'Phrase Structure' in Indus Script Continued

In my previous, I described inscriptions containing a single ‘phrase’ or segment and those containing two such ‘phrases.’  I noted that most inscriptions consisting of a single ‘phrase’ contain only the medial segment (abbreviated M) as analyzed by Korvink (2008).  There are, however, a few instances where the inscription apparently contains only the so-called prefix (abbreviated P).  Only one clear example exists, as far as I know, of an inscription with the third segment alone, the terminal (abbreviated T).  Given this data, I tentatively conclude that M is obligatory except in rare instances, when P is sufficient.  But usually both P and T are optional.  Among inscriptions comprising two ‘phrases,’ there are many PM and even more MT sequences.  PT occurs rarely.  Other theoretical combinations either do not occur at all (PP, MP, TP, TM, TT) or are quite rare (PT).  For the moment, I cannot determine whether MM occurs as this seems to depend on definition of M.

At this point, I will continue the discussed with 3-phrase sequences.  The complete sequence of P plus M plus T can be termed a “sentence” for this purpose, although we cannot determine whether the symbol sequence actually represents any linguistic element.  Given the sequences we have seen thus far, we expect to see PMT and MTM in large numbers; there may also be a few PMP, MTP, PTP, PTM, and MPT.  If M can be duplicated (MM), we may also MMM where an inscription contains three lines or an object has three inscribed sides.

Segment	P	M	T
P	PP	PM (72+)	PT (3+)
M	MP	MM	MT (136+)
T	TP	TM	TT

Theoretically possible two-phrase inscriptions in the Indus script with those not found gray.

Segment	PP	PM	PT	MP	MM	MT	TP	TM	TT
P	PPP	PPM	PPT	PMP (1)	PMM (5)	PMT (357)	PTP (0)	PTM (1)	PTT
M	MPP	MPM (0)	MPT (0)	MMP	MMM (1)	MMT (3)	MTP (3)	MTM (120)	MTT
T	TPP	TPM	TPT	TMP	TMM	TMT	TTP	TTM	TTT

Theoretically possible three-phrase inscriptions with those not found in gray.

I find 357 examples of PMT (the postulated complete ‘sentence’) and 120 examples of MTM (or, more properly, MT-M, indicating two ‘sentences,’ one containing two ‘phrases’ and the other containing a single ‘phrase’).  A few other types do occur, all rare:

·         PMP (or rather PM-P) appears once (M-367),

·         PMM (PM-M?) appears at least five times (M-577, M-1341, H-244, H-702, C-11, and possibly L-48 and C-6, depending on the interpretation of a single sign),

·         MTP (MT-P) appears three times (K-16, M-1482 and 1483 with MT as P, M-1103),

·         PTM (PT-M) appears once (M-257),

·         MMM occurs once (H-363, with a M-only inscription on each of three sides),

·         MMT may occurs three times (on tablets M-483 and -484, side A contains a M-only inscription while side B contains MT, giving either M-MT or MT-M; on C-33, the top row of a two-line inscription is M while the bottom row is MT; on Ad-7, side A contains a M-only inscription while side B contains MT, giving either M-MT or MT-M),

·         although MPM (M-PM) and MPT (M-PT) seem to be possible, they do not occur.

Thus, there are several theoretically possible phrase combinations that do not actually occur.  The hypothesized sequence, (P)M(T) predicts most but not all sequences that occur.

When it comes to inscriptions of four phrases, the situation is similar.  The following occur:

·         PM-PM (4 examples: M-325, M-1190; K-4, C-24)

·         PM-MM (2 examples: M-1188, C-23)

·         PM-MT (2 examples: possibly M-573  and -574, M-1169)

·         PT-MT (1 example: M-267)

·         MT-PM (7 good examples: M-378, M-380, M-1310, M-1314, M-1340, M-1474-81 & -525; H-420; though all but the first should be encoded differently – perhaps as (MT)PM -- because they show possible embedding)

·         MT-MM (5 good examples): duplicates M-599, M-600, and M-1555 through M-1560 (with MT-M on side A, additional M on side B, hence MT-M.M); H-239 (MT side A top row plus M side A bottom row, additional M on side B, hence MT-M.M); H-350, H-352 through H-357, H-987 with MT side A, M side B, M side C, hence MT.M.M)

·         MT-MT (9 good examples): M-980, M-1224 (with a MT sequence on each of two sides, hence MT.MT), M-1429, M-1452 (MT.MT), H-58, H-132, H-160, H-874, H-909

·         PMT-P should be possible but I find no unequivocal examples.  The closest possibility is M-1127: BI-QUOTES // LOOP MAN HOLDING SLASH / BIRD BETWEEN PARENS // POT // (2^ND ROW) FAT EX / PINCH.  It is conceivable that the bottom row should be “read” before the top as this would yield a more typical PMT sequence (though still with an extra constant in the prefix).

·         PMT-M (37 examples): M-7, M-28, M-67, M-68, M-79, M-101, M-154, M-171, M-235, M-266, M-296, M-309, M-391, M-393, M-509 and M-510,M-421, M-598 (PMT.M with inscriptions on two sides), M-629, M-634, M-697 (PMT.M), M-720, M-777, M-792, M-821 through M-824 (PMT.M), M-833 (PMT.M),  M-851, M-921, M-1107, M-1150, M-1445 (PMT.M), Nd-1, Q-6 (PMT.M), H-21, H-129, H-360 (PMT.M), K-28, C-1.  In some cases, the inscription is on a single line, in other cases on two lines on a single side, and in still other cases on two different sides of an object.  The shortest inscriptions of this type contain five signs, two for the P, one each in the M and T.  The longest inscription of this type includes 12 signs, near the upper limit of “long” texts.

I find over 300 inscriptions incorporating still longer and more complex strings, to be discussed in the next post.