writing limits are quattrolinear (four lines), a refinement of the trilinear limit system.
Quattrolinear scripts are written *between* four lines. In palaeographic terms these limit lines are called the ascender, headline, baseline, and descender. Quattrolinear scripts are trilinear scripts moved downward to the former central zone to accommodate ascenders, such as an ‘l’, within the horizontal upper and lower limit lines. In these systems, the ascender line is the former upper limit and the descender the former lower limit of the trilinear systems. The headline is the former upper and the baseline the former lower limit of the central writing zone. In England, until well into the seventeenth century, quattrolinear systems retained the movement of the trilinear systems. Like trilinear systems, quattrolinear limits permit the inclusion of variant forms. Variant forms originally recorded different phones of the same symbol; today, “lowercase letter” is our name for one of the variant forms.
The Sumerian and Akkadian cuneiform writing systems use trilinear limits. So do all of their descendants, including the Ugaritic, Phoenician, Pre-Second Century CE Hebraic, and the Ionian and Western Greek systems. In all these systems the symbols hang from the top line and move up and down and from side-to-side within the writing zone.
When we examine ancient texts, if the design of the writing system calls for totally bilinear limits, we are dealing with a mystically oriented society; if they use trilinear limits, we are seeing a practically oriented society. This distinction between magical-mystical and commercial-practical becomes of even greater importance following the arrival of Christianity, since each choice of a component sub-system gives us religious information about party affiliation.
Although the components of a writing system are an integral part of any such system, the *purpose* of a writing system can change. There is one essential difference between modern and ancient writing systems; this difference is of critical importance to our understanding of ancient texts. Modern writing systems are semantically based; their primary purpose is to transmit data, the content. Ancient writing systems are phonetically based; their primary purpose is to record speech, the words as spoken along with all those components that identify the text as *that* text.
Accustomed as we are to semantic-based systems, we erroneously assume that the size, format, and script of a text are irrelevant; hence, scholars working with older texts concentrate on the content. Content, that is, the text itself, is certainly important as the content, determines:
- 1) The choice of script,
- 2) The size of a page, and
- 3) the format.
These three elements, in turn, are constrained by religious and political affiliation. The importance of the relationship between these components of a writing system and a given culture can hardly be overemphasized. The correct script, size, and format was required for the content to be accepted as authoritative. It still is today.
Writing systems are perhaps the most conservative of all cultural artefacts. Semantic-based or no, our modern writing systems use hierarchies of scripts, sizes, and formats to identify a text. These components of a writing system are still constrained by political or religious affiliation. All of these interdependent components of a writing system are important today. None of them are modern in origin.
Scholars somewhat wistfully comment upon the diversity of the types of Graeco-Roman Egyptian documents and how even the same class of document will vary according to the time period in spite of the fact that the ancient legalese is quite formulaic in character as well as rather rigid and limited as to types of format permitted.4 The relationship of the size of a document to its content is as rigid as the formats and formulaic language.5 While it would certainly be easier for modern scholars if we could simply separate documents by size, unfortunately, it is not that straightforward.
Because each culture adapted the Phoenician writing system to its own needs, this led to a multiplicity of authoritative and official formats and sizes — each characteristic of a given locale. It all started, however, back in the third millennium BCE at Sumer.
The Sumerians were very methodical: there was a place for everything and everything was in its place. Archaeologists digging at Sumer found the archives nicely organized — by size. Official documents, clay tablets in this case, were stored in baskets. Each basket was carefully indexed as to contents, and the tablets for each type of transaction were all of one size and one format. Size and format depended upon the content of a document.
The oldest examples of writing from Sumer are on very small clay tablets — about 1-1/2 to 2 inches long by 3/4 to 1-1/4 inches wide. As technology improved, the size of the writing surfaces grew. Over the aeons, the association of a certain size and shape with respect to content solidified; first to custom and finally into prescription. This has had profound and lasting effects upon the concept of authoritative or official documents. The largest texts were always the formal documents issued by the ruling government: The Law.
Although clay is cheap and easy to obtain, it is heavy. Its weight limited the size of an archival tablet to what one person could conveniently lift. Accordingly, the largest tablets we find are the ancient law codes: 14-15 inches in height by 8-9 inches in width. Because archival clay tablets were baked to harden and preserve them, the edges did not crumble and there was no need to allow for margins. The tablets are entirely covered with writing. By the time leather or parchment began to appear, the relationship of size to content had already become a rule. Parchment, however, requires a margin to protect the written text. Accordingly, the writing area retained the — by now — rules for size relative to the content with the addition of a white space margin. The size of the reading area remained the same — 14 inches high by 8-1/2 inches wide.
Meanwhile, on the African side of the Ancient Near East, the Egyptians were writing on papyrus, a material made from the leaves of a reed-like plant. Papyrus is produced by laying strips of the inner layers of the plant (pith), one edge on another. The height of a papyrus roll was limited by the height of the strips of pith peeled off the inner layers of the plant. A leaf 12 inches high was costly while a leaf 14 inches high was very rare. A papyrus roll 9 inches or more in height was always a luxury item.
The Greeks had originally received their writing systems from the Asian side. When Alexander and his army conquered Egypt, they brought with them the Asian Near Eastern rules governing size. These size rules created a problem because they required that the law codes and government orders be the highest and the largest documents. Papyrus rolls high enough to attain the “correct” 14 inch size were in very short supply. The Greeks solved the problem by rotating the direction of writing ninety degrees. Greek official documents were 9 to 10 inches high by 14 or more inches wide. In turn, when Rome conquered Egypt from the Ptolemaic Greeks, the Romans adopted the rotation; they also increased the width of their legal documents to show that they were the ones in charge.
The writing material forced a ninety degree rotation in order to maintain the correct size hierarchy. This simple physical difficulty played an important role in creating a distinction between traditions. By the third century BCE the great diversity among authoritative and official formats and sizes crystallized into two primary streams: Semitic and Graeco-Roman.
On one side we find the Semitic descendants of the Phoenician system, which used different official scripts, but retained ancient practices almost intact. In this tradition, authoritative official texts were written in narrow columns, by utterance or breathings, that is, as spoken, and suspended from the upper writing limit. For example, the Hebrew practice was (and is) to divide the 8-9 inches reading area into two narrow columns, each approximately 3-1/2 to 4 inches wide. Even after the codex gained popularity over the scroll, the format remained essentially unchanged. Non-official, yet authoritative, documents were written in broad columns. Law codes were 14 inches in height with a reading area of 8-9 inches in width, _plus_ margins. Writings were somewhat shorter, around 11 inches. Secular documents, such as tax receipts, were smaller yet and varied in size depending upon the type of tax. Harking back to Akkad, deeds of sale, for example, were always narrow, around 3 to 3-1/2 inches in width, but varied in height depending upon the status of the seller.
On the other side we find a combined Etrusco-Roman-Greek tradition. Although all the heirs of the Phoenician systems were “essentially” bilinear, Etruscan was completely so. Formal texts were written from right-to-left and employed the medial point as its primary word divider. Borrowed from the Etruscan, Latin was also bilinear. There is little question that Etruscan influence on Roman writing systems was very strong: Old Roman cursive writing is taken directly from the Etruscan symbol set.
The earliest forms of punctuation used at Ugarit were a medial point as a word divider and a bar as a “sense” divider.6 Etruscan occasionally employed a double point much like our modern colon to indicate “end phrase,” but Latin retained the older form of word division by medial point and bar as sense divider. Eric G. Turner comments that under Greek influence, by the third century BCE, the Romans changed from point as word divider to writing in _scripto continuo_.7 Nevertheless, Giessen, Universitatsbibliothek, Papyrus Ianda, 90, which uses the medial point and bar, shows that these components were still in use in Rome as late as the first century CE.
Greek contact influenced Rome at an early date; but, contact is a two way street. Although Imperial Rome admired and imitated Greek culture, the evidence shows that Romans clung to their ancestral official formats and sizes inherited from the Etruscan tradition. Instead of Greek traditions influencing Roman documentation, with the Roman expansion to the east, official Greek sizes began to conform to Roman standards. This conformance naturally became more rapid after the Greeks lost their dominant position. Eventually, the strictly bilinear Etrusco-Roman official sizes and formats fused with the trilinear Greek authoritative ones. The combined Latin Graeco-Roman official documents appear written in broad columns and in _scripto continuo_. Official single sheets were 12 inches high by 14 or more inches wide while official papyrus codices were 12 inches in height by 8-9 inches in width, *including* margins. Writings were 9-9-1/2 inches in height. Authoritative, but non-official, texts appear in narrow column format. In this tradition, the size of a tax receipt also depended upon the type of tax and the issuee. (People had to pay for the papyrus in their receipts. A typical low status receipt runs 3 x 5 inches.) Deeds of sale retained the ancient Akkadian practice and appear as very narrow leaves of papyrus.
Some sizes, however, are the same no matter what the political affiliation. A single size and format of document appears in both traditions. These texts run around 8 to 8-1/2 inches in height by 4-1/4 inches in width, or roughly a sheet of modern letterhead paper folded in half horizontally. In size, the resulting folded paper emulates the writing surface of a wax tablet.
Wax tablets were the scratch pads of antiquity. Wooden frames filled with wax, they were used to teach children to write, for a student to take lecture notes, and for an author to jot down ideas. Easily erasable for use at another time, the size of a wax tablet became the signal for disposable text. When paperback books first appeared, their size, which was that of the ancient wax tablets, signaled “entertainment,” disposable material. It still does.
In Graeco-Roman practice, the size of a document within a class depended upon the perception of the social status of the originator. While during the early centuries CE perception of status generally coincided with actual status, these carefully maintained distinctions eroded along with the Roman Empire. In all areas, however, the largest documents were always The Law. Under the Imperium, The Law was Roman Imperial; under early Christianity, The Law was the Pentateuch.
Today, which tradition we follow depends upon when and from whom an area received its writing system. The continental areas at the heart of the old Roman Empire followed the Graeco-Roman tradition. English sizes followed the other tradition, a tradition that was taken to the colonies in what is now the United States. In ninth- and tenth-century England, official secular law codes, such as AElfred’s, were 12 inches in height. During the late Medieval and Renaissance periods, pipe rolls, legal records, were 14 inches in width by the length of the roll. Non-religious (secular) authoritative texts were written on leaves 11 to 12 inches in height to distinguish them from The Law. Today authoritative but not official texts, reference books, such as the _Encyclopedia Britannica_ or the _Oxford English Dictionary_, are 11 to 12 inches in height. Legal writing pads are still 14 inches in height. We still automatically associate size with authority.
The formats of our modern texts are also governed by these two ancient traditions. We still have a hierarchy of formats. Prose texts use a format differentiated from poetic texts. We associate multiple columnar formats with newspapers, newsletters, and encyclopaedias. Mathematical and chemical formulas are always printed in a different format from the body of a text. Our responses to format are as automatic and unconscious as our association of size with authority.
The reading width in both traditions remained 8 to 9 inches in width. The recurrent dimension of 8 to 9 inches wide is not arbitrary. It is a result of human physiology; the maximum width for the scanning human eye to process and comprehend data. While the ancients did not have modern scientific research to confirm this point, they did have pragmatic knowledge. These dimensions were retained when writing transferred from clay to scroll and thence to codex: 8 to 9 inches is the normal reading area of an open scroll – as it was of cuneiform official tablets. This influenced the layout of the codex, early books. It still does today.
Although all of our modern Western printed books tend to use “Roman” type fonts, we can still tell at a glance whether a text is in English or in a Romance Language, in German or in Icelandic. Greek and Arabic texts are equally easily distinguishable, as is anything in Chinese or the Cyrillic scripts. One term on the list of sub-systems may appear odd; nevertheless, “prescribed” is correct. As Jack Goody points out, scripts are an integral part of a culture’s identity.8 In antiquity, scripts were a people’s visual statement of independence and identification; We have national flags. Flags are our visual statements of identity. We get very upset if a change to this visual identifier should be suggested from within. We fight fiercely when an outside power enforces a change upon us. We consider it a public shame if a foreign flag should fly over our territory. Yesterday, our ancestors fought fiercely to retain their visual cultural identity, their identifying script.
Scripts are the graphic representations of a given symbol set. A symbol set and its graphic representation, a script, are not the same thing at all and should not be conflated. All writing systems, be they alphabetic, mathematical, logical, or musical, use a carefully defined, finite symbol-set. All members of a symbol-set are mnemonic aides; they were back at Akkad, they still are. The symbol <, for instance, is a mnemonic for “less than”; ‘A’ is a mnemonic for the phone. The choice of mnemonic- to-symbol is, in fact, infinite; but an infinite set would not be very useful. The Law of Parsimony requires that the minimum number of symbols be used to do the job. The number of alphabetic symbols depends upon the needs of a language.
Sumerian was a completely syllabic language and used around 770 symbols in their set. The Sumerian system worked on the rebus principle, that is, B+4 = “before”, it also used super- and sub-scripted markers to avoid the need for extra symbols. The Sumerian symbol-set, however, was not phonetic-based.
Sumerian is an isolate language, unrelated to the Semitic languages found elsewhere in the Mesopotamian valley. Akkadian is a member of the North-East Semitic language family; a consonantal language, it did not need as large a number of symbols in its set. The Post-Sargonic symbol set consisted of a restricted group of monosyllabic signs that included some common Sumerian syllabograms. Perhaps 100-150 symbols were required for everyday transactions in the Akkadian period. Besides the symbols, the Akkadians used an external subscript marker system to show whether the symbol stood for itself or as a phone in a semantic unit. Sense dividers consisted of spaces. Paragraphs were separated by larger spaces. While Akkadian scribes tried to fit a sentence on one line, if words ran over, they would indent the right-hand margin to show continuation and write the remainder on the next line down.
The Sumerian syllabograms were reserved specifically for xenographic exchange, that is, the use of style A in a text written in style B. In Akkadian texts, for example, if the reference to a king were secular, they used the Akkadian graph; if the reference was spiritual, they used the Sumerian one. Writing systems, as we have already seen, are extremely conservative. All Western writing systems are descendants of the Phoenician writing system as passed down and modified from those in use at Ugarit, Akkad, and Sumer.
For the next 3,500 years, no matter where or when, among any peoples who had borrowed the North Semitic writing systems, *archaic* graphic-symbol forms were reserved to distinguish a spiritual reference from a secular one. We find this purposeful distinction in Neo- Babylon and in the use of the Paleo-Hebraic tetragrammaton among the Dead Sea Scrolls. The secular-spiritual distinction appears in Greek use and in the Latin use of the _nomina sacra_. This distinction shows up in the late 9th century CE in England where they used ‘cyning’, with a ‘c’, for the secular realm, and ‘Kyning’, with a ‘K’ for the spiritual one. When during the late 10th century ‘kyning’ became ‘king’, the need for the distinction was filled by the introduction of yet another form of ‘k’, the ‘Christos-K’, to perform the same function. Our modern use of italics comes from the use of xenographic exchange to indicate ‘foreign word’ in early texts. Xenographic exchange enabled the reduction of the necessary number of symbols in a set because it permitted the introduction of variant forms.
1750 years after Sargon’s innovative reduction of the Semitic symbol-set, some astute Ugaritic system designer saw that, if the symbols were *alphabetic*, the set could be reduced to thirty or less symbols — including variant forms. The number of symbols depends upon the level of formality. Official texts are written from left-to-right and use the full thirty symbol set; informal texts are written from right-to- left and employ a form of shorthand. The Ugaritic symbol set has separate graphic symbols for a number of consonant-plus- vowel phones. Among these differentiated graphic symbols we find three distinct graphs of ‘alep’ that record three distinct a-phones and two distinct forms of heh (e-phones). The alphabet was born at Ugarit. Ugaritic, however, still used a clumsy cuneiform script-system to graphically represent their symbol-set.
Phoenicians adopted the Ugaritic alphabetic cuneiform system and refined it even further. Like Ugaritic, Phoenician is a North-West Semitic language. The Phoenicians created a new script design, a simplified version of the triangular cuneiform of Ugarit. They also invented a different way to record the different qualities (phones) of the vowels. They reduced the number of symbols to twenty-two plus variant forms. The result was the first true alphabet.
The Phoenician solution was very simple. Instead of relying on completely different graphic forms to indicate quality, they designed a script that uses variant forms of a basic graphic symbol shape. Each variant of a basic shape indicated a different phone associated with *that* shape. Like Ugaritic, Phoenician used three different a-phones attached to the consonant aleph. In the Phoenician system, a different vowel phone is associated with each of the variant forms. Greek borrowed the first form for their alpha and Etruscan the third for their ‘A’. Which of the three forms a group borrowed from the Phoenician system tells us something about their language. Likewise, heh (modern e) as well as vav (modern u or v) had variant forms. While it may seem odd that one Semitic form should give different modern letter symbols, the Phoenician, and likewise the Hebrew, symbol may be either vocalized (shuruk) or consonantal (v). The two remaining vowels of our modern alphabet, ‘i’ and ‘o’ remain unchanged — but for different reasons. Although a consonant, the Semitic yod plus vowel apparently sounded much like the Greek iota (jot) and Latin i, that is, the modern English long ‘e’ phone. The symbol ‘o’ (‘ayin’ in the Semitic languages) is the measurement (mensural) base for all script design.
Although the art of script design is extremely complex, the basics are very simple. Proportional scripts (and fonts), bilinear, trilinear, or quattrolinear, are designed relative to the width of the minim and the symbol ‘o’. The unit of measure is the minim — the width of a vertical stroke — as determined by the width of the writing instrument. In palaeographic circles, the term “minim” sometimes refers only to the upright stroke of an ‘m’ or ‘n’. The minim, however, as the basis of mensuration in all Western script systems refers to the width of the upright stroke — whether tall or short.
The writing instrument dictates the width of the ‘o’. An ‘o’ is the width of a “standard” space. The ‘o’ determines the distance between the baseline and the headline. This may appear complicated, but it is not. These design specifications can be summed up as follows:
- The writing instrument establishes the width of a minim.
- A minim is the width of a vertical stroke.
- The minim is the unit of measurement (scale or method of mensuration).
- The writing instrument establishes the width of the ‘o’.
- The ‘o’ establishes the width of a standard space.
- The ‘o’ establishes the distance between the baseline and the headline.
The number of minims per ‘o’ varies with the design, but is usually around three. Within the ancient writing systems, the width of the minim is the measure for horizontal movement; the height of the ‘o’ (ayin in Hebrew and Aramaic fonts) is the measure for vertical movement. Because a written text has a built-in scale, the scribe only has to change the width of his or her pen to change the scale (Figure 5). The writing instrument will determine the distance between headline and baseline when a completed font design is actually performed by a scribe.
The standards were set more than five thousand years ago and apply equally well to an inscription on an ancient stele, a Roman Monumental inscription, a tiny fourteenth century book of hours, and a modern printed paperback book. These standards are used by calligraphers to this day and are still the basis of professionally designed proportional scripts, typefaces, and computer fonts. Although there have been attempts across the centuries to vary this basic script design, no-one has yet found anything better for the creation of well balanced, pleasing, and readable fonts. Our modern scripts and fonts are all descendants of the new Phoenician script design.
Scripts are carefully designed finite systems. A script is composed of a limited number of graphic symbols designed to work together within the constraints of a system design. This holds true whether referring to 24th century cuneiform, 9th century BCE Phoenician, 3rd century BCE Hebrew, 4th century CE monumental Greek, 8th century Official Latin, or modern computer fonts. Scripts do not simply develop, nor are they merely collections of various available forms. Only when viewed from a distance of millennia can scripts be said to develop. Development implies a continuum; it suggests that one letter form changes here, another there, until finally a totally new script arrives. Methods develop; scripts do not develop — they mutate. There may be an unfinished quality to random shards, but ancient formal or official inscriptions and tablets display fully formed graphic symbol sets designed to work within their respective writing systems. Ancient cuneiform examples exhibit distinctions between scripts and fonts, their mutated descendants (Figure 4).
Graphic symbol (a) in Figure 4 illustrates the oldest known form, the syllabogram (a single syllable graphic symbol). The symbol was drawn on the clay. In illustration (b), the symbols rotate 90 degrees and angularize to suit contemporary writing materials: clay and wedge-stylus. B is a script. Example (c) is a font, that is, a mutation of script (b). Example (d) is a new script design, and (e) is a font, a mutation of new script (d).
The fact that scripts mutate rather than develop is a very important distinction. The mutations allow us literally to see history. The scripts and their fonts tell us something about the people who used them. We can, for example, determine political and religious policies from whether a group designed a new script or used a font.
When in the mid-twenty-fourth century BCE the Akkadians conquered Sumer, they adopted the Sumerian writing system. While the scripts make it clear that the Sumerian national script was replaced, they also record Akkadian policies. We know from documents that Sargon I was revered and his reign looked back on as the example of the perfect kingship. The scripts tell us why: Sargon I thought in terms of unification rather than despotism.
Sargon’s new script had to be recognized by his Sumerian subjects as “official,” yet it also had to have his personal stamp to differentiate his voice from his predecessors’. The result is a mutation of the Sumerian official script. We can see Sargon’s approach in font (c) of Figure 4.
On the other hand, the Old Babylonians (Biblical Chaldeans) thought in terms of complete control and despotism. They pulled down the “national flag” of Sumer and Akkad and replaced it with their own. The Babylonians designed a completely new official script (d) and official voice(see Figure 4 again). The Assyrians again clearly thought in terms of unification. Their official script is a mutation (e), a font, of the Babylonian official script(see Figure 4 again). While the Old Babylonian and Assyrian conquerors changed the script, they nevertheless continued to use the cuneiform writing system.
The Northern Semitic languages separated into North-West and North-East branches. The simplification process that we see in the North-West Semitic texts is the norm and usual. This simplification process appears to proceed at a slower pace among the North-East Semitic peoples. A professional “academic” scribe in the Neo-Babylonian Empire (625-538 BCE) still had to know around 600 symbols. The old saying that appearances are deceiving is more than a cliche in this case. The new Babylonian rulers were not direct descendants of the old ones. The superficial conservatism really is another form of borrowing authority on their part as a tie to the Old Babylonian Empire: *archaizing*. The borrowed archaized cuneiform was used for “official” legal and religious documents, but not for everyday use. For foreign affairs and commercial transactions, the Neo-Babylonians used the writing system designed and implemented by the Phoenicians.
Commerce had held the primary role in the distribution and dissemination of writing systems. For nearly 3000 years, scripts had been the national flags of peoples; they signaled cultural identity. In the fourth century CE, with the recognition of Christianity as a legitimate denomination, scripts followed religion. Once only the signal of cultural identity, the scripts now also signal religious party affiliation. A different script *always* means a different religious group. The absoluteness of script-as-authority cannot be stated often enough. Prior to the invention of the printing press, liturgical letter forms had religious significance and their use, shape, and script-model were prescribed, party- affiliated, and obligatory. There exist many examples of the rigidity of script-as-obligation. One glance is sufficient to determine the religious affiliation of the Syriac parties. The Syriac scripts, an offshoot of cursive Aramaic, divide into numerous styles according to party. Syriac texts, translated from Hebrew and Greek originals, were originally written in the authorized script called Estrangela. The sub-divisions of the Syriac party were Nestorian, Assyrian, Jacobite, and Melkite. Each of these sub-divisions had their own mutations, fonts, of the official Estrangela script. Later, after the Muslim conquest of the area, Syriac texts were written in another official script called Garshuni. Transcribed Greek scriptures (rather than translated) are written in two other Syriac sub-divisions: Mandaean and Manichaean. Once a group has chosen an official script, nothing which is not written in that model, to quote the Parsis, “can claim to be considered as part of the sacred literature.” It is no accident that the two major surviving communities of Parsis use two distinct styles: the older straight ‘Indian’ in India, and the Persian influenced ‘Iranian’ in Iran. Kufic was the official script for the Qur’an until the eleventh century. Syriac remains the liturgical script of the Kurdish remnants of the Assyrian sub- party. To this day, we unconsciously associate language or political affiliation with script.
We say that we have only twenty-six letters in our modern English alphabet; a moment’s thought shows that this is obviously a false statement. At a minimum, we have fifty-two alphabetic symbols: twenty-six capitals and twenty-six lower- case variants of the capitals. Similarly, within our Western “Roman” fonts, we still use xenographic exchange to create a hierarchy of script styles. Words printed in all capital letters normally indicate a heading. Bold face type tells us that this is a sub-heading or is of special note; italic type indicates foreign language or book/journal title. These hierarchies of script styles are impossible to achieve without xenographic exchange and variant forms. We use variant forms on a regular basis: We unthinkingly associate type style with subject matter.
We associate authority with a type style, too. Although we do not consciously make decisions on whether a font has those little triangular additions we call ‘serifs’, the text must be printed in a serifed type to carry authority. Sans-serif fonts signaled ‘advertisement’, material not to be taken seriously, back at Pompeii. Today, we still associate a sans-serif type style with ‘advertisement’.
Because writing equaled power and control, we see reforms every time there is a change of power structure. We may even refer to this phenomenon as the “Winner’s Standard Operating Procedure.” The meaner the origin, the more grandiose the reform. The new ruler must have a better background than any of his subject peoples, so genealogies are created for him, for example, Cyrus. If the new ruling power is a council or city, they write histories to create a background for themselves, like they did in Rome. The winners reform the language to meet some nonexistent “ancient” standards, as we can see under Augustus Caesar or Charlemagne or AEthelwold or Queen Victoria. They change the “official” script to a new one. If they cannot create a good background for themselves, they may purposely “archaize,” appropriate an older “official” script, as their national script, like the Neo-babylonians. The “Winner’s” cycle repeats again and again; it has from the very beginnings of recorded history. It is a common way of saying that we, too, have a good background — and a history.
Early Christians rejected the Official Roman writing system as “pagan”: they used the Semitic system that we see in the Qumran scrolls. Diversity, however, set in with great speed. By the 4th century, either accepting or rejecting the influence of Neo-Platonism and Graeco-Egyptian mysticism, Christian parties looking for their own official scripts and voices adopted these already established, ancient, Winner’s techniques for their own writing systems.
By the time we finally have enough documents to compare against each other, we find that the North African-Semitic parties were adhering to the Semitic traditions for religious texts; on the Alexandrian-Roman side, these parties adopted the Imperial Roman traditions. These sizes and formats remained when the codex became the most common form of official Christian document.
The codex, as opposed to the scroll, tends to create a canon. What had been contained in separate documents is now placed together between two covers. A codex gives us a surprising amount of information about religious parties, for which texts are included between the covers tells us what texts were considered authoritative by a given group. The content, however, does not tell us *which* group; their writing system does give us that information.
Among the earliest of the Greek Bibles, we find two codices from the early 4th century. One, the Codex Sinaiticus, was at the monastery of St. Catherine in the Sinai until the 19th century. The official story is that Constantine Tischendorf pursuaded the monks of Saint Catherine’s to give the codex to Tsar Alexander II of Russia in exchange for protection. Tischendorf prepared a small edition, based on around 43 pages, which was published at Liepzig. He then is said to have given the Codex to the Tsar. The Codex was in Leningrad until 1933 when 347 of its estimated 730 leaves were legally purchased by the British Library (B.L. Additional MS. 43723), where, quite understandably, the manuscript has pride of place. Fragments of 3 more leaves are in Saint Petersburg. About a dozen more leaves came to light at St. Catherine’s after a fire in 1975. We have good reason to doubt the official tale; the story they tell at Saint Catherine’s is quite different. They say that Tischendorf claimed he wanted to study the Codex at his leisure, “borrowed” it, and never returned. The fact that another 43 leaves are in the University Library at Leipzig, a suspicious number and a suspicious place in light of the small edition produced by Tischendorf and published at Liepzig, lends support to the version told by the monks at St. Catherine. The Sinaiticus probably was produced at Alexandria.
The other early 4th century Greek Bible is the Codex Vaticanus (Vatican, MS. Greek 1209). The Vaticanus was a gift. It may have been brought to the Council of Ferrara-Florence in 1438-39 by the delegates from Constantinople. Its place of origin appears to have been Upper Egypt.
While the two codices have much in common, there are some substantial differences. The first major difference between these two Bibles is their content. Both codices originally contained the texts of the Old and New Testaments. Due to a misreading of the texts by the first editor, for nearly a century it was assumed that The Vaticanus included part of the Pseudepigraphic 4 Maccabees. Modern research has shown that The Vaticanus never contained anything besides the Old and New Testaments, including the Pauline material. The Sinaiticus, on the other hand, held additional material — including the Epistle of Barnabas, the Shepherd of Hermes, and the Old Testament Pseudepigraphic 4 Maccabees.
We cannot really speak *in general* of a ‘canon’ at this date, for canonical material fluctuated wildly depending upon time and affiliation. On the other hand, both codices are written on vellum, the finest grade of parchment; both are written in Monumental Biblical Book hands. The vellum and the formal book hands places both codices into Class 1 books according to known Ancient standards and price lists found in Roman discussions on book preparation. Class 1 books are authoritative; thus, we already know that we are looking at a difference in what was considered *their* canon by two different groups.
Both codices are approximately square. The Vaticanus is roughly 10.7 inches in height. The Sinaiticus is written in four columns per leaf; the Vaticanus in three. Both are written in breathings, that is, the number of syllables that can be said in one breath. The two codices also share another fate: both have suffered from later editorializing, such as, the addition of punctuation, or the insertion of later “corrections” to bring the text into conformity. (Editing-to- conform is another of the Winner’s Standard techniques.) The Sinaiticus was written by three scribes who clearly were trained at the same scriptorium. The Vaticanus, originally written by two scribes (A and B) using the same script, was ‘freshened’, that is, sometime around the 10th or 11th centuries, a third scribe (C) went over the entire codex overwriting the text letter-by-letter. Freshening usually occurs when the original ink is fading and the text is becoming difficult to read. Because Scribe C left one leaf as it was, we know that, while this Scribe added the accent marks, he simply overwrote the original script and did not change the design.
All Greek fonts are related as they are all descendants, mutations, of the same script. The differences are in the details; the second major difference between the codices is their fonts. These two Biblical fonts are totally different designs (Figure 6).
We can see immediately that we are looking at the canonical texts of two different religious parties. The fonts tell us more: we have here two examples of the age old practice of borrowing authority. The question is, of course, whose authority? If we can determine from whom authority was borrowed and by which method, we have a head start on determining which font belongs to which group.
The first obvious point of difference is that the script of the Sinaiticus uses both thick and thin lines, while the script of the Vaticanus is monoline (one thickness). This difference is a result of the types of pens used to write the texts. Both the type and cut of a pen are part of a script design. While both scripts are written with reed pens, the pen of the Sinaiticus is a chisel-point and that of the Vaticanus a flat-cut nib. A chisel-point nib is cut at a slight angle; when it meets the writing surface, it automatically produces a thick and thin line as the graphic symbols are formed. A flat- cut pen nib meets the writing surface squarely; it will automatically produce a monoline.
We can see the effect of the pen in the form of the graphic-symbols. In the Sinaiticus, the uncial-a is pointed; in the Vaticanus, the form is rounded. Where we see a flattened base on the lobe of the rounded-a form, we can also see the hand of the overwriter: Scribe C. The ‘O’ is also different. In the Sinaiticus the symbol has thick and thin strokes; in the Vaticanus, the ‘O’ is monoline and flattened. Second, as already noted, contact is a two-way street. While Rome borrowed much from Greece, Greece borrowed much from Rome. By the 2nd century BCE, Roman presence in the Hellenic world was well established. By the 1st century BCE, official Roman scripts display thick and thin strokes, serifs, tight kerning (the separation between two letter-symbols),9 and a narrow ‘o’ base (the basis of measurement). On the other hand, official Classical Greek scripts and fonts show monoline strokes, no serifs, loose kerning, and a wide ‘o’ base.
More to the point, the oldest clear example of an authoritative serifed Greek font shows up in a fragment of Deuteronomy dated to the first half of the 2nd century BCE from Egypt (John Rylands Library, Papyrus 111,458, fragments, Manchester, England). Deuteronomy was hardly an official text of the Ptolemaic Greek government; it was an official text of the Jewish population of Alexandria. That a Graeco-Judean authoritative font would have serifs is to be expected: official Square Aramaic fonts are serifed. Further, the serifs of this font design follow the Aramaic practice of heavy serifs as opposed to the thin serifs used by Rome. Other than the serifs in Aramaic style, the font follows Greek practices: it is monoline, the mensural base of this script design is still the wider Greek ‘o’, and the written text still displays very loose kerning.
All of our early examples of Greek fonts with serifs are from Ptolemaic Egypt, not from Seleucid Syria or mainland Greece. Official serifed Greek scripts only begin to replace the Classical Greek sans-serif fonts following the Roman conquest of Egypt. Further, the Greek script designs retained the old wide ‘o’ base until long after Rome was a solid presence in the Eastern Mediterranean basin. The change in ‘o’ base does not occur until around the middle of the 2nd century CE. Our oldest example of this new Greek Book script can be seen in a fragment of Hesiod’s Catalog (Greek poet, lived 8th- century BCE). The script design of the Hesiod fragment incorporates many aspects of the Official Augustan Roman Capitals. It uses a narrower, Roman style, ‘o’ base; as a result, its proportions are quite different. The graphic symbols are narrower and taller in relation to the ‘o’, and, to allow for tighter kerning, rounded forms, such as epsilon (e) and omicron (little o, the mensural base) are much rounder than in older Greek script designs. Although we can see the serifs on the fragment of Deuteronomy, which would suggest a Semitic tradition for the Sinaiticus, upon closer examination, this does not prove to be the case. The designers of the Sinaiticus font followed in the footsteps of Sargon I when he borrowed the authority of the Sumerian graphic design for his voice. The Biblical font of the Sinaiticus clearly borrows its authority from Rome. The placement of the serifs and the design of the font follows the Official Augustan Roman model, not the 2nd-century BCE Ptolemaic Graeco-Judean one. Like the earlier Hesiod fragment, the kerning between letters is tight; the omicron (o) is very round, as is the epsilon, and the letter-symbols are very narrow and tall in comparison with the ‘o’.
On the other hand, the font of the Vaticanus follows the other old practice for borrowing authority and used at Neo-Babylon: archaization. This font design incorporates many features that appear in the oldest existing Greek scripts. The kerning is quite large and the ‘o’ is flattened and broad. The epsilon has an almost straight-back. Forms such as the beta (B) and the lambda ( /\ ) hark back to the ancient angular carved models. The script of the Vaticanus displays deliberate archaization, a return to the authority of Classical Greece.
There were a number of parties who chose Classical Greece as their model. When we think of Classical Greece, we think of Athens and philosophy, literature, and sculpture; we do not ordinarily pay much attention to their love of mysticism and mysteries in religion. Their writing limits, however, should remind us of this point. Unlike Ionian or Western Greek writing systems, the Classical Attic writing system used bilinear limits — or rather — as bilinear as could be managed with their symbol set, what Turner refers to as “essentially” bilinear.10 Both of these early Greek codices use the bilinear limits that were a feature of Attic, Egyptian, and Etrusco-Roman writing systems. This tells us that these codices were not the official texts of any of the ultra-conservative subordinationist parties, such as the Anomaeans. We also can rule out the Montanists, Donatists or any other groups who were strongly against the Hellenization of Christianity. This anti-Hellenism factor reduces the field somewhat and leaves us with this font as the official Biblical voice of either the monophysites or one of the compromiser factions among the subordinationist parties.
We have quite a bit of evidence that the Sinaiticus was produced by affiliates of the Alexandrian-Roman parties. Another Class 1 book, the 5th-century Codex Alexandrinus, like the Sinaiticus, contains material in addition to the Old and New Testaments. Eusebian and Athanasian material was inserted before the Psalms. Now bound in 4 volumes, the first four books contain the Old Testament, plus the Pseudepigraphic 3 and 4 Maccabees, Psalm 151, and the 14 liturgical canticles. The Epistles of Clement are appended to the New Testament. According to the table of contents, the Psalms of Solomon were originally part of the Codex, but they have been lost.
The Alexandrinus has grown in size. Written on leaves 12.6 inches in height by 10.4 inches in width, it is now the height of Official Imperial documents. This codex uses a much broader two column format rather than the four very narrow ones of the earlier Siniaticus.>>11 The text is still written in breathings. New paragraphs are marked by letter-symbols in the white space between columns. The font is a mutation of the Romanized font of the Sinaiticus (Figure 7). The size tells us that the Alexandrinus is an Official and Authoritative document. The font and the inclusion of some of the *same* pseudepigraphic material in the Alexandrinus strongly suggests that, in the Sinaiticus, we are seeing the Official identifying script of the Athanasian faction of the Alexandrian-Roman parties. The Alexandrinus also tells us that pseudepigraphic material, such as 4 Maccabees, was still considered part of the Official Alexandrian-Roman canon in the 5th century.
While we cannot be certain to which party the official script of the Vaticanus belongs, we do know what these three codices tell us: In these volumes we are seeing the battle for authority in action. Even if we did not know the later history and the outcome of the battle, the two Romanized fonts in the other two codices, as well as the size of the Alexandrinus, tell us who won.
Who, what, when, where, why are the researcher’s questions. The answers are visible in the texts; the writings sytems give us social, political, and religious information about a society, even today. The primary difference between our modern approach to a text and the ancient approach is one of consciousness. Today, we do not know why we require the components of our writing systems; we simply use them. The ancients knew the meaning and relevance of every part of a writing system. The texts contain more than words; we would do well to remember this fact. The manuscripts and their writing systems will speak to us, too — if we will listen.
2 For scripts as authority, see Stanley Morison, _Politics and Script: Aspects of authority and freedom in the development of Graeco-Latin script from the sixth century B.C. to the twentieth century A.D. The Lyell Lectures 1957. Edited and completed by Nicolas Barker_. (Oxford: Clarendon Press, 1972).
5 Some of this material appears in from R. I. Altman, “The Size of the Law: Document Dimensions and their Significance in the Imperial Administration,” in _Shifting Frontiers in Late Antiquity: The Self, the Other, and the Law_. H. Sivan, ed. (forthcoming) and is used courtesy of the editor.
6 The “bar” is a line drawn (or inscribed) from the upper to the lower limit. This form of punctuation is not to be confused with the chironomic musical instruction to increase volume or stress: the tifcha/oXeia/virgule. Although the chironomic symbol may be as long as a punctuation bar, it is written above the syllable to be stressed.
9 Each graphic symbol has a ‘frame’, the white space around it. What we are looking at, however, is the white space *between* symbols. There is no paleographic term to describe this area. ‘Kerning’, a term borrowed from printing, already exists and accurately describes this space. We need not invent a new term.
10 Turner, _Greek Manuscripts_, 20. Every one of Turner’s examples includes the statement “essentially bilinear” followed by a list of the exceptions to the “bilinear rule”. 11 By late in the 4th century, Latin documents of the Alexandrian-Roman parties use the Imperial broad column format for official Biblical texts.
Avrin, Leila. _Scribes, Script and Book: The book Arts from Antiquity to the Renaissance_ (Chicago, American Library Association, 1991).
Bagnall, Roger S. _Reading Papyri, Writing Ancient History_. (London, New York: Routledge, 1995).
DeFrancis, John. _Visible Speech: The Diverse Oneness of Writing Systems_. (Honolulu: University of Honolulu Press, 1989).
Falivene, Maria Rosaria and Jennifer R. March, trs. Italo Gallo. _Greek and Latin Papyrology_ (London: University of London, 1986).
Goody, Jack. The Interface between the Oral and the Written, (Cambridge: Cambridge University Press, 1987).
Kenney, E. J., ed. “Books and Readers in the Roman World.” Cambridge History of Classical Literature, Vol 2. Latin Literature. (Cambridge: Cambridge University Press, 1982). 3-32.
McKinnon, James. _Music in early Christian literature_. (Cambridge: Cambridge University Press, 1987).
Morison, Stanley. _Politics and Script: Aspects of authority and freedom in the development of Graeco-Latin script from the sixth century B.C. to the twentieth century A.D. The Lyell Lectures 1957_. Edited and completed by Nicolas Barker (Oxford: Clarendon, 1972).
Taubenschlag, Raphael. _The Law of Graeco-Roman Egypt in the Light of the Papyri: 332 B.C. – 640 A.D._ 2nd ed. (Milano: Cisalpino-Goliardico, 1972).
Turner, E[ric] G. Greek Manuscripts of the Ancient World, (Oxford: Clarendon, 1971).
Wiseman, Donald John. “Books in the Ancient Near East and in the Old Testament.” _Cambridge History of the Bible_, vol. 1. (Cambridge: Cambridge University Press, 1970). 30-47.
Reproduction beyond fair use only on permission of the author.