The Other Side Garment spec sheet
Spec sheet ← All writing
The Other Side · Garment spec sheets Revision A · August 2026
Garment spec sheet 02

The Tamil configuration

Dress and ornament in the Tamil sources, in two dated layers: the Sangam corpus, c. 300 BCE – 300 CE, and the epic Silappadikāram, which is later. The second layer is where almost all the detail is. It is labelled throughout, and it is not Sangam.

How to read this sheet

Layer A — Sangam corpus, c. 300 BCE – 300 CE. Thin on garments, and thinner than most popular accounts admit. What it gives, it gives in passing.

Layer B — Silappadikāram, composed somewhere between the 2nd and 6th centuries CE and usually placed in the 5th. Extremely rich: named ornaments, materials, construction, worn head to toe. Anything drawn from it is post-Sangam and must be dated as such wherever it is shown.

1 · Layer A — what the Sangam corpus supports

c. 300 BCE – 300 CE
Element What the sources give Status
Lower garment Unstitched cloth wrapped from the waist down. Read in the secondary literature as running from the loins to the ankles, with the torso left uncovered. Attested
Kachchu / vambu In this period, a waist-band holding the lower garment in place, worn by both sexes. Not a breast-band — see the correction below. Attested
Oliyal An upper cloth, associated with wealthier households and worn by women and men alike. Attested
Breast-band Contested. One scholarly reading finds no reference to a breast-band anywhere in the Sangam classics; the same account notes that Akanāṉūṟu appears to refer to a kind of kucha-bandha. Treat as unresolved, not as established. Disputed
Cotton and weaving A working cotton-weaving economy. Widowed women employed in cotton work are referred to as paruthi pendir. Attested
Flowers and scent Flowers worn in plaits and tufts — jasmine among them — and sandal and flower perfumes. Beads and shell ornament in common use. Attested
The sari A single continuous garment covering upper and lower body has no place in the Sangam works, and none in Pallava-period art either. Absent

Correction we are making to our own work. Our Sangam plates were captioned with a kachchu breast-band. That is a later usage. In the Sangam period the word denotes a waist-band; the breast-band called kachchu, vambu or vaar in Tamil is traced to Satavahana influence reaching Tamil country in the medieval centuries, and by the Chola period it was a stitched, embellished garment made by tailors — which is neither unstitched nor Sangam.

The breast-band on our Sangam figures is therefore our own addition, and is now labelled that way. It is defensible as a modesty choice in an illustration; it is not evidence.

2 · Layer B — Silappadikāram

Post-Sangam · composed between the 2nd and 6th centuries CE, usually placed in the 5th

This is the source worth building from, and the reason to be strict about labelling. The epic turns on an ornament — its title means the tale of an anklet — and it describes dress, jewellery and the cities that made them in detail no earlier Tamil text approaches. The courtesan Madhavi's ornament is listed from head to toe in a single passage.

Named piece As described
Mekalai A dancer's girdle of thirty-two strands of large pearls, worn over a blue cloth embroidered with flowers. The most makeable single item in the whole epic, and the most specific.
Pullaham Head ornament, the anchor of the hair set.
Toyyaham (sidevi) The central component of the pullaham — a round set with gemstones, with pearl dangles.
Thenpalli, vadapalli Curled, centipede-like ornaments worn on either side of the head, behind the pullaham. Commonly misidentified today as the sun-and-moon pair.
Kandigai, sarapalli Further pieces in the same head-to-toe list.
Silambu The anklet the epic is named for, and the object the plot turns on.

The setting matters as much as the list. Puhar is described as a city of gold, pearls and precious stones, with the jewellers holding its main street — an ornament economy, not just an ornament vocabulary. And much of this jewellery is still worn by dancers and brides today, which is worth saying out loud: the continuity is real in this layer, and it is a fifth-century continuity, not a Sangam one.

Reference to pin. The head-to-toe passage is cited in the translation literature at the Kadaladukādai section, verses 76–98. We are checking that against a printed edition before it goes on a plate caption.

3 · Making the lower garment

As in sheet 01, no source gives a measurement, and as in sheet 01 that is a feature of the garment rather than a gap in the record: an unstitched cloth has no size, only a wearer. Take the wearer's waist-to-ankle length and waist circumference, and cut from the formulae.

Configuration Cut length Cut width
Full length, waist to ankle Waist circumference × 2.5 + 45 cm for the tuck Waist-to-ankle measurement + 8 cm
Working length, tied high (field, water) Waist circumference × 2 + 35 cm Waist-to-knee measurement + 8 cm
Waist-band (kachchu / vambu) Waist circumference + 40 cm tie allowance 6–10 cm, folded from wider cloth

Derived, not attested. The formulae are ours. Handloom cotton in a plain weave is the default cloth; the corpus supports a cotton economy and gives no weave count. Cut long on the first attempt.

What we do not know
  • Whether any breast-band was worn in the Sangam period at all.
  • Measurements, weave counts and dye colours, in either layer.
  • How much of the epic's ornament was real dress and how much was literary display.
  • Regional difference between Chera, Chola and Pandya country.
Do not
  • Call kachchu a breast-band in a Sangam-period context.
  • Put Layer B ornament on a pre-300 CE figure without dating it in the caption.
  • Add a stitched blouse, a fitted band, or elastic to either layer.
  • Reach for Chola-period temple jewellery sets — they are five to fifteen centuries later than Layer A.

4 · Captions to use

Layer A

Reconstructed from the Sangam corpus, c. 300 BCE – 300 CE. Attested: the wrapped lower cloth, the waist-band (kachchu), an upper cloth (oliyal) in wealthier households, flowers worn in the hair. Our addition: the breast-band — the corpus does not clearly name one.

Layer B

Ornament from Silappadikāram, a Tamil epic composed between the 2nd and 6th centuries CE — post-Sangam, and later than the period usually claimed for this dress. Attested: the named pieces and their materials. Inferred: how they sat on the body.

5 · Sources

  • The Sangam corpus — Akanāṉūṟu, Puṟanāṉūṟu, Kuṟuntokai and the Pattuppāṭṭu. Dress appears incidentally, in simile and description, not in any dedicated passage.
  • Silappadikāram, attributed to Ilangō Adigal. Composed between the 2nd and 6th centuries CE; commonly placed in the 5th. Post-Sangam throughout this sheet.
  • Studies of pre-Pallava and Pallava-period cultural history — for the reading that the Sangam classics carry no breast-band reference, and for oliyal and kachchu as waist-band.
  • Costume-history surveys of medieval South India — for the later arrival of the stitched breast-band into Tamil country and its Satavahana antecedents.
  • Nanditha Krishna, on traditional South Indian jewellery, and the translation literature on Madhavi's ornament list.

Poem-and-line references for Layer A, and the printed-edition reference for the Madhavi passage, are being pinned and will be added in the next revision. Where this sheet says a thing is disputed, we would rather be shown the passage than be right.

Reconstruction, not an object record. Attested, inferred and dated layers are marked throughout.