Michel Davoust (2002), "Mayan script and society from
the second through the tenth centuries", p. 171;
in A History of
Writing: From Hieroglyph to Multimedia, ed. Anne-Marie Christin, pp.
169-178.
Maya writing combined logograms and phonographic symbols, which expressed sound sequences of the shape V or CV. This is of course a syllabic system: there are no phonetic symbols that are just C (contrast Egyptian). Unlike Sumerian-Akkadian there are also no syllabic signs that stand for CVC or VC. This page from the Dresden Codex illustrates some painted signs, integrated with a nonlinguistic scene. The bars and dots are numbers.
|
image from Kettunen and Helmke (2005)
Maya writing survives in a wide range of contexts, including stone monuments, pottery, and books. These stelae are from Copán and are dedicated to the king known as "18 Rabbit". They are dated 731-32, and are quite typical of the stelae used for monumental inscriptions, generally describing the birth, accession, and exploits of rulers. The Mayan word for these monuments, appropriately, is TE TUN "tree of stone". |
A distinctive feature of Mayan (and some other Mesoamerican) writing is the way that glyph blocks are read in paired columns that is, left to right within pairs of glyph blocks and overall in the text, but proceeding down a pair of adjacent columns before moving rightward. The general structure of a Maya glyph block includes a main sign and from zero to several "affixes" to that sign. These "graphic" affixes do not necessarily correspond to grammatical affixes in the spoken language; they're just a way of arranging signs together in a block. According to Grube (1994), in the 1500-year history of Maya writing there are 650–700 recorded signs in total, most of which were limited in use by time or region. Any particular scribe probably never used more than 250 signs. |
This wall panel from Piedras Negras, called "Lintel 3", depicts a night-time chocolate-drinking feast that took place in 749. The fourth king of the dynasty was celebrating his first k’atun, or 20-year period of rule. The column on the left gives the date; note the large "initial series introductory glyph" (ISIG) and the calendrical glyphs (with numbers to their left). |
This painted tomb is from the Early Classic Period, at Río Azul in the northeastern Petén region. The Long Count on the back wall gives the birth date of the deceased as the equivalent of September 29, 417. |
These pages are from the Dresden Codex, one of the few surviving Maya books, largely containing astronomical data and calculations. |
Michael D. Coe and Justin Kerr (1998), The Art of the Maya Scribe, Abrams, pp. 15, 32, 118.
A word like BALAM "jaguar" could be written in a variety of ways, illustrating the possibilities in the Maya script. Since the phonetic symbols stand for CV, the final [m] of [balam], if indicated phonetically, has to be written with a silent vowel. Thus [CVC] is written CV+C(V).
More precisely this word is pronounced bahlam, with an [h] that is not represented in the writing system. |
Besides BALAM, another word that can be found written many ways is PAKAL "shield", part of the name of an important ruler, and frequently found in the inscriptions.
|
Over time, many logographic signs came to represent a phonetic syllable that was based on the pronunciation of the word to which the logogram referred. This yielded a syllabary alongside the logographic system.
Phonetic signs become common only after about 400 CE, in the Classic period when the writing system reaches its fullest development. The restriction to a maximum size of CV can be seen in the adjustments that occur when a logogram is adapted to phonetic use; extra material beyond CV is dropped for the syllabic use. These three signs are used to write pakal "shield", which of course does not mean "net-fish-completion", illustrating the purely phonetic use of the signs. |
An interesting convention in Maya writing is the use of two small dots to indicate that an adjacent syllabic sign should be doubled in the reading. Notice that the repeated signs in these words are written out just once. This convention resembles the use of quote marks (") as "ditto", i.e. the same again.
The numeral "2" is a convenient way to transliterate this special sign, which is often called a determinative since it affects the reading of an adjacent sign. |
This cup was used for drinking chocolate (from the cacao plant, which originated in Central or South America). Pottery objects often included short texts. You can see the dot determinative in the top line on the left side. |
There are two major theories regarding the origin of writing in Mesoamerica, which reached its fullest development under the Maya: these are the Olmec writing of Veracruz and the Zapotec writing of Oaxaca. One theory — which has the widest acceptance — looks to the Olmec region (modern Veracruz), where a number of possibly pre-linguistic representations resemble elements found in Maya writing. The Olmecs were an early advanced culture who may have spoken a Zoquean language. (You may recognize the large stone sculptures of Olmec heads from the Simpsons.) |
It is well established that the Olmec had an influence far beyond their homeland. |
For example, the Humboldt Celt, an ax-like tool from about 900 BCE, contains many symbols similar to Maya glyphs. Some ideas about the meaning of signs are indicated to the right. Note in particular the seeming affixal structure in the rain-god headband, including signs resembling Maya <u>. The interpretation of such inscriptions remains uncertain but is quite suggestive of a connection to the nearby and later Maya culture. |
Justeson 1986
In the Oaxaca region of southern Mexico, early Zapotec monuments contain elements that figure in Maya writing as well, and constitute what is probably the most advanced writing in the early record. In particular, at Monte Albán we find the typical two-column style of arrangement, similar numerals, and main signs with affixes. Stelae 12 and 13 shown here date from 500-400 BCE. |
Many of the danzantes or "dancing" figures (actually, they're depictions of slain enemies!) found on a building at Monte Albán have apparent short texts, often what is probably the person's name near the mouth. The texts are not well understood. |
There are also longer texts, such as Danzante 55. Given the relative proximity of these regions and the nearness in time of the developments, it is likely that multiple influences (including these and others) converged to produce the most sophisticated writing system of the ancient Americas, Maya hieroglyphs. |
John F. Harris and Stephen K. Stearns (1997), Understanding Maya Inscriptions: A Hieroglyphic Handbook, 2nd edition, University of Pennsylvania Museum, p. 117-128.
The only other script of Mesoamerica that appears to have been as linguistically sophisticated as the Maya script and at a slightly earlier date is from southern Veracruz. The general system of writing has been called Epi-Olmec, Isthmian, Tuxtla, or La Mojarra, after the most important text, La Mojarra Stela, a four-ton basalt monument bearing dates corresponding to May 21, 143 and July 13, 156 CE. This inscription is slightly before than the earliest datable Maya monuments. Interpretation of the script is still ongoing, since this stela was discovered only in 1986; a recent decipherment claims it is a Zoquean language. |
La Mojarra has the same Isthmian script as the smaller Tuxtla Statuette, which until recently was unique in the record of writing systems. The statuette is dated March 14, 162, so that there is currently little evidence that Isthmian writing was used for any length of time. This diagram shows the full amount of text, spread around the statuette. |
Michael D. Coe and Justin Kerr (1998), The Art of the Maya Scribe, Abrams, p. 66.
This page from the Codex Borbonicus illustrates the rich illustrations of these codices; it dates from right around the time of the Conquest (1519–1521). There is considerable symbolism but no indication of a linguistically represented narrative. The main image is the god associated with a 13-day period in the calendar; each day is represented by a pair of boxes that as a group occupy the rest of the page. These include day-signs (numbered by dots) and more deities (with birds). The alphabetic writing in Nahuatl and Spanish was often added later to such codices. |
The Aztec writing system was not as linguistically advanced as that of the Maya (despite the fact that it developed later). Representation of placenames was primarily based on the rebus principle, but this had not (by the time of Spanish arrival in 1519) developed into a system that made general phonetic use of signs. These basic roots here, and thus the phonetic values of the signs, are in the first row; the towns have names have related to "water" but the second element is a rebus for an unrelated meaning. Thus while a is used logographically (it actually refers to the root "water" here), each second element is used in rebus fashion.
In Aztec, the "absolutive" suffixes -tl, -tli, and -li that you will see on most words simply mark a noun that does not carry any other suffix; they are not part of the root, and thus not part of the sound carried by the basic sign. |
These remaining examples of placenames also illustrate the use of the sign for water logographically, with additional rebus elements. In most cases, not all elements of the name are actually written, and the "reader" is required to provide the missing information from their knowledge of existing placenames. You should be able to see how the shape or details of the sign convey at least part of the meaning found in the placename.
|
This page from the Codex Nuttall shows the somewhat more connected nature of Mixtec codices. This one dates from the 14th century, i.e. clearly pre-Conquest. According to an interpretation by Alfonso Caso (1965), this page (starting at the upper right and snaking toward the lower left) contains genealogical information. The palace at the upper right shows a man named 5 Crocodile Sun of Rain (the 6 is said to be an error) and the woman 9 Eagle Garland of Cacao Flowers. Because they face each other, they are married. Between them is their marriage year 6 Stone and the day 7 Eagle (corresponding to the year 992). Below them are their children: the first son 12 Motion Bloody Tiger, born in 7 House (993); the second son 3 Water Heron; and a daughter 3 Lizard Jade Ornament. The remainder of the page gives further genealogies. Other pages in the codices also give historical events, such as conquests and sacrifices. |
Moving to South America, we'll look briefly at the recordkeeping of the Inca Empire. Quechua was the official language of the empire, but the Inca ruling class actually spoke a different language, Puquina, of which few records survive. |
The Incas used complex collections of knotted strings called quipu (Quechua khipu) as a form of recordkeeping. See also a very complex quipu, at the Penn museum.
|
Three different types of knots, plus meaningful blank spaces, were used, along with colors. The quipu use a decimal place notation: simple knots (s) for places above the units, and a long knot (L) with multiple turns for the units place. If the unit is 1, then a single figure-eight (E) is used.
|
The reason we can verify these practices is that additionally many quipu include top cords that total the other strings (here, 123 + 112 + 20 = 255). Here the main (primary) cord has three pendants, two of which have a subsidiary. The top-cord subsidiary totals the figures on the pendant subsidiaries (322 + 111 = 433). Zero is shown by the absence of knots in a position. The use of knots for numbers is well understood, but it has also been claimed that they represent non-numerical information as well, possibly placenames for example, or perhaps commodities that are being counted. There appears to be no dependable evidence, however, for fully expressed phrases or sentences. |
Marcia Ascher and Robert Ascher, 1981. Code of the Quipu: a study in media, mathematics, and culture. University of Michigan Press, pp. 29, 31.
We now shift to North America. There was no known indigenous writing before European contact, but there were various means of recording information. The Dakota Indians of the northern plains of the U.S. on occasion used a series of symbols, one for the passage of each year (or "winter"), to keep track of events beyond the memory of living individuals. This famous example of a winter count was maintained by Lone Dog of the Yankton band of Dakotas, living in Montana, covering 1800-1871. The symbols begin in the center of a buffalo hide and spiral outward. These symbols are examples of ideographic proto-writing, since they are not connected to particular words, but rather to events or ideas that are suggested by the image. |
This hide is now on display at the Smithsonian's National Museum of the American Indian.
Below are a few examples of symbols used in Lone Dog's Winter Count.
For more examples, see the detailed list of the symbols and the events to which they refer. |
Since Dakota personal names are often based on the words for animals or activities, they are relatively easy to represent or at least, to evoke pictorially. The following list of warriors is from 1883. |
Robinson (1995)
Here are six specific examples of names of warriors in the list.
Here are the pictorial representations of those names, so you can see how they work. Notice that the names are often evoked rather than given precisely, since a verb like "spare" requires a fully linguistic writing system, whereas this is an ideographic approach. Other names, such as "red horn bull", are easier to depict. |
Cherokee is an Iroquoian language originally spoken primarily in what is now Georgia and South Carolina. It is related to languages of New York and neighboring regions (such as Mohawk, Seneca, Tuscarora). Beginning around 1809, a Cherokee man named Sequoyah began to design a writing system for his people's language. It is reported that he could not speak or read English, but believed that a written language would confer important benefits. Whether or not he understood that English writing was alphabetic, Sequoyah designed a syllabary for his own language. In the fall and winter of 1838–39, most of the Cherokee nation was forced to march to new territory in Oklahoma; because of harsh weather, inadequate food, and an exhausting pace, 4000 people died during the journey, named the Trail of Tears. Adoption and dissemination of the syllabary occurred shortly before these events. |
In its original form of 1821, the syllabary consists of apparently arbitrary cursive symbols. The smaller letters here are the later version shown immediately below. Originally Sequoyah also proposed special symbols for numbers. These were not, however, officially adopted by the tribe; instead, Arabic numerals are used. |
In 1827 a new version was adopted, one that is based graphically on Latin (and Greek) letters, many with various modifications. This is the form that was officially adopted by the tribe, and survives today. The syllabary was rapidly adopted by members of the tribe, resulting in quite a high literacy rate. It is still used for writing Cherokee today. |
Though the shapes of many symbols are obviously based on Latin letters, they have completely different values.
Note that the letter <v> is used in transcriptions of Cherokee and some other southeastern languages for a nasalized vowel [ə̃], similar to that in the English word huh. Also, letters for voiceless stops (p, t, k) are used for aspirated stops, as typical in English, while the letters that technically represent voiced stops (b, d, g) are actually voiceless unaspirated stops, as in the English words span, stan, scan. I'll follow that practice here for the sake of consistency. |
The perennial problem of consonant clusters is handled somewhat irregularly. There is one symbol that stands for a consonant alone, specifically [s], as in this word for "five". This use of a special consonant-only sign in the context of a syllabary is similar to the Japanese kana symbol for the moraic nasal <N>.
|
For other clusters, various words have a silent written vowel. This strategy is more like Mayan writing, although there's no clear pattern to the choice of character (e.g. not synharmony, nor always the same silent vowel).
|
However, a syllable-final glottal stop [ʔ] or [h] is unwritten.
|
Like many writing systems, the Cherokee syllabary does not distinguish certain "prosodic" features such as vowel length and pitch. Thus these words are written the same, but have different pronunciations. The accents give information about tone on that syllable, where á is high, a is mid, and à is low.
Notice that the differences in meaning can be rather important: they are distinguished in the spoken language, but not in the syllabic writing. |
Examples from Durbin Feeling, et al. (2003) A Handbook of the Cherokee Verb: A Preliminary Study. Cherokee Heritage Center, Tahlequah, OK.
Cree is an Algonquian language of central and eastern Canada (called Montagnais in the east). It is related to many Native American languages of Atlantic region (including Unami Delaware or Lenape, spoken where Philadelphia now stands). Cree can be written in Latin letters, but there is also a syllabary. It was invented around 1840 by a white minister living in Manitoba, James Evans. It is used for writing Cree in central Canada together with some nearby related languages such as Ojibwa. By 1879 it was adapted for the unrelated Inuit language in the Arctic region of Canada, specifically Inuktitut. |
A striking feature of the Cree syllabary is that a single basic shape is used for each consonant series. By convention, the symbol <c> is used for "ch", i.e. [č]. The various vowels are distinguished by the orientation of the basic sign.
It makes sense to call this system a syllabary because each graph normally represents V or CV, as with other syllabaries such as Japanese kana and Linear B. The core system is also pseudo-alphabetic in the systematic relationship among signs that expresses vowels though the independent representation of the vowel is the orientation of the sign rather than a separate graphical mark. It somewhat resembles the Indic and Ethiopic alphasyllabaries in the way that changes to the basic sign are used to indicate vowels. Where did this unusual syllabary come from?
|
Cree makes a phonological distinction between the short vowels [i, o, a] and the long vowels [î, ô, â]. Unlike these, [e] is always long, so it can also be written [ê]. (The circumflex is the conventional way to write a long vowel in Cree romanization.) In Cree writing, a long vowel is indicated by placing dot above the syllabic character containing that vowel. This use of a diacritic is exactly parallel to the use of a circumflex (or other length mark) in the Latin transcription. |
A purely alphabetic aspect of the system is that [w] is always written as a separate character. This sound is special in the Cree language because it is the only one that can occur as the second element in a cluster of consonants at the beginning of a syllable. Thus Cree has not only [ka] but also [kwa], and so forth. Some special method is necessary to represent such clusters. When followed in speech by a vowel, a dot is written either before or after the syllabic character. The choice depends on the regional variety of the syllabary. The spoken version actually has the [w] in the middle of the syllable, which is why there's no obvious place to put the dot. Logically, this dot should always precede a vowel-only sign for a syllable like [wa], but for the sake of consistency, the dot follows it in those regions where that's the practice for [kwa]. At the end of a syllable, [w] is written with a small circle °. This follows the general practice of special symbols for syllable-final consonants (discussed below). Also, in some regions the character " is used alphabetically for [h]. In other regions, [h] is simply not written. |
More generally, any consonant at the end of a syllable (i.e. not followed by a vowel) is written as a separate character. In Western Cree, special characters unrelated to the general syllabic signs are used for syllable-final consonants; these reflect the original Cree practice. In Eastern Cree, a small raised version of the Ca character is used for this; these are an innovation, originating in the Inuktitut use of the syllabary.
See a Cree text with many of these special consonant-only symbols. See also the back cover from the journal Inuit Studies (with French/English and then Inuit) which illustrates the Inuit version of the syllabary. |
This term normally refers to an alphabetic system that avoids special symbols — i.e. is more practical to use every day. In particular, linguistic symbols such as [ʃ] and [š] are avoided in favor of digraphs such as sh or other solutions. (This was especially true in the days of typewriters, but is not a technical problem in the age of Unicode.) But there can be complex choices and necessary compromises. Often linguists have different views of the best orthography compared to the community of speakers. As Leanne Hinton put it:
This is an issue that arises for linguists trying to help develop a workable orthography for a language without a written tradition. |
Hinton, Leanne. 2003. "Orthography Wars." Presented at the WAIL conference, UC Santa Barbara.
The following table illustrates some choices made by the Havasupai tribe of Arizona — traditionally living in and around the Grand Canyon — in developing an orthography for everyday use.
Some
choices are obvious from the viewpoint of English orthography — the main
influence for non-linguists – but note a couple facts. Because English
voiceless stops and affricates (p, t, k, ch) are aspirated in
many positions, those spellings are a natural match to aspirated stops
in Havasupai. Because the /r/ here is a tap, unlike the English approximant r,
it seemed more similar to English flapped d between vowels (as
in ready). That meant that a special symbol was needed for /t/,
which would otherwise have used d. The strike-through Given local familiarity with Spanish, the character ñ was a natural supplement to the English notations. But the very closely related Hualapai tribe decided to develop a different orthography, partly to distinguish the identities of the two tribes. Thus rather than ñ ŧ the Hualapai use ny đ; and whereas in Havasupai a predictable vowel inserted in consonant clusters is unwritten, such as mse: "to be afraid", in Hualapai it is written explicitly, mise:. The name ha-vasu-pai is literally "water - blue/green - people", for reasons evident from the photograph of Havasu Falls. |
The Zapotec languages, spoken mainly in Oaxaca in Southern Mexico, have very complex contrasts in their vowels, including tone and phonation. Linguists use careful notations to keep the different vowels distinct, which can number up to 33 different vowels and vowel combinations. A few are shown there, with the tone named at the right.
This discussion is based specifically on the dialect of San Lucas Quiaviní. |
While necessary to know the exact pronunciation, these notations are very cumbersome, and speakers are generally resistant to learning and using them. In connected text, someone who knows the language can figure out the words with less explicit information, so the same orthography need not be used all the time.
By a regular process of simplification — essentially reducing all the vowel contrasts to the simple vowel, as well as certain consonant simplifications — a much less cumbersome notation is created, but one that is still complete enough for native speakers. |
This difference can also be exploited to make the use of dictionaries easier. By using the simplified spellings as guide words, even someone who doesn't understand all the special notations can locate the relevant entry. It's also easier for a linguist who knows the diacritics, since one can find the right part of the page more quickly. |
Munro, Pamela. 2003. "Preserving
the Language of the Valley Zapotecs: The Orthography Question."
Language and Immigration in France and the United States: Sociolinguistic
Perspectives. University of Texas.