Saturday, October 1, 2011

Analysis of Indus Inscriptions: Part II

Seal M-626 with inscription: FAT EX / PINCH // PRAWN / ZEE / CROSSROADS EX // POT //
(probably three units of information).

In the previous post, I mentioned the longest continuous Indus inscription, which appears in three lines on a seal (M-314).  According to Korvink’s analysis, there are also some inscriptions with multiple units of information in a single line.  One of these reads: FAT EX / PINCH // PRAWN / ZEE / CROSSROADS EX // POT // FAT EX / PINCH // CIRCLED TRI-FORK // POT (M-626) (note that there is another line in the inscription, but this single line is the focus here).  In this inscription, the “X” shaped sign at the beginning (FAT EX) couples with the long and short lines (PINCH) to make up a prefix (P).  The appearance of the first POT, a terminal sign (T), signals the end of the first unit of information.  This is followed by the same prefix as before (FAT EX / PINCH).  The second unit of information ends with the same terminal sign as the first (POT).  The only part that differs in these two units of the inscription is the middle (M).  The first medial segment contains three signs (PRAWN / ZEE / CROSSROADS EX), the second having only one (CIRCLED TRI-FORK).  Thus we may represent the components of this inscription as follows: PMT+PMT (+M).

According to the hypothesis that seal inscriptions contain names and titles, one would assume that this relatively long sequence must contain two such name + title sequences.  Korvink comments: “The two units of information here, rendered similarly, must represent two analogous units of information (two titles?) and not a title with appended auxiliary information” (2007: 73).  In the eyes of this author, this situation is problematic, because one person would not have more than one title and would have no practical use for a seal with someone else’s name and title alongside his own. 
Elamite cylinder seal impression, with image and an inscription of three lines.

I see two possible explanations for such compound inscriptions, however.  In the first place, some ancient peoples allowed a single person to bear more than one title.  In the previous post, I noted examples from Mesopotamia and from Egypt where a single person certainly had more than one title.  Another such example comes from Turkey: AN-LUGAL-..., (name 1) city ruler of the god Assur (title 1), son of Ikunum (name 2), city ruler of the god Assur (title 2) (Collon 2005: 42-3, no. 136, Period IV, 2000-1500 BCE).  Here, a single title is repeated, as the seal owner and his father both served as city rulers.  There are also multiple examples following the following pattern: Ibbi-Sin (name 1), god of his country (title 1), king of the Four Quarters (title 2): Aham-arshi (name 2), the scribe (title 3), son of Babati (name 3), is your servant (title 4) (2005: 36-7, no. 118).  In this case, the first named person has two titles, while the second essentially has two as well (scribe and servant), with a patronymic that includes a name, in addition (son of Babati).  Archeologists characterize the Indus Valley Civilization as more egalitarian than either Egypt or Mesopotamia.  So the multiplication of kingly titles found in Egypt and Mesopotamia may not be equally characteristic of the Harappan leaders.  But it may still be the case that a single Harappan held more than one title, even if this was relatively uncommon.
Early cuneiform cylinder seal with image and inscription.

In the second place, I would compare such compound Indus inscriptions with inscriptions from the Near East that contain names and titles of two people.  The most common form for these includes a patronymic, an indication that the first person (with his title) is the son of the second person (with his title).  Another common form occurs when the first person named is the king, sometimes given more than one title, followed by the name of the seal owner, who is the “servant” of the king.  A third type indicates that the seal owner (with or without an accompanying title) is the servant or beloved of a deity (with or without accompanying titles).  Examples of the first two types are given in the previous paragraph.  A combination of the third type with elements from the others is as follows: Ilum-muttabbil, purification-priest of (the goddess) Inanna of Zabalam, son of Shu-ili (2005: 45, no. 157).  Thus, I consider that the hypothesis of a nominative function for Indus seals remains a possibility.  It is not disconfirmed either by long inscriptions that contain more than one unit of information, or by comparison of inscriptions of varying lengths that serve the same function.  Of course, this does not prove that the hypothesis is correct.
Indus bas-relief tablet H-187 with inscription (right to left): STRIPED TRIANGLE /
TRIPLE TRIANGLES // POT / COMB (no prefix, two-sign terminal).

Korvink notes that if one accepts the thesis that the inscriptions contain names and titles, one must decide where the name ends and the title begins.  He points out four inscriptions containing the sign sequence STRIPED TRIANGLE / TRIPLE TRIANGLES, demonstrating the difficulty of this determination.  In the first two examples, these two signs comprise the whole of the inscription, there being neither a prefix nor a terminal.  (I do not find the precise inscriptions he cites in the first two volumes of the Corpus and his citation by numbers given in Mahadevan’s concordance, which makes it difficult to match them with the artifact numbers found in the Corpus).  However, M-1307 and M-1353 both contain these two signs plus the most frequently occurring terminal, POT.  Another inscription is identical except for the addition of a second terminal sign, COMB (H-187).






In the third example, the same two triangular signs are preceded by a common prefix (CIRCLED VEE / BI-QUOTES), as well as followed by the same terminal (POT).  In the fourth and last example, the same two medial signs are preceded by the same prefix as in the previous example (CIRCLED VEE / BI-QUOTES before STRIPED TRIANGLE / TRIPLE TRIANGLES).  Two additional medial signs follow: BOAT / SKEWERED CHEVRON.  At the end, there is a terminal pair (TRIDENT TOPPED POT / POT).
(containing a prefix, the same medial segment as in H-187, and a single-sign terminal).

In these examples, the last two are sufficiently long, in Korvink’s estimation, to contain a title (as they contain the three elements, PMT), but the first two are not (containing either M alone or MT, depending on whether one follows his examples or mine).  Assuming that the seal’s function would be the same regardless of the inscription’s length, Korvink concludes that the nominative function is ruled out. 
BOAT / SKEWERED CHEVRON // TRI-FORK TOPPED POT / POT (containing a prefix, a medial segment
with the same sequence as in H-187 but two additional signs, and a two-sign terminal differing from H-187).

However, inscriptions from Near Eastern cylinder seals provide examples of similar variations – a name alone, a name plus a second name joined by either a patronymic or statement of servanthood, a name plus occupation, a name plus association with a deity, or a combination of a number of these possibilities.  Thus, I consider the nominative function to remain viable for Indus seals.
Funeral monument of King Wadj, showing the royal serekh (a building) with the snake
indicating his name, and the royal falcon on top (after Malek 2003: 36).

In early Egyptian, in a similar way, the king’s name may be the only true writing on a tag or seal.  But other symbols occur with the name as indicators of kingship, e.g., the falcon of Horus or the Set animal (or both), the serekh, occasionally a “rosette.”  The falcon atop a serekh bearing the snake glyph of his name appears on a stela of King Wadji from his tomb at Abydos (Malek 2003: 36).  The royal rosette occurs on the palette of Narmer and on the mace head of King Scorpion, both objects from the temple at Hierakonpolis (2003: 29-31).  The presence of two or more such indicators of status for a single king’s name is not unusual.  For example, the pharaoh’s name is also enclosed in a serekh at the top of each side of Narmer’s palette, as well as the rosette behind him.  And the falcon perches on the serekh of King Wadji, both indicators of royalty.  In these cases, the falcon, the serekh, and the rosette are symbols, each of which represents royal status.  Thus, these symbols have meaning, but they do not function as signs in a script on these early objects.
Detail from the mace head of King Scorpion, showing the king, the (damaged) scorpion
before him indicating his name, and the royal rosette (after Malek 2003: 29).

If the Indus script functioned in a manner similar to the earliest Egyptian, we might expect such elements as the STRIPED TRIANGLE and TRIPLE TRIANGLES (quoted in Korvink’s example) to represent the personal name of a high-status individual.  The presence of this pair as the whole inscription would be perfectly adequate on a tag or label, as it would indicate that the object or container to which the tag was attached belonged to this individual or was intended for this individual.  Adding a terminal to a two-sign name might not serve a linguistic function, such as the honorific ending that Fairservis proposed (1992: 173).  Instead, it could be a way of providing more information, say, a place name or, as Mahadevan suggests, a representation of social status, similar to the Egyptian rosette. 
Mixtec woman named 6-Monkey (her birthday, indicated by the monkey head on her back and the six circles resembling beads on a string) Serpent Quechquemitl (glyph attached to her head by a line) Warband Quechquemitl (indicated by her
garment), a person with three names (after Smith 1973: 221).

In the Mixtec codices of Mesoamerica, again there are symbols that represent names, often of humans but sometimes of deities (Smith 1973).  A person is depicted in pictorial form, with his or her name written most often as a date (the birthdate).  Some have a single name, the birthdate, while others have two, the birthday plus an epithet or title, and occasionally someone has three: 3-Monkey (a man, 1 name, p. 225); 9-Movement (a male deity, p. 296); 7-Grass “talking Tlaloc” (a man, 2 names), and 6-Monkey “Serpent Quechquemitl” “Warband Quechquemitl” (woman, 3 names) (1973: 221).  Here, the significance of a person or god having one name, two, or even three, is not well understood.  Nevertheless, it is clear that a single name serves the same nominative or identifying function as a name plus title(s), or a series of names.
Mixtec glyph representing the place "Hill of the Road"
(after Smith 1973: 224).

In the Mixtec codices, there are also glyphs representing specific places, some made from a semi-standardized symbol for a hill, with additional elements specifying which hill is meant (“Hill of the Road” and yucu ñu ycu p. 224).  Other place names contain a rectangle decorated with a stepped design, along with specifying elements (ytno diyuchi, with a triple motif resembling a lit candle, ytno nino maa with a person standing on the rectangular element, p. 231).  Still other place names are represented by a building atop either the rectangle or the hill motif (Tilantongo from the map of Teozacoalco on the rectangle, the same place sign on the hill from Codex Nuttall, p. 234).  Actual dates of historical events are also represented, with the so-called “A-O year sign,” alongside a “week” glyph, plus circles to indicate the day numeral (e.g., beneath two place signs in the Lienzo of Zacatepec, p. 270).  Water is another symbol, with rivers depicted in a semi-standardized form as well (p. 272).  Thus, toponyms (place names) occur as well as names of persons and deities, even though much of the Mixtec language is never encoded in this proto-writing system.  This suggests a second possibility for the nominative function of Indus inscriptions.  They may represent places rather than (or in addition to) names and titles of people.
Mixtec glyph for the place Tilantongo (after Smith 1973: 234).

In Egypt, some of the earliest tags may include hieroglyphs representing the name of a town or nome (an administrative section of the country something like a province or county).  This may indicate the area from which the tagged commodities came (O’Connor 2009: 144-145).  According to this interpretation, the Egyptian proto-king who was buried in Tomb U-j at Abydos received funeral equipment from 45 different towns.  We might hypothesize that the simplest Indus inscriptions – those containing only a “medial” segment – symbolize locations, whether names of villages and towns, larger areas such as regions, or smaller areas such as quarters of a city.  Then the addition of a prefix might add information on the affiliation of a town, similar to the Egyptian specification of a particular town as being in a certain nome.  The terminal might indicate the larger region, perhaps one of the cultural regions that Possehl describes: Kulli domain, Sindhi domain, Sorath domain, Northwestern borderlands, Harappa domain, Cholistan domain, Anarta Chalcolithic, and Eastern domain (2002: 7).
King Tut's names, on left as carved and painted on a cartouche-shaped chest,
on right as embossed on book cover (Hawass n.d.).

The cartouche containing the name of a pharaoh might contain some of his titles, some of which convey geographical information of a general sort.  For example, a cartouche-shaped chest found in Tutankhamun’s tomb contains his name, followed by three symbols indicating that he was lord (the shepherd’s crook) of Upper Egypt (a pillar resembling a butter churn) and of Lower Egypt (the sedge plant).  There are, thus, ten symbols inside this cartouche (Hawass 2009: 4).  Three spell out the divine Ram’s name, Amun (with the zigzag water, n, being superfluous since its sound is also comprised in the gameboard above it, mn).  Three spell out the first syllable of the king’s name, Tut (bread loaf, chick, bread loaf, each indicating a single consonant).  One glyph by itself represents the middle syllable of his name (the ankh symbol, “life”).  The three signs at the end are an abbreviated title in which each symbol represents at least one word, with grammatical elements unrepresented. 
Canopic jar of Tjuya with four lines of hieroglyphs (after Davis 2000: Pl. XVII).

A canopic jar belonging to Tjuya also bears a brief text, a magical inscription, but this one serving a protective function (Hawass 2009: 135): “Words spoken by Selket: My two arms will embrace what is inside, extending protection around Qebehsenuef who is therein, the one revered before Qebehsenuef, the Osiris, king’s ornament, lady of the house, Tjuya.”  The first phrase is a standard opening of many inscriptions (ddi n ___, “Words spoken by ...”).  The spell indicates that the scorpion goddess, Selket, will protect the viscera in the jar, which are also guarded by one of the sons of the falcon god, Horus.  The name of the son, Qebehsenuef, reveals that it was intestines that were stored in this particular jar.  The other phrases refer to the lady whose remains were partly contained in this and three other canopic jars, namely Tjuya.  She was “the one revered before Qebehsenuef,” she was dead and therefore “the Osiris” in this case, and in life she had been “the king’s ornament” as the king’s mother-in-law.  She was the mother of Tiye, the Great Royal Wife of pharaoh Amenhotep III.  She was also “lady of the house” as a noblewoman, possibly serving as regent for the youthful king, alongside her husband, Yuya.  Again, though a single person, Tjuya had a number of titles, some of them specific to her role in life, some of them more general and shared with other people in death (e.g., “the Osiris”).
Detail from ivory armband depicting the oba of Benin gripping two mudfishes -- the ivory,
the costume worn by the oba, and the mudfishes are all royal prerogatives (after Cable 1984: 126).

Turning to sub-Saharan Africa, we can find instances of symbols used as identifiers in the absence of writing.  The kingdom of Benin provides several examples, where the majority of these standardized symbols are royal in nature (Cable 1984: 110-137).  Items carved from elephant ivory and cast in gold take the forms of royal emblems such as the leopard, elephant, and mudfish.  The oba or king also had the sole right to wear jewelry of coral.  He had many bronze or brass plaques that also included such symbols alongside more naturalistic representations.  “At important public ceremonies his [the oba’s] ritual costume would include carved ivory masks, worn in groups of three on the belt of his traditional kilt.  Only he had the right to wear a mask shaped like a leopard....The original white color of the ivory, which provided a startling contrast with the famous red coral beads that adorned the oba’s sacred garb, symbolized purity and peace” (1984: 122). 
Leather fan with ornamentation, showing the royal elephant at the top,
a symbol of the oba of Benin (after Cable 1984: 134).

In this case, neither the zoomorphic symbols of the oba nor the materials specific to his status (coral, ivory, bronze, and gold) represent words.  Nevertheless, they are still meaningful symbols and, in a way, serve a function akin to naming (i.e., the nominative function).  The individual oba is not distinguished from others, his predecessors and those who will follow him, by these kinds of symbols.  But then, the oba was considered a god, immortal, though one who changed his face periodically.  It was therefore not important to distinguish the individual symbolically as it was for the Egyptians and Sumerians.  It is possible that the Indus symbols had a similar type of non-linguistic meaning as well, representing statuses or institutions, in effect, rather than individuals. 
Detail of inscribed brick from the ziggurat of Choga Zanbil (after Potts 1999: 226, Pl. 7.7).

On a final note for this post, I will briefly address the possible narrative function of the script.  As Korvink observes, if the Indus script was intended to serve a narrative function, we would expect to find considerably longer texts (2007: 74).  Some researchers insist that these did indeed exist, but were written on perishable materials and are lost to us.  However, in Mesopotamia and Egypt such longer texts were not confined to perishable media.  Elamite kings often had bricks inscribed with narrative messages (Potts 1999: 226 Pl. 7.7 and 232 Fig. 7.9).  The first example depicted contains 6 lines of text that together cover the whole of one side of the brick.  The second contains 10 lines, though the brick itself is broken.  In each case, there are considerably more than 30 signs, a total never matched in any Indus inscription.  Statues and figurines also bear inscriptions in the Near East, sometimes on the base but at other times on the back, the arm, or the torso of the figure.  In the Indus Valley, both bricks and statues have been excavated, but none bear inscriptions.  Therefore, as Korvink states, most scholars agree that the Indus script did not serve a narrative function.


Cable, M. and the editors of Tree Communications, Inc. 1984. Treasures of the World: The African Kings. New York: Select Books.

Davis, T.M. 2000 (and 1907). The Tomb of Iouiya and Touiyou: The Finding of the Tomb. London: Duckworth.

Hawass, Z. n.d. Tutankhamun and the Golden Age of the Pharaohs. Washington, DC: National Geographic.

Korvink, M.P. 2007. The Indus Script: A Positional Statistical Approach. Gilund Press (Amazon).

Malek, J. 2003. Egypt: 4000 Years of Art. London: Phaidon.

Potts, D.T. 1999. The Archaeology of Elam: Formation and Transformation of an Ancient Iranian State. Cambridge: University Press.

Smith, M.E. 1973. Picture Writing from Ancient Southern Mexico: Mixtec Place Signs and Maps. Norman: University of Oklahoma.

No comments:

Post a Comment