An Indus Syntax?

Detail from seal M-38 with inscription: 2 POSTS / BLANKET / MAN HOLDING DEE-SLASH / TRI-FORK /
BI-QUOTES // FISH UNDER CHEVRON / WHISKERED FISH / 2 POSTS + FISH / PRAWN + ZEE +
CROSSROADS EX // POT (double slashes indicate division of the inscription into parts -- prefix, medial
segment, and terminal -- while "+" indicates the signs on either side form a close pair or triad).

This post is a discussion of an article published last year by S.Sinha, A.M. Izhar, R.K. Pan, and B.K. Wells (2010: 1-17).  The title of their work indicates they have performed network analysis of Indus inscriptions and found syntactic organization.  My first reaction to the title was that such “syntactic organization” had been found previously, on more than one occasion.  G.L. Possehl describes a number of analyses that made such claims (1996).  G.R. Hunter as early as 1929 noted the function of the POT and COMB as final elements – or terminal signs to use M. Korvink’s terminology (Possehl 1996: 85).  Similar conclusions come from the work of Father H. Heras (p. 110-112), a Soviet Russian team of scholars headed by Yu. V. Knorozov (p. 115-122), a Finnish team including A. Parpola (p. 124), and others.  M. Korvink’s analysis, which I have referred to repeatedly, came after the publication of Possehl’s review, but also notes such organization (2008).

In their preliminary discussion, Sinha et al note that the ancestral scripts that later became Mesopotamian cuneiform, Egyptian hieroglyphs, and Mayan writing had limited connections to spoken language (2010: 2).  They state, “In almost all cases, a writing system became more or less capable of expressing spoken language only after centuries of development....”  This is certainly correct for cuneiform and hieroglyphs (my knowledge of Mayan writing is too limited to assess the accuracy of the statement).  However, the earliest hieroglyphs may not be true writing, but rather proto-writing (O’Connor 2009: 144-147).  Proto-cuneiform, like proto-Elamite, is also considered not closely tied to speech (Englund 2001: 1-43).  The first tablets contain only numerical symbols, at the second stage containing ideographs as well, representing “owners” (people and/or institutions) and commodities.  When speaking of a symbol system that eventually evolved into a fully developed writing system, it is common to extend the term “writing” to the earliest examples of the script.  But that does not really demonstrate that the earliest inscriptions are indeed true writing.  Thus, the fact that the Indus script shares characteristics of the earliest proto-cuneiform, proto-Elamite, and hieroglyphic scripts may well suggest that the Indus script itself is proto-writing rather than a fully developed writing system (as noted in several previous posts).

A long bead of lapis lazuli from Mari in Syria, dating to the Early Dynastic IIIB period,
2400-2250 BCE, with a short inscription in early cuneiform: "For [...], Mesanepada, king of
Ur, the son of Meskalamdug, king of Kish, has consecrated [this bead]" (after Aruz 2003: 143). 

That caveat aside, the analysis of Sinha et al is quite interesting.  They begin by defining the corpus they use which is based on Wells’ database (2010).  This includes the inscriptions from all three volumes of the corpus, as well as some unpublished data (Sinha et al 2010: 3).  These inscriptions come from 3,896 artifacts, from which a subset of 2,393 are selected as being complete, unbroken, and legible.  From this a further subset is taken for analysis, comprising inscriptions that occur on a single line.  The selected sequences contain from 1 to 13 signs, with a median length of 4 signs.  As a comparison set, the authors created randomized variations on each inscription, in which the signs were combined in every possible order, with statistical measures averaged.  If the signs of the Indus script are arranged purely randomly, the statistics from the comparison set would match the statistics from the actual corpus.  The fact that results generally differ between the two sets is the evidence the authors cite for syntactic organization.

To begin with, the authors note 21 signs that only occur individually, as 1-sign inscriptions (2010: 4).  They then identify 128 signs that only occur at the beginning of an inscription, a group they term “beginners.”  The third group of 43 signs, which appears only at the end of inscriptions, is referred to as “enders.”  This leaves 401 signs, some of which sometimes occur at the beginning, some of which sometimes occur at the end, but all of which also occur in other positions.  Out of this large group, the authors isolate 127 signs that are never initial or final.  By comparison, the randomized set shows very different beginners and enders, far fewer beginners but considerably more enders.  “We observe that four beginners in the empirical set...and two enders...never occur as beginners and enders (respectively) in the 100 randomized trials we carried out.  It implies that the occurrence of these signs always at the beginning or end of a sequence (and never in any other position) may be highly significant, and certainly not a result of simple chance” (2010: 5).

One troubling feature of the three lists and of the conclusion drawn from this analysis is the large number of singletons.  As noted in previous posts, one of the striking characteristics shared by proto-cuneiform, proto-Elamite, and the Indus script is the very large proportion of signs that occur only once in the corpus.  Of approximately 1,900 proto-Elamite signs (not including numerals), over 1,000 are singletons (Dahl 2002: 1-2).  Around 300 additional signs occur only twice, while the vast majority – 1,700 in all – appear fewer than 10 times.  This leaves a relatively small core of 200 “common” signs, out of which just 20 occur 100 or more times.  What effect does this characteristic have on the three lists provided by Sinha et al?  As it happens, of 21 signs that occur solo, 10 are singletons in the entire corpus and only one appears 9 times.  Thus, all are rare.  In addition, of the 128 “beginners,” 66 are singletons.  A similar proportion are singletons among the “enders” – 26 out of 43 signs.  It seems misleading to say that a singleton is found only at the beginning or end of an inscription.  If it is a singleton, it only occurs once, which is not enough to establish a rule! 

An administrative text in proto-cuneiform from Jemdet Nasr, dated to 3000-2900 BCE
(from Aruz 2003: 41, copy by author).  The inscription includes around 20 symbols that
are separated into compartments.  It concerns distribution of grain as represented by
the signs resembling ears of barley or wheat.  No Indus inscription has four occurrences
of a single sign in such non-contiguous arrangement, i.e., not side by side.

After comparing the number of sequences begun or ended by these signs, the authors observe that “most of the beginner signs appear in only 1-3 distinct sequences” with a maximum of 8 sequences, while “a particular ender sign can appear at the end of a maximum of 4 distinct inscriptions” (2010: 5).  With such small numbers, the results may not be random but they do not seem particularly robust either.

Sinha et al also look at the distribution of the number of links each sign has (its degree) and the strength of these links (2010: 5-6).  The authors describe their results concerning the degree as indicating that “the variation of sign relations is much more restricted in the empirical sequences than would be the case had each of the sequences been put together randomly” (2010: 6).  Their graph indeed shows a difference of degree between the real or empirical set and the randomized set, but that difference does not appear particularly great.  When it comes to strength, the authors must agree that the results are almost the same between the two data sets, inasmuch as “this is governed by the frequency of occurrence of individual signs” (2010: 6).  Clearly, a common sign will have links with more signs than will a rare sign, whether or not their arrangements are random.

Thus, the 10 most frequent signs in the empirical set have the most links.  These signs include POT, BI-QUOTES, FISH, 2 POSTS, COMB, WHISKERED FISH, SPEAR, TRI-FORK, 3 POSTS, and VEE IN DIAMOND.  Arrows in one of the figures show in which direction the links occur among these signs.  For example, VEE IN DIAMOND precedes but does not follow POT; POT both precedes and follows WHISKERED FISH and COMB; FISH precedes but does not follow SPEAR, while SPEAR precedes but does not follow 2 POSTS.  The profusion of arrows is hard to interpret, but the description summarizes the situation: “about half of all potential connections [are] present” (2010: 6).

When Sinha et al discuss medial signs, they seem to be on firmer ground.  The intersection of two sets of signs – those with the most connections with preceding signs and those with the most connections with following signs – yields an inner core of 26 signs most likely to appear in medial position (2010: 7).  Of these, only 5 occur less than 100 times apiece, so the statistical measures are more convincingly robust.  Further statistical manipulations of the data yield a group of 35 signs forming the “core lexicon” (2010: 7-8).  They include medial signs: SINGLE QUOTE, BI-QUOTES, 3 QUOTES, 3 POSTS, WINGED MAN, HAIRY HUNCHBACK, FISH, MAKRED FISH, FISH UNDER CHEVRON, WHISKERED FISH, TRI-FORK, RAKE, MALLET, CUPPED POST, POT, CRAB, CIRCLED FORK.  They also include signs that are most likely to precede this medial group: SINGLE POST, PINCH, CROSSROADS EX, POTTED ONE, TRI-FORK TOPPED POT, CIRCLED E-FORK, CIRCLED VEE, CARTWHEEL, OVERLAPPING CIRCLES, VEE IN DIAMOND, BOAT (and three others occurring slightly fewer than 100 times each).

An Egyptian amuletic circle of gold and silver dating to 2000-1800 BCE (after
Pinch 2006: 111).  Most of the symbols are hieroglyphs, including wedjat eyes,
djed pillars, the ankh knot, and possibly the face (shown en face!).  But this is
not writing -- each symbol communicates its own message independently of position.

The authors then focus on determining which signs occur in pairs by comparing the probability of a given pair’s occurrence in the empirical data versus the probability in the randomized data set (2010: 9).  They find some groupings that Korvink and others identified previously: FOOTED STOOL + PINWHEEL; MALLET + FORK; CUPPED SPOON (or CUPPED POST) + 3 POSTS; 3 POSTS + SPEAR; STACKED 7 + EF TOPPED EXIT; 2 POSTS + FISH; PRAWN + ZEE; ZEE + CROSSROADS EX; CROSSROADS EX + POT; TRI-FORK TOPPED POT + POT; POT + MAN; POT + COMB; CIRCLED VEE + BI-QUOTES; VEE IN DIAMOND + BI-QUOTES; CARTWHEEL + BI-QUOTES; CARTWHEEL + PINCH; BOAT + PINCH; FAT EX + PINCH.  They also find some groupings that seem unintuitive by Korvink’s analysis, including BI-QUOTES + CIRCLED FORK; BI-QUOTES + WHISKERED FISH; BI-QUOTES + FISH UNDER CHEVRON; FAT CEE + POT. 

Here I have underlined the signs identified as terminals by Korvink and placed his prefix elements in italics.  Thus, the reader can see that where Korvink divides the inscriptions into three segments – prefix, medial portion, and terminal – Sinha et al simply note which combinations are most frequent.  The signs that occur most frequently at the beginning of the medial segment will thus seem to form pairs with the most frequent prefix constant (i.e., the final sign in the prefix).  The signs that occur most frequently at the end of the medial segment will similarly seem to form pairs with the most frequent terminals.  Now, how does one determine whether or not this type of – perhaps fortuitous – occurrence is significant?  Following the method of Sinha et al, one constructs a control group of fake sequences in which signs occur in random order.  Then one points to the differences between the set of random sequences and the real sequence, concluding that this difference indicates syntactical organization.  I suppose I am suggesting that there is organization all right, but not necessarily any syntax behind that organization. 

The main source of my discomfort with the study’s conclusions lies in the final section where the authors construct segmentation trees for the longest 33 inscriptions, each containing 10 or more signs.  The second tree they show is M-38 (see illustration above): 2 POSTS / BLANKET / MAN HOLDING DEE-SLASH / TRI-FORK / BI-QUOTES / FISH UNDER CHEVRON / WHISKERED FISH / 2 POSTS / FISH / PRAWN / ZEE / CROSSROADS EX / POT.  They find a strong sign pair to be 2 POSTS + FISH and another, equally strong, is PRAWN + ZEE.  So far, I would agree. 

Sinha et al find a third pair that I find problematic: FISH UNDER CHEVRON + WHISKERED FISH.  Further, they link this pair fairly strongly with the preceding BI-QUOTES, forming a triad.  In Korvink's analysis, BI-QUOTES belongs with the first four symbols, all five together forming the prefix.  Korvink also notes that the various "fish" signs tend to occur in a semi-standardized order (2008: 37), with FISH UNDER CHEVRON typically the first of any series, followed by FISH, WHISKERED FISH, DOT IN FISH, MARKED FISH in that order (though no inscription contains all of these).   He also notes the pair 2 POSTS + FISH, which may precede the otherwise standard sequence (2008: 38).  In M-38, though, this pair follows the other "fish." 

Sinha et al further show a strong link between the pair PRAWN + ZEE on one side and the following sign, CROSSROADS EX (2010: 11).  This seems clear enough.  But then they link this triad to the final symbol, the terminal POT.  This link with POT is stronger, in their view, than that between the symbols preceding BI-QUOTES, all elements of the prefix according to Korvink's analysis.  Thus, Sinha et al would divide the prefix between an initial "phrase" of four signs on one side, and the prefix constant along with the first two elements of the medial segment on the other side.  I find this segmentation tree -- as well as the other two shown -- to be unconvincing.

The authors’ description of their first example (with my sign names substituted for their numerical designations) reads thus:

The 3-sign cluster “LOOP ARMED MAN-DOUBLE BOATS-SPEAR” at the beginning of the sequence is the initial phrase, and is separated from the rest by sign SQUARE AY.  The medial sequence is broken into two parts “VEE IN DIAMOND-BI-QUOTES-FISH UNDER CHEVRON” and “DOT IN FISH-CUPPED POST-3 POSTS.”  The sequence ends with the terminal phrase “2 POSTS-FAT EX IN DIAMOND-POT.”  We observe that each of these four sub-sequences obtained by this analysis also  occur as units in other inscriptions in the WUCS dataset, thereby verifying the accuracy of the segmentation procedure (2010: 11, emphasis added).

Let us examine this type of verification, seeking the sign pairs in the corpus that Sinha et al designate.  I find 10 examples of 2 POSTS + BLANKET in the concordance of Koskenniemi and Parpola (1982: 101-102).  In two of these other than the cited sequence, BLANKET is followed by the MAN HOLDING DEE-SLASH (KP2027, M-119, M-1188).   There are other examples of this "man" followed immediately by one or the other FORK (14 instances in KP, 1982: 33).  Five of these place the FORK just before BI-QUOTES.  Interestingly, other instances have this apparent pair at or near the end of the inscription rather than in the prefix.  Since BI-QUOTES is one of the most frequent signs, it is not surprising that it often occurs immediately preceding FISH UNDER CHEVRON.  But if this prefix constant is joined with the pair, FISH UNDER CHEVRON + WHISKERED FISH, then the occurrences are fewer (6 examples in KP, 1982: 45-46).  There are multiple examples of 2 POSTS + FISH (66 examples in KP, 1982: 99-100).  Equally compelling is the link between PRAWN + ZEE (45 examples in KP, 1982: 53).  Of these 45 pairs, many but not all are followed by CROSSROADS EX (27 examples).  Two of these include another sign or two between CROSSROADS EX and the terminal (POT in each case).  In two inscriptions, the sign following CROSSROADS EX is either missing or illegible.  Thus, I find Korvink's description more convincing as it separates the terminal from the medial signs.  It seems to me that the "link" between CROSSROADS EX and POT is essentially fortuitous and due to the fact that POT is the most frequent sign by far.

In conclusion, I am not any more impressed with this article than I was by the work of Yadav et al (2008: 39-52).  Sinha et al, like the previous team, demonstrate that there are some patterns in the sequences of Indus signs.  But neither group seems to have developed a way to determine which of these patterns are likely to be syntactic and which are the result of other characteristics, such as frequency.  That is, if one finds that feature X correlates with feature Y, this may indicate a direct link between the two.  On the other hand, it may only be a clue that some other (unobserved) feature Z directly links with both and thus the apparent link between X and Y is really an indirect one.

A proto-cuneiform inscription on the so-called "Blau" monument shaped like a scraper (after
Aruz 2003: 39, drawing by author).  The "scraper" is from Jemdet Nasr in Mesopotamia and dates
to 3000-2900 BCE.  The text records an economic transaction with land exchanged for such goods
as wool, silver, goats, and (most importantly?) beer.  Note that there are over 30 signs, more than
on any known Indus artifact, though this is proto-writing. 

The conclusion of Sinha et al is tempered by their recognition that there is still no generally accepted sign list; where Mahadevan lists 417 signs, Wells enumerates over 700 (2010: 12).  They state that “an important open issue that needs to be settled is the robustness of these results with respect to the sign list being used” (loc. cit.).  Part of this open issue can easily be settled, though.  For example, among the “beginners” that appear only in initial position are some that would not be included if Wells had not classified them as independent signs rather than variants.  For example, Wells always distinguishes multiples of a sign from the same sign’s individual occurrences: DOUBLE SKEWERED CHEVRONS, QUADRUPLE TRI-FORK, DOUBLE CHEVRONS, DOUBLE EF TOPPED EXITS, DOUBLE CUPS, TRIPLE BOATS.  Each of these signs, when found individually, occurs in non-initial positions.  Wells also distinguishes reversed signs from their mirror images, something not all other list-makers do (unless the reversed sign appears in an inscription along with its mirror image).  If we remove both of these types of examples and remove all the rare signs (those that occur only once or twice in the corpus), then we are left with a much smaller inventory. 

In fact, if we require a sign to occur at least five times – which is still far too rare for meaningful statistics – the list of “beginners” would be significantly reduced.  If this reduced list were to be compared to sign lists of other researchers, even fewer would remain.  For example, Wells has ODD STACKED 7 appearing in initial position only.  But there are non-initial examples in the KP list where all variants of ODD STACKED are conflated (1982: 96).  MAN WITH WING might remain on the list of beginners, but KP show one instance where this sign is final (M-1054).  This fact is not noted by Sinha et al because they analyze only complete, unbroken inscriptions and M-1054 shows an initial sign to have been broken off.  CAGED MAN HOLDING POST is a singleton in KP’s concordance although Wells gives it a frequency of 5 (1982: 33).  FIGURE EIGHT WITH LADDER indeed appears only initially, but the similar BEE WITH SLASH, possibly a variant, is medial (1982: 56).  The BED is only a beginner if one separates the variants according to the number of legs; the one with 6 legs occurs initially as Sinha et al observe, but other variants appear both medially and finally (1982: 148).  SKEWERED CIRCLE is probably a variant of SKEWERED DONUT, which appears medially in 3 inscriptions (1982: 70). 

Inscription from a foundation figure of Lugalkisalsi, dating to the Early Dynastic IIIB period,
2400-2250 BCE (after Aruz 2003: 65, drawing by author).  It reads: "For Namma [a goddess],
wife of An [a god], Lugalkisalsi, king of Uruk and king of Ur, erected this temple of Namma."
Note that with 12 or 13 symbols, this is equivalent in length to some of the longest Indus
inscriptions, but with three non-contiguous repetitions of LUGAL, "king," it shows features
quite rare to Indus script.

SHISH KEBAB IN TRIANGLE is almost certainly a variant of BISECTED STRIPED TRIANGLE (1982: 22).  The latter sign occurs at least as often in medial position as initially.  Besides this fact, KEBAB IN TRIANGLE only occurs in a single inscription that is duplicated multiple times on tablets, as far as I can tell.  The duplicate inscriptions should be removed from the frequency count per Sinha et al's stated procedure (2010: 3).  STRIPED BOWTIE WITH EXTRA RIBBON occurs once in medial position (Kd-8).  REVERSED FOOTED STOOL is non-initial at least 6 times (Koskenniemi and Parpola 1982: 131).  BLANKET has a number of variants with differing numbers and arrangements of inner strokes or “ticks.”  Wells’ version with 8 “ticks” may well be a beginner, but other variants certainly occur medially and even once in final position (1982: 138-140).  Wells divides ASTERISK UNDER TABLE into two signs, one with an eight-legged star resembling the keyboard asterisk, the other showing the “X” form with extra strokes only between the upper and lower “legs,” not across the middle.  Presumably the KP list subsumes the two into one group with all in initial position (1982: 132).  The similar EX UNDER TABLE is always final except for one instance which is initial – perhaps this one instance was an error for ASTERISK UNDER TABLE!  The version of POTTED 3 with slashes cross the sign may be initial, but other variations are medial (1982: 178).  Two examples with slashes appear in what may be final position on a broken seal (H-689).  CRAB IN STRIPED LEAF TOPPED POT is another sign that is most likely a variant of several others, one of which occurs medially (L-11).  VEE & TRI-FORK IN DIAMOND appears in both medial and final position according to KP (1982: 201). 

It is rather tedious to examine each symbol in this way, but the point should be clear.  If researchers use the KP list, the same statistical measures would yield results that differ from those reported for Sinha et al using Wells’ list.  Lest the reader think my objection applies only to beginners, I will quickly note the enders as well.  If we remove signs occurring fewer than 5 times, this group would then be reduced to PANTS (one variant only since the KP list, which groups together all variants, shows 2 examples in initial position), EF PRONGED CHEVRON UPON STACKED 6 IN POT (which KP show as occurring in only 2 distinct inscriptions), CAGED SLASHES IN OVERLAPPING CIRCLES, SEPTUPLE STACKED ROOFS (which KP show as a singleton), and RAINY CARTWHEEL.

The foundation figure of Lugalkisalsi, bearing inscription (after Aruz 2003: 65).

At the end of their paper, Sinha et al address some of the problems of assuming Indus script is true writing.  Clearly they would prefer to define writing very generally, terming it “a system of intercommunication by means of conventional visible marks” (2010: 12, following Gelb 1963: 11).  Even so, they admit that “often writing tends to be narrowly defined as a means of efficiently encoding speech” (loc. cit.).  What they fail to note is that even Gelb went further with his definition and separated what he called the semasiographic stage from the phonographic stage of writing (1963: 12).  In today’s terminology (at least the terms I use), he separates proto-writing from true or fully developed writing.  In other words, early “writing” is often loosely connected to speech.  A given sequence of symbols may be expressed in a number of ways in speech.  At this point, those who define writing narrowly would not term the symbol system writing. 

There is good reason for defining writing narrowly in this way, since not all communication systems that use visible marks qualify as writing even to the most generous.  Gelb gives examples of such, including carvings and paintings on stone such as those in caves at Lascaux, petroglyphs in Africa and the Americas, mason’s marks from Anatolia, the Nsibidi system used in Nigeria (similar to the Adinkra symbols I referred to in some earlier posts), and so on.  We could also include Navaho sand paintings.  They are visible and they communicate – at least to those who know enough of the Navaho culture to interpret them.  But they are not writing and not intended to be writing.  In a similar way, we can point to symbols in our own culture that are not part of a writing system – the conventional hearts that appear on Valentines, the cross, star of David, and star and crescent that represent major religious faiths, the icons on computer screens, some simple depictions found on traffic signs, etc.  These communicate and they are symbols.  But they are not writing.

When scholars study proto-cuneiform, they may not always distinguish between the semasiographical stage, or proto-writing, on the one hand and the phonetic stage on the other.  After all, the symbols gradually evolved from proto-writing into true writing.  The same may be true of Egyptian hieroglyphs – the earliest symbols found on pre-dynastic tags mostly evolved into later glyphs with specific phonetic realizations.  But not all early writing systems continued to evolve in this way.  We know, for example, that proto-Elamite failed to evolve into true writing, dying out instead.  After a lapse of centuries, it was replaced by Linear Elamite, which is phonetic writing.  Similarly, the Indus script did not evolve continuously into a fully developed writing system used later on.  Instead, it died out and after a lapse of several centuries was replaced by clearly phonetic writing.

The authors note another problem, namely the fact that the inscriptions are so short (2010: 12).  They want to counter this by claiming that “many early writing systems exhibit such brevity,” citing early Sumerian and early Egyptian.  Early proto-cuneiform (if that is what they mean by early Sumerian) does include very brief texts.  Oddly enough, though, even in proto-cuneiform there are some that are longer than any Indus inscription (see tablets dating to Uruk III phase in Nissen, Damerow and Englund 1993: 22-23).  For example, a single tablet listing pigs bears 30 or so sections, each containing a single wedge (here functioning rather like the bullets in a modern bulleted list) and one or more non-numerical symbols (fig. 26).  And this is just on one side!  Even if we count symbols occurring on multiple sides of objects from the Indus Valley, there are no inscriptions with 30 signs.  So, while some brief texts are understandable in early proto-cuneiform economic documents, others are longer without necessarily containing phonetic writing.  The same is probably true of Egyptian hieroglyphs, as noted previously.

An ancient Chinese inscription from a bronze vessel (after Wieger 1965: 372).
Wieger interprets this to mean "In the presence of the Ancestor, after libation,
offering of jade wine and tow, by the son, a grandson having recently been born."
But for the first phrase, "in the presence of the Ancestor" there is just one symbol,
the character second from the bottom on the left which resembles the Indus MAN.
There may have been no simple, one-to-one correspondence between symbol and word
at this stage of (proto-?) writing.

In their effort to show Indus script to be nothing unusual, the authors go as far as to cite early Chinese, specifically symbols found on bronze vessels from Anyang that may predate even the Shang oracle bones (though the exact date depends on who is doing the dating).  “The majority of these inscriptions comprise only a clan sign and an ancestor dedication,” they say.  If clan signs and symbols (early Chinese characters) representing specific ancestors are the only types of signs, then this would also be better characterized as proto-writing.

An Egyptian tag from the pre-dynastic period (after O'Connor 2009: 145).
Archeologist G. Dreyer reads this two hieroglyphs, the elephant 3b and the
triple mountain peaks dw, thus spelling the name of Abydos.  O'Connor thinks
this represents the name of a town, though not necessarily Abydos.  Neither
seems to address the third element, a schematic tree.  (The circle upper left is a hole).

Perhaps early (proto) writing is always brief, the authors suggest, because “the main use of writing was as a mean[s] of maintaining accounts, list and other economic records” (loc. cit.).  But this is mainly true in the Near East.  In China, the oracle bones record divinations rather than economic transactions.  Among the Maya, early writing records astronomical and ritual information.  Mesoamerican proto-writing such as that used by the Mixtecs includes symbols for dates – which include numerals – and the names of persons.  There are also semi-conventionalized methods for indicating specific places and few other items of information.  The earliest Egyptianglyphs are also mainly indications of specific persons and places.  Thus, writing does not always serves first as an economic account and therefore we cannot assume that this is the reason for the brevity of Indus texts.

In one way, it does not matter what definition one uses.  One can still study a “script” even though it is not fully developed writing.  Several scholars now focus mainly on proto-cuneiform or proto-Elamite, for example, and they have been able to make considerable progress in interpreting the early, enigmatic accounts.  They do not try to excuse the difficulties of their documents by changing definitions to suit them.  Neither should scholars of the Indus script.  It is whatever it is because the people who once used it found it sufficient for their purposes.  In the end, then, is there syntactic organization in the Indus script?  I am not convinced we know, although the signs are not completely random.

REFERENCES

Aruz, J. 2003. Art of the First Cities: The Third Millennium BC from the Mediterranean to the Indus. New York: Metropolitan Museum of Art and Yale University Press.

Dahl, J.L. 2002. “Proto-Elamite Sign Frequencies” in Cuneiform Digital Library Bulletin 1: 1-3.  Available online at http://cdli.ucla.edu/pubs/cdlb/2002/001.html

Englund, R. 2001. “The State of Decipherment of Proto-Elamite” available online at http://cdli.ucla.edu , to be published in S. Houston, ed., First Writing. Cambridge: Cambridge University Press.
Gelb, I.J. 1963. A Study of Writing. Chicago: University of Chicago Press.

Korvink, M.P. 2008. The Indus Script: A Positional Statistical Approach. Gilund Press.

Koskenniemi, K. and A. Parpola. 1982. A Concordance to the Texts in the Indus Script. Helsinki: Department of Asian and African Studies, University of Helsinki.

Nissen, H.J., P. Damerow, R.K. Englund. 1993. Archaic Bookkeeping: Early Writing and Techniques of Economic Administration in the Ancient Near East. Paul Larsen, transl. Chicago and London: University of Chicago Press.

O’Connor, D. 2009. Abydos: Egypt’s First Pharaohs and the Cult of Osiris. London: Thames & Hudson.

Pinch, G. 2006 and 1994. Magic in Ancient Egypt. London: The British Museum.

Possehl, G.L. 1996. Indus Age: The Writing System. Philadelphia: University of Pennsylvania.

Sinha, S., A.M. Izhar, R.K. Pan, B.K. Wells. 2010. “Network analysis of a corpus of undeciphered Indus civilization inscriptions indicates syntactic organization” available online at arXiv:1005.4997v1

Wells, B.K. 2010. Epigraphic Approaches to Indus Writing. Cambridge, MA: Oxbow Press.

Wieger, Dr. L. 1965. Chinese Characters: Their Origin, Etymology, History, Classification and Signification. A Thorough Study from Chinese Documents. New York: Paragon and Dover (reprint of 1927 edition by Catholic Mission Press, originally printed 1915).

Yadav, N., M.N. Vahia, I. Mahadevan, H. Jogelkar. 2008.  “A statistical approach for pattern search in Indus writing” in International Journal of Dravidian Linguistics, 37 (1): 39-52.

Complex Signs in Proto-Elamite and Indus Script

Handmade replica of a proto-Elamite tablet.  Repeated symbols are numerical.
The third from the bottom on the left shows a symbol between "less than" and
"greater than" signs, a form of ligature also found in the Indus script (A + B + A).

I recently reread J.L. Dahl’s article on complex graphemes (or signs) in proto-Elamite and noted how often this author’s remarks also apply to the Indus script.  For example, he begins by noting that, although proto-Elamite was clearly inspired by proto-cuneiform, the Iranian system “exhibits a high degree of independent development of both sign repertoire and text structure” (2005: 1).  Similarly, the Indus script arose around the time that contact between the Indus Valley and Mesopotamia is documented in cuneiform, so it is quite reasonable to suggest that cuneiform (and perhaps Linear Elamite from Iran) influenced the development of the Indus script (see dates in Parpola 1994: 18).  However, even if the idea for the Indus script came from outside – from Iraq or Iran – the forms of the signs and the patterns of their use differ, indicating that same “high degree of independent development” cited by Dahl for proto-Elamite.

Dahl divides the complex signs of proto-Elamite into two basic types, with further subdivisions.  The first category is formed by the placement of a numeral in, on, or next to a non-numerical sign.  He terms these Complex Capacity Signs (CCS).  While we cannot tell whether Indus signs represent symbols for capacity, there are signs resembling CCS’s in this script.  For example, the POT acts as a terminal marker when there is nothing inside it.  But when an almost identical sign contains an apparent numeral (one to four “quotes”), the result is a ligature that behaves quite differently.  POTTED ONE, POTTED TWO, and so on are medial signs, not terminals.  Similarly, the CUP sometimes contains an apparent numeral, so there is a CUPPED TWO and a CUPPED THREE (KP signs 311 and 312). 

Seal H-76 with inscription, using a proto-Elamite transcription method, as described in this post:
MAN+FORK / HUNCHBACK / POTx1Q / CHEVRON%FISH / WHISKERED FISH / WINGED MAN / POT.
(Note that the addition of a "quote" inside the "pot" changes the sign's position from terminal to medial).

In a discussion of the BLANKET sign in a previous post, I noted the hypothesis of W. Fairservis that the differing numbers of “ticks” inside the rectangle might have significance (1992: 100-101).  There are BLANKETS with 4, 5, 6, 7, 8, and 10 short strokes (“ticks”) attached to the inner top and bottom.  In addition, sometimes those with 4 such strokes also have additional strokes in the center, usually two (“dashes”).  Occasionally, the sign contains “ticks,” “dashes,” and an additional “C” shaped mark attached to one side.  Fairservis thinks it possible these represent various weights.  However, he consider it unlikely that the CUP and POT containing “numerals” represent quantities (1992: 87). Without claiming to know the purpose of such “numerical” inclusions, I merely point out the existence of Indus ligatures containing both “numerical” and “non-numerical” elements.

The second type of complex grapheme in proto-Elamite is that containing two non-numerical symbols, according to Dahl (2005: 1).  These can be subdivided into three groups, the first of which is termed framing (written A + B + A).  The second subtype is formed by placing one symbol inside another (written A x B).  The third is simply the grouping of two symbols in close proximity, one after the other (A + B).  We can symbolize several Indus ligatures in a similar fashion.  The CEE symbol, for example, can appear singly (CEE) or doubled (CEE + CEE).  Additionally, a different sign may stand between two CEES (e.g., CEE + WINGED MAN + CEE).  Since there is also a reversed CEE (BACK CEE), a symbol can also be framed by “parentheses” (CEE + BIRD + BACK CEE).  In some cases, the positions of the “parentheses” seem to be reversed (BACK CEE + SKEWERED CHEVRON + CEE).  Less frequently, one sign appears inside another (MALLET x FOOTED STOOL).  Using the same type of notation, we can rewrite POTTED ONE as POT x SINGLE QUOTE (or, more briefly, as POT x 1Q). 

Detail of seal fragment M-1106 with partial inscription, now transcribed:
CUPxSPOON / MALLETxFOOTED STOOL / FOOT (?) ....

Not all types of ligatures can be written with these symbols, when it comes to the Indus script.  In some cases, the choice of one transcription system over another must be purely subjective.  For example, “caging” is a relatively common qualification of an Indus sign.  If we wish to define this as framing, then CAGED MALLET consists of three signs: STACKED 2 + MALLET + STACKED 2.  But if we wish to define this as enclosing one sign in another, then the same sign becomes STACKED 4 x MALLET (because the MALLET is inside STACKED 4).  Where a POST is attached to another element by a slanted line, we might choose to “read” the group from left to right on seals, from right to left on tablets (e.g., KP 31 is MAN + SINGLE POST, or more succintly, MAN + 1P; KP 74 is 1P + PRAWN).  Alternatively, it might be better to ignore the side an attachment is on, instead always writing the name of the dominant element first, followed by the (often smaller) attached portion (MAN + 1P, PRAWN + 1P).

Indus ligatures sometimes add an element above or below the main sign as well, something not demonstrated for proto-Elamite.  We could press the symbol for division into service to represent this relationship, placing the upper element first in the “equation,” followed by the lower element.  Thus, FISH UNDER CHEVRON would become CHEVRON ÷ FISH.  Since this symbol is not on the standard keyboard, this is not convenient, however.  Since the percentage sign resembles the mathematical division sign, I suggest we use it instead: CHEVRON % FISH; TABLE % EX.

Seal L-45 with inscription, now transcribed: BLANKETx8Q / POTx1Q / FISH /
4Sx(CHEVRON%FISH) / TABLE / GRAIN EAR.

In proto-Elamite, the third type of complex grapheme consists of two signs placed side by side.  The result is not a true ligature, but “it can be established that they form a distinct semantic unit that can be replaced by a CG of the second type” (Dahl 2005: 1).  In other words, a single concept is sometimes written A x B (with B inside A) and sometimes A + B (with B next to A).  This kind of replacement of a sign pair by a ligature cannot be demonstrated for the Indus script.  Thus, it is not a good idea to write the pair with the same “+” sign used in examples of attached elements.  Instead, we can pull the equal sign into use for this.  Thus, the common pair TWO POSTS plus FISH can be rewritten as 2P = FISH.

Another complication introduced by the Indus script is the combination of three elements in a single ligature.  For example, using notation similar to that for proto-Elamite, we can rewrite CIRCLED FORK as CIRCLE x FORK.  But if CAGED CIRCLED FORK is a FORK inside a CIRCLE inside a STACKED 4, how do we write it?  It seems reasonable to simply extend the system, writing the outermost element first and proceeding toward the inside: STACKED FOUR x CIRCLE x FORK (or more succintly 4S x CIRCLE x FORK).  Similarly, CARTWHEEL IN FAT EX IN DIAMOND can be shortened to CARTWHEEL x FAT EX x DIAMOND.  Or the BOXED BI-RAKE WITH EAR can be rewritten as (BOX x BI-RAKE) + EAR, using parentheses to show that the EAR is not attached to the BI-RAKE inside the “box” but to the BOX (the outer symbol containing the BI-RAKE).  When the attachment is inside, the position of the parentheses can show this as well.  Thus, TRI-FORK & VEE IN DIAMOND becomes DIAMOND x (VEE % FORK) – the VEE and FORK touch one another, both being inside the DIAMOND, with the VEE above the FORK (KP 390).  Where the FORK attached on the outside of the DIAMOND and the VEE is inside (KP 392), the transcription differs: (DIAMOND x VEE) + FORK – the FORK is attached to the outside of the DIAMOND, which encloses the VEE.

Detail from seal M-831 with inscription, now transcribed: COMB$MAN / CIRCLEx1Q.

Occasionally, one element is placed right on top of another element, so that the two overlap.  This is sometimes seen in Egyptian hieroglyphs, where a long but short glyph crosses the median of a tall and thin glyph, e.g., D59, where the arm, D36, forms the horizontal portion of a sort of cross shape and the leg, D58, forms the vertical portion.  Often such a ligature is read as if one glyph sat next to the other.  This type of ligature appears in the Indus script in the COMB BELTED MAN, where a horizontal “comb” seems to lie right on top of the MAN’s waist.  To represent this type of combination, I suggest using the dollar sign which itself resembles the letter “S” with a vertical stroke drawn right through it ($).  Thus, the COMB BELTED MAN becomes COMB $ MAN.  The signs I term “skewered” show this type of combination as well: SKEWERED CHEVRON is thus rewritten 1P $ CHEVRON.  This same character framed by inverted parentheses (formerly SKEWERED CHEVRON BETWEEN BACK CEE & CEE) can now be rewritten thus: BACK CEE + (1P $ CHEVRON) + CEE.

In some cases, though, it may be preferable to continue referring to a complex symbol with a simple name.  The BEARER, for example, might be seen as a ligature of MAN with two CIRCLES and possibly the TABLE (as the shoulder yoke).  But BEARER is much briefer and clearer than CIRCLE + (TABLE $ MAN) + CIRCLE.  When doing so shortens some of the clumsier designations, the use of symbols +, x, =, %, $, and parentheses is useful.  When the use of such symbols obscures the unity of a sign or becomes cumbersome, a simple name is preferable.

As a final note, Dahl observes that while proto-Elamite scribes shared a common stock of symbolic elements, the system was not fully standardized, a characteristic that “allowed the scribes who wrote the proto-Elamite tablets to generate signs in an ad hoc manner when needed, usually relying on a set of basic signs” (2005: 7).  This same situation characterized proto-cuneiform as well.  It seems quite likely that it was true of the Indus script, too.  This may be the reason that (1) there are so many rare signs (they were invented on the spot and used only for a short period by a small number of “scribes”) and (2) so many signs are vanishingly rare (the vast majority occur in the Corpus fewer than 100 times).

REFERENCES

Dahl, J.L. 2005: “Complex Graphemes in Proto-Elamite” in Cuneiform Digital Library Journal (3): 1-15. http://cdli.ucla.edu/pubs/cdlj/2005/cdlj2005_003.html

Fairservis, W.A. 1992. The Harappan Civilization and Its Writing: A Model for the Decipherment of the Indus Script. Leiden: E.J. Brill.

Parpola, A. 1994 and 2009. Deciphering the Indus Script. Cambridge: Cambridge University.

Sentences in the Indus Script?

The city seal of Ur, bearing possible writing or proto-writing (after Collon 2005: 106, no. 455).

Dominique Collon describes the inscriptions found on Near Eastern cylinder seals (1987/2005: 105-7).  The earliest examples, dating to the third millennium BCE, seem to contain only names, whether names of cities, deities, or humans.  More information appears in later seals, with names and multiple attendant titles by the time of Ur III (2334-2000 BCE).  Old Babylonian seal inscriptions generally give the owner’s name, his or her filiation (son or daughter of named person number two), along with a third element indicating that the owner is the servant of a particular deity or, less often, a king.  Professions sometimes appear in early inscriptions, but disappear by this time.  Prayers begin to appear at the end of this period.  By the time of the Kassites, prayers and incantations are common and can be fairly long.  An example of this type reads, “Oh Marduk, sublime lord, prince in whose hands the power of decision in heaven and on earth has been vested: the servant who worships you, by your look may he be happy” (2005: 107, no. 460).  This dates from after 1500 BCE, a time when the Indus script had essentially disappeared from use.  But the presence of some full sentences on these Mesopotamian seals suggests to some researchers the possibility that their neighbors, the Harappans, might also have written complete sentences on their own seals.

Inscription from a cylinder seal that depicts heroes battling lions; it reads "Naram-Sin,
god of Akkad, Urag, scribe, is his servant" (after Aruz 2003: 208, no. 134).

In this post, I will examine some of the inscriptions whose structure might suggest sentences in the Indus script.  A number of scholars have examined the inscriptions in Indus script, looking for positional regularities (signs that typically begin or end an inscription) and for small groups of signs that generally appear together (usually pairs of signs, but sometimes groups of three).  Among these, the study by M. Korvink carefully distinguishes between structural regularities on the one hand, and possible meaning on the other.  He classifies the components of the inscriptions as (1) prefix, containing one or more variable signs plus a single constant; (2) the medial segment, which often contains one or more fish-like signs in a standardized order; and (3) the terminal, containing one or more signs in a standard order (2007).  I abbreviate these sections as P, M, and T, respectively in the following.

A minority of inscriptions include sequences other than PMT (in which each segment is optional).  These are the focus of this post.  The first inscription I will analyze is from seal M-670:

CUPPED POST / THREE POSTS / SPEAR / SINGLE QUOTE / SINGLE POST / POT-HATTED BEARER

The first two signs form a common pair (CUPPED POST / THREE POSTS), together making up M.  They are followed by SPEAR, which is T.  The presence of the next sign, SINGLE QUOTE, is rather surprising, since it is normally the constant sign in P.  I will assume here that this sign retains this function despite its position after a terminal.  The fifth sign, SINGLE POST, must be another M, followed by another T, POT-HATTED BEARER.  Thus, the inscription as a whole seems to be analyzable as MTPMT.  In some inscriptions, such an analysis indicates simply two units of information, each of which could be an independent inscription: MT + PMT.  But there is an anomaly in the group of signs preceding SINGLE QUOTE.  Korvink notes that the constant causes other signs to move to initial position to become part of P (2007: 25-26).  The signs that usually appear as variables in P, though, are found in M when not followed by the constant.  What appears to be going on here, instead, is that a whole “phrase” (MT) is functioning as the variable.  We might express this as ^(MT)PMT.  In other words, the initial MT phrase + constant is P in this case.  Such an analysis removes the anomalous character of the inscription.

Seal M-782 with inscription: TRIPLE BRICK / POT / SINGLE QUOTE / CAGED TETRAPOD / POT.

Another example of a “phrase” as the variable in P is seal M-782:

TRIPLE BRICK / POT / SINGLE QUOTE / CAGED TETRAPOD / POT

The TRIPLE BRICK is M, followed by the most frequently occurring T, POT.  This MT “phrase” again appears to be the variable portion of P, as it is followed by a constant, SINGLE QUOTE.  There follows another MT sequence (CAGED TETRAPOD + POT).  Again, the overall analysis is ^(MT)PMT.

A third, similar example is found on seal M-1135:

CUPPED THREE / COMB / BI-QUOTES / LAMBDA / POT

Once more, the initial sign is M, with the second sign T.  This MT phrase precedes a constant, BI-QUOTES, thus forming P.  After this complex segment, the rest of the inscription includes a single sign for M and repeats the same T for an overall analysis of ^(MT)PMT.

Bar seal M-1267 with inscription: VEE IN DIAMOND (partially restored here) / BI-QUOTES / HUNCHBACK / FLANGE TOPPED POT / POT / SINGLE QUOTE / BI-RAKE (note the presence of two different prefix constants here).

A more complex type appears in M-1267:

VEE IN DIAMOND / BI-QUOTES // HUNCHBACK // FLANGE TOPPED POT / POT / SINGLE QUOTE // BI-RAKE.

In this example, the first five signs form a complete series, P (VEE IN DIAMOND + BI-QUOTES), M (HUNCHBACK), and T (FLANGE TOPPED POT + POT).  Since this precedes a prefix constant, SINGLE QUOTE, the whole initial PMT can be analyzed as the prefix variable.  The final sign, BI-RAKE, is a second M.  The overall analysis is ^(PMT)PM.  The same pattern appears in the inscription M-1284: SINGLE POST / STRIPED TRIANGLE / BI-QUOTES // CIRCLE TRI-FORK / WHISKERED FISH // SPEAR // SINGLE QUOTE // BI-RAKE, or ^(PMT)PM.

Bar seal H-143 with inscription: TWO POSTS / FAT EX IN DIAMOND / POT / SINGLE QUOTE / FLAIL OVER FOUR QUOTES / CIRCLED TRI-FORK / POT HATTED BEARER (identical to H-8 except the FLAIL in H-143 is AY in H-8).

All the examples quoted thus far have been from Mohenjo daro.  But this interesting phenomenon also occurs at Harappa.  Inscription H-8 has the same structure as the last two examples above: BOAT / CROSSROADS EX / BI-QUOTES // CUPPED SPOON / THREE POSTS / MALLET / QUAD-FORK // POT // SINGLE QUOTE // PRICKLY CORN HOLDER, or ^(PMT)PM.  The previous type above appears in the Harappan inscription H-27: TWO POSTS / FAT EX IN DIAMOND // POT // SINGLE QUOTE // AY ON FOUR QUOTES / CIRCLED TRI-FORK // POT HATTED BEARER, or ^(MT)PMT.  The inscription on H-143 is identical except that instead of AY ON FOUR QUOTES, there is a FLAIL OVER FOUR QUOTES.

Broken seal C-24 with inscription: DIAMOND / CHEVRON / SINGLE QUOTE / BI-QUOTES / TWO POSTS / CAGED FISH / BI-QUOTES / CUPPED TWO / MAN HOLDING DEE-SLASH / TRI-FORK (last 3 signs are broken).

A somewhat different complex example comes from Chanhujo daro, where C-24 reads: DIAMOND / CHEVRON / SINGLE QUOTE / BI-QUOTES // TWO POSTS / CAGED FISH // (2^nd row) BI-QUOTES // CUPPED TWO / MAN HOLDING DEE-SLASH / TRI-FORK.  The signs in the second row are not complete because the seal is broken horizontally.  Before the BI-QUOTES in this row, there is a small loop that could be part of a broken sign.  I am assuming that it is not, instead forming the ear of an iconic animal (not the unicorn, since there is no horn, and not the elephant, which has no such ear).  But it might be the head of an anthropomorphic sign, since the MAN HOLDING DEE-SLASH has a head of almost the same shape.  However, if I have transcribed the inscription correctly, its analysis differs from previous examples in that the first line forms PM.  The second line, beginning with BI-QUOTES, would seem to make of the first a prefix variable.  Then the last three signs form a second M.  The overall analysis is ^(PM)PM.

In the examples cited and in other, similar inscriptions, the structural complexity does not prove that these are actually sentences.  This complexity is similar to that of sentences with embedded clauses.  In English, subordinate clauses follow the noun or pronoun they describe (as in “That’s the man who bought the car,” where “who bought the car” describes “the man”).  In Chinese and in Turkish, such clauses precede the noun or pronoun they describe.  Languages such as English, with clauses that follow, are termed right-branching, while the Chinese or Turkish type is termed left-branching.  Some scholars have interpreted similar evidence as indicating that the language underlying the Indus script is left-branching (Koskenniemi and Parpola 1982: 12, citing their own conclusion and that of the Soviet scholars, Y.V. Knorozov, M.F. Albedil, B.Y. Volchok).  This is one reason that Parpola and others think that the Indus script represents a Dravidian language.

In some Bronze Age scripts, though, a single or double stroke functions as a punctuation mark of a sort, rather than a phonetic or ideographic symbol.  This is the case in Assyrian cuneiform and in the Linear B writing of Mycenaean Greece.  It is possible that this is what SINGLE QUOTE, DOUBLE QUOTE, and PINCH are in the Indus script.  That is, these may simply be marks separating units of information.  Each prefix would then represent one unit of information and the following M(T) a second such unit.  In the complex examples cited above, there would be more, namely three units of information. 

This may be the best analysis for now, rather than interpreting the apparent complexity as evidence of embedding.  Where I have analyzed an inscription above as ^(MT)PM(T), I now amend this to MT “ M(T), with the double quotation mark indicating the presence of a prefix constant which marks a boundary between the first and second units of information.  Presumably the three constants do not all indicate precisely the same thing, though, since occasionally there are two different ones in the same inscription (as in H-8 and M-1284).  This implies that they are not just separating two units, but also have some significance or meaning of their own. 

In conclusion, then, there is apparent complexity in some inscriptions.  Such complexity may be interpreted as evidence for sentences.  But given what is known about other scripts, it seems more likely that there is a simpler explanation.  There is stronger evidence that some inscriptions are compound, in that they contain more than one unit of information.  But what seems to be embedding (and thus complex sentences) on first examination is better analyzed simply as the separation of the multiple units.

REFERENCES

Aruz, J., ed. 2003. Art of the First Cities: The Third Millennium BC from the Mediterranean to the Indus. New York: Metropolitan Museum of Art and Yale University Press.

Collon, D. 1987 & 2005. First Impressions: Cylinder Seals in the Ancient Near East. London: British Museum.

Joshi, J.P. and A. Parpola. 1987. Corpus of Indus Seals and Inscriptions. 1. Collections in India. Helsinki: Suomalainen Tiedeakatemia.

Korvink, M.P. 2008. The Indus Script: A Positional Statistical Approach. Gilund Press (Amazon).

Koskenniemi, K. and A. Parpola. 1982. A Concordance to the Texts in the Indus Script. Helsinki: Department of Asian and African Studies, University of Helsinki.

Shah, S.G.M. and A. Parpola. 1991. Corpus of Indus Seals and Inscriptions. 2. Collections in Pakistan. Helsinki: Suomalainen Tiedeakatemia.