|Three types of proto-cuneiform "nine": N14 (circles), N01 (wedges), and N18 (slashed circles).|
As usual, I begin the discussion of the Indus symbols made with nine strokes by noting apparent numerals. There are three different signs that are made up of nine short strokes grouped together. The first of these displays the marks in a simple row: NINE QUOTES (IX 1). Koskenniemi and Parpola include this sign in their list as KP129(a), but Wells does not note it. Neither does Fairservis, who states that there are no occurrences of this form “nine” (1992: 62). If it does actually occur (and I have no reason to doubt Koskenniemi and Parpola), it is probably a singleton or at least very rare.
|Detail of seal H-23 with inscription: STACKED NINE (5 X 4) / QUAD-FORK / |
HAIRY HUNCHBACK / WINGED MAN / POT.
The second apparent numeral is STACKED NINE (IX 2), also known as KP129(b), W219, and Fs O-14. I designate this form in my database on inscriptions as STACKED NINE (5 X 4) in order to differentiate it from a second stacked form, one to be discussed shortly. Fairservis does not hesitate to identify this one as the number nine, adjectival form, noting two occurrences (1992: 62). This identification is rather surprising since he elsewhere suggests that the Harappans had a base eight number system – in which case, they should use a distinct symbol for “eight” and that plus a single quote for “nine.” That fact demonstrates to the satisfaction of most scholars that the Harappans did not have a base eight number system, or at least the evidence does not support the existence of such an unusual number system. In any case, this form includes five short strokes in an upper row, with four short strokes in another row beneath them.
In proto-cuneiform, “nine” occurs in a variety of forms, most of which are not quite like this Indus sign. Only the sign designated 9(N58) takes this precise shape. Another form of “nine” includes the same configuration of strokes, but is rotated 90 degrees (N57). Three types of proto-cuneiform “nine” are written as a column of four wedges alongside a second identical column, with a ninth wedge that is larger than those in the columns at the base (N01, the same but with larger wedges as N34, and the larger wedges with two short strokes added to the right and left sides of each wedge as N36). The very similar pattern of five marks in a column alongside a second column of four marks occurs in six other forms of “nine” (simple wedges as N08, circular indentations as N14, larger circles as N45, small circles with a diagonal stroke added to each as N18, circles with two strokes on each side as N19, and circles with three diagonal strokes crossing each as well as a fourth stroke to the right as N20).
There are also two different forms of the numeral “one” that contain nine short strokes, in proto-cuneiform. In the first of these, a long wedge bears nine horizontal strokes across it (N12). In the second, a large circular impression contains a “stacked nine” of five strokes over four (N63). There are many different types of numeral in this system because there were distinct enumerative systems for different types of item. In much the same way, in modern English, we have special terms for measures of liquids that differ from the terms for weight, which differ yet again from terms for length or distance. There are eight ounces in a cup, two cups in a pint, two pints in a quart, and four quarts in a gallon (in the U.S., that is). But when speaking of weight, we say that 16 ounces make a pound, even though “ounce” is the same word in both systems. In measuring length, we say there are 12 inches in a foot, three feet in a yard, and so on. If we wrote these varying measures in a proto-cuneiform fashion, we might use N01 numerals for liquid measures, N02 symbols for weight measures, and N08 signs for length measures.
Moving on, the third Indus sign comprising nine strokes is STACKED NINE (3 X 3 X 3), or IX 3. It occurs in the literature as KP141(b), W227, and Fs N-2. Fairservis does not consider this a numeral, but describes it as a depiction of a river or stream, meaning “water.” Since the individual strokes are not perfectly vertical, but “slashes” (and backslashes), it is possible that he is right about this not being simply a second version of “nine.” Wells, for his part, notes this sign as a singleton from Mohenjo daro (identified as Marshall No. 273). However, I find one of these at Harappa (H-322) and one from Lothal (L-47).
|Proto-cuneiform LUM, "manure" (upper left) and "nine" (N19, upper right);|
non-numerical Luwian nu (lower left) and ki (lower right).
Among the Luwian hieroglyphs, there is one that is made up of short, vertical strokes in three rows of three. It is not a numeral, though it looks like it ought to be. Instead, it functions as a phonetic symbol, representing the syllable nú. There may still be an indirect relationship with a number, though, because “nine” in Luwian derives from Proto-Indo-European *newn (the second "n" needs a small circle beneath it to denote a syllabic nasal) (Watkins 2000: 58). The Luwian "nine" itself is unknown. If any of the Indus signs has a phonetic value – a possibility that has yet to be proven – some or all of the apparent numerals might serve a parallel phonetic function.
Then again, this could be just another way of writing “nine” since proto-cuneiform provides such a parallel. One variant of the N19 form of “nine” is made up of nine circles, each with two strokes attached to either side, in three rows of three. Still, there is another analogous proto-cuneiform sign, one made up of stacks of diagonal marks. In LUM, seven strokes angle one direction alongside another stack of seven strokes, these angling the other direction, and a final set of seven angles like the first. Thus, the whole symbol resembles a stack of “Z’s” (rotated 90 degrees). Having no numeric significance, this sign came to mean “manure” and “cloud” (Halloran 2006: 164).
Sometimes elements that resemble numerals are combined with something else, in various writing systems. In proto-cuneiform, for example, a “stacked eight” of four strokes over four appears beneath a chevron, the whole thing rotated 90 degrees (GI6). This came to mean “night, shade.” Thus, while it appears numerical, the presence of the chevron changes its function to a semantic one. Similarly, in Luwian, a “stacked four” is bracketed by two lines resembling elongated “Z” shapes to form ki, perhaps the first syllable in a word meaning “fourth, quarter” or something similar (< Proto-Indo-European *kwetwer- / *kwetwor; Watkins 2000: 45).
|Tablet H-205 with inscription (from right to left): DIAMOND WITH BACKSLASH / CUP /|
STACKED SEVEN BETWEEN PARENTHESES / POT.
Indus signs provide a parallel, with STACKED SEVEN BETWEEN PARENTHESES (IX 4), also known as KP136(b) and W204. Wells notes a total of 16 occurrences, with two from Mohenjo daro, two from Harappa, and 12 from Lothal. Of the last group, items L-161 through L-170 are duplicates. There is also SEVEN QUOTES BETWEEN PARENTHESES (IX5). It appears in the literature as KP136(a) and W214, a singleton from Harappa (H-156). A third example may be a variant of the latter, SEVEN POSTS BETWEEN PARENTHESES (IX 6). It appears only in Wells’ list, as W215, where it is a singleton from Chandigarh (Ch-2). The distinction between “quotes” (or short strokes) and “posts” (or long strokes) is not always entirely clear, especially where there are no other signs for comparison. But in these two cases, the distinction is clear enough.
|Inscription from H-156: SEVEN QUOTES BETWEEN PARENTHESES / POT.|
|Inscription Ch-1 (right to left): DUBYA / SEVEN POSTS BETWEEN PARENTHESES / POT.|
Not all the nine-stroke symbols appear to be numerals, though. There is another STRIPED MALLET (IX 7), this one with three internal stripes. It is also identified as KP280, variant “B” of W470, and Fs N-10. Fairservis suggests that it represents a sign for a place, perhaps in the sense of a determinative like Egyptian O49 (a “fat ex” in a circle). Wells finds a total of 23 occurrences of striped “mallets,” but does not indicate which ones contain what number of stripes. They appear at Mohenjo daro, Harappa, Lothal, and Kalibangan. However, I see this or a similar sign 16 times at Harappa and 20 times at Mohenjo daro, in addition to the single instance from Lothal and that from Kalibangan. If my count is correct, then there are 38 in all.
|Detail from seal H-1 with inscription: FAT EX / BATTERY / STRIPED MALLET (3 stripes) / CARTWHEEL.|
The next sign is CAGED MALLET (IX 8), also KP277, W471, and in Fairservis a combaintion of G-9 + P-9. Fairservis defines the “mallet,” which he sees as an enclosure with a pillar, as “number nine; cow(s).” The four dots (or short vertical strokes) surrounding the central element are drops of milk, he thinks. Putting these two together makes “tribute (of milk cattle); to flow (as with water).” Since there are several better possibilities for a “nine,” I think it most unlikely that the “mallet” element has that meaning. In any case, Wells cites three examples of the CAGED MALLET, one from Mohenjo daro and two from Harappa. I see three from Mohenjo daro and three from Harappa (two of the latter being duplicates), giving a slightly larger total of six occurrences.
|Detail from seal H-67 with inscription: SINGLE POST / CAGED MALLET.|
We come to a very peculiar sign next, STRIPED MALLET BETWEEN BI-QUOTES (IX 9). It appears as a separate symbol only in Wells’ list (W474), one which he thinks is a singleton. However, I see quite a few duplicates of the copper tablets bearing this sign: M-519 through M521, M-575 through M-577, and M-592. If each object bearing the sign is counted, there are seven occurrences. Some scholars count elements in the duplicates only once apiece. Following this procedure, there are three distinct occurrences. In each case, the “striped mallet” portion is adorned with a short “quote” mark on each side of the “handle,” near the top.
While there are no parallels for the four-mark “caging” that appears so often in the Indus script, I do find an analog for this peculiar two-mark “cage.” It occurs in proto-cuneiform, where TAR~a – which alone resembles the Indus BI-QUOTES – appears in combination with other elements to form additional signs. Alone, this symbol came to mean “to cut, separate.”
|Tablet M-519A with inscription (right to left): POTTED TRIPLE SLASHES (?) / CUPPED TRIPLE SLASHES (?) / MAN HOLDING QUOTE / STRIPED MALLET BETWEEN BI-QUOTES / SINGLE POST / MAN HOLDING POST / DOUBLE GRIDS (probably the same inscription as on M-592).|
The last sign for this post is EXIT UNDER TABLE (IX 10). It is a singleton, elsewhere known as KP286 and W478. I reminds me of an Old Chinese character, ting2, “the phonetic ding1 [which resembles our letter “T”]...replaced the [mouth, later a square] at the bottom [of a previously discussed character, gao1]....Pavilion, terrace” (Wieger 1965: 191).Proto-cuneiform, which so often provides parallels, does not have a similar sign. There are elements resembling thicker versions of the “table” (ZATU 750), other symbols comparable to the “exit” (ZATU 737 and |ZATU 737 x SZE~a1|). And there are signs made up of one symbol beside or beneath another (e.g., |SILA3~a x HI|). But these individual pieces do not combine in a manner resembling the Indus EXIT UNDER TABLE. This will become a familiar pattern as we proceed through the remainder of the Indus signs, those containing nine or more strokes. The more complex signs have analogs much less often than the simple ones, with a few notable exceptions. But I will have more to say about this later.
|Detail from seal M-742 with inscription: EXIT UNDER TABLE / PINCH / TWO POSTS / DIAMOND / MAN.|
|Old Chinese ting2, "pavilion."|
|ZATU 737, three variants.|