The quickest test is botany. Persian ornament tends toward stylised, repeating, geometrically resolved plant forms — the arabesque and the islimi scroll. Mughal ornament, from the 1630s onward, tends toward a single flowering plant observed the way a botanist would draw it: identifiable species, visible stamens, one plant per panel. One is a pattern. The other is a portrait.
Stand in front of the cenotaph screen at the Taj Mahal and look at what is inlaid into the marble. Not the calligraphy — the flowers below it.
You can identify them. Not “a flower”: a poppy, a tulip, an iris, a narcissus, each with the right number of petals and its leaves attached in the right arrangement. Whoever cut those stones had looked at actual plants, or at a very good drawing of actual plants.
Now put a Safavid tile panel from Isfahan beside it. The vocabulary is unmistakably related — the same palmettes, the same curving stems, the same instinct for a filled surface. But the flowers do not resolve into species. They resolve into geometry. Follow any stem and it will loop back and meet itself.
That difference has a date and a cause, and it is the whole of this article.
One inspired the other. The other made it Indian.
The short answer: nine ways to tell them apart
| Attribute | Persian (Safavid and earlier) | Mughal (especially post-1630) |
|---|---|---|
| Plant form | Stylised, idealised, composite — a “flower” not tied to a species | Botanically observed — poppy, tulip, iris, narcissus, lily, identifiable by petal count and leaf form |
| Composition | Continuous scrolling field (islimi), infinite repeat | Discrete single flowering plant, often one per panel or niche, with visible ground line |
| Symmetry | Mirror and rotational symmetry, often four- or eight-fold | Bilateral symmetry about a vertical axis; the plant stands upright |
| Stem behaviour | Stems interlace and pass over and under each other continuously | Stems rise from a root or tuft and terminate; less interlacing |
| Fauna | Hunting scenes, lions, deer, phoenix (simurgh), hybrid creatures | Parrots, peacocks, hoopoes, butterflies — observed rather than mythic; often sparse |
| Signature plant | Cypress, almond, pomegranate, the boteh | Poppy and tulip above all; then champa, mango, and native flowering species |
| Borders | Multiple concentric bands, dense guard stripes | Fewer, wider bands; the field is allowed to breathe |
| Palette | Cobalt and turquoise dominant; strong blue-white contrast | Wider range; on marble, the field is white and the colour is the inlay stone |
| Spatial logic | The pattern fills the surface it is given | The plant is placed within an architectural frame it does not fill |
Print that table and take it to a museum. It works on carpets, tiles, textiles, manuscript borders and carved wood.
Tell them apart in five seconds
- Look for a ground line. If the plant grows out of a tuft, a mound or a visible root, you are almost certainly looking at Mughal work. Persian floral fields rarely give a plant somewhere to stand.
- Try to name the flower. If you can say “that’s a poppy,” it’s Mughal. If the best you can do is “that’s a flower shape,” it’s Persian.
- Follow one stem. If it loops back into the design and never ends, it is an islimi scroll — Persian vocabulary. If it starts at a root and stops at a bloom, it is Mughal.
- Count the plants in the panel. One upright plant, centred: Mughal. Continuous repeat with no obvious count: Persian.
- Check what’s in the corners. Persian design fills corners with the pattern’s own logic. Mughal design tends to leave the frame as a frame.
None of these is infallible. All of them together are reliable, and the failures are interesting — see the section on the hard cases below.
Why the difference exists: a lineage, honestly stated
There is no version of this story in which Mughal ornament is not downstream of Persian ornament. Saying otherwise is bad history and it insults people who know better.
The Mughal court was Persianate from the start. Babur, who founded the dynasty in 1526, was a Timurid prince writing in Chaghatai and Persian. Under Akbar (r. 1556–1605), the V&A records that Hindu and Muslim artists from northern India worked directly alongside Iranian masters, and Persian became the official administrative language in 1578. The painters, the calligraphers, the ateliers, the technical vocabulary — Iranian.
So the interesting question is not where did it come from. It is when did it stop sounding Iranian.
The answer is roughly the second quarter of the seventeenth century, and it involves an unlikely source: Dutch and Flemish botanical prints.
European Jesuit missionaries brought engraved herbals to the Mughal court. Mughal artists saw plants drawn under the conventions of European botanical illustration — single specimen, upright, isolated on the page, drawn from observation so that a physician could identify it. The V&A notes that flower designs in Mughal album borders were “derived from European engravings, probably seen by Mughal artists in the borders of Netherlandish prints.”
Under Shah Jahan (r. 1628–1658), that convention was absorbed, translated, and set into stone. The flowering plant became the defining motif of the reign — carved in relief on marble dados, inlaid in parchin kari (the Mughal pietra dura, itself developed after Florentine panels reached the court), woven into carpets, embroidered onto sashes.
So the Mughal floral style is a genuine hybrid: Persian grammar, European botanical observation, Indian flora and Indian hands. That is a more interesting story than either “it’s all Persian” or “it’s uniquely Indian,” and it is the accurate one.
Where the accent changed: what Indian artisans actually did to the vocabulary
Three specific substitutions are worth naming, because they are where you can see the dialect forming.
The cypress gave way. The cypress is a Persian signature — an evergreen with strong Zoroastrian associations of eternity and endurance, and one half of the origin story of the boteh. Mughal and later Indian work keeps it, but crowds it with plants that grow in the subcontinent: champa, mango, plantain, lotus.
The mango entered the pattern. The boteh — the teardrop that Europe would eventually call paisley after the Scottish town that mass-produced imitations in the 1860s — is Persian in origin, likely a fusion of a floral spray with a cypress. In India it acquired a local name and a local reading: ambi, from the Sanskrit for mango. The form barely changed. What it meant changed completely.
Birds became observable. Persian ornament is fond of the simurgh and of hybrid creatures that do not exist. Mughal work under Jahangir — whose court painter Mansur was, by any reasonable standard, a natural-history illustrator — prefers birds you could point at: parrots, hoopoes, peacocks. The peacock in particular carries so much prior Indian iconographic weight that its arrival is not a borrowing at all.
That is what a dialect is. Same grammar, different nouns, different things worth saying.
The hard cases (where the five-second test fails)
Any honest guide has to include the places its own method breaks.
Kashmir shawls, eighteenth and nineteenth century. The boteh here is Persian in ancestry, Indian in execution and increasingly European in market. Trying to sort a Kashmir shawl into one bucket is a category error.
Deccani work. The Sultanates of Bijapur, Golconda and Ahmadnagar had their own direct maritime links to Iran, largely independent of Delhi. Deccani ornament can be more Persian-looking than contemporaneous Mughal work while being no less Indian.
Late Mughal and Company-period revival. By the nineteenth century, workshops were producing deliberate pastiche for European buyers. The motifs are correct and the logic behind them has evaporated. Date matters more than vocabulary here.
Anything made for export after 1850. Motifs were mixed to order. A Persian border on a Mughal field on a piece made in Punjab for a London merchant is not a puzzle to be solved; it is a commission.
Carpets generally. The Mughal carpet workshops were staffed by weavers with Persian training producing for Indian patrons, sometimes to Persian cartoons. The Ashmolean’s Jameel Centre holds Mughal carpets whose attribution has been argued over for a century.
If a source tells you the distinction is always clean, that source is selling something.
Why any of this matters if you are not a historian
Because it changes what you see, and because it changes what you are being sold.
Walk through a furniture showroom in Jodhpur or a home-décor floor in Delhi and you will see “Mughal-inspired” applied to almost anything with a curve in it. Most of it is neither Mughal nor Persian — it is a generic orientalist floral that belongs to no tradition and answers to no rules. That is not a moral failing, but it is worth being able to see.
The pieces that are genuinely working within one of these vocabularies look different in a way you can now name. A carved panel with a single upright poppy rising from a tuft inside a cusped arch is making a specific, seventeenth-century, north Indian argument. A repeating interlaced scroll on a marble tabletop is making a different, older, Iranian one. Both are legitimate. Knowing which is which is the difference between owning an object and owning a decoration.
It also tells you something about the maker. Motif vocabularies are transmitted through workshops, and a carver who can execute a correct parchin kari flowering plant learned it from someone who learned it from someone. Hemraj Kumawat, whose hand-cut thikri mirror work we have documented, went from a classroom in Tonk to Jaipur at eighteen to learn a specific regional idiom. That lineage is legible in the work if you know what to look for.
The Folklorei view
Design historians have argued about Mughal-versus-Persian attribution for a hundred and fifty years, and the argument has not been settled because it is partly the wrong question. Ornament is not owned. It moves along trade routes, gets carried by migrating craftsmen, is copied badly and then copied well, and acquires new meanings in each place it lands. The boteh went Sasanian Iran → Mughal India → Kashmir → Paisley, Scotland, and came back to India as a print called paisley. That circuit is the normal condition of pattern, not an aberration.
What can be owned is competence. Somebody in a workshop knows how a poppy is drawn in this tradition, how many petals it takes, where the leaf attaches, how deep the undercut goes at the stem. That knowledge is local, embodied and losable in a generation.
Go back to that cenotaph screen. Somebody looked at a real poppy, and then cut its shape into marble so precisely that four hundred years later you can name the species from six feet away. That is not decoration. That is a person paying attention, in stone, permanently.
Frequently asked questions
Is Mughal art the same as Persian art?
No, though it descends directly from it. Mughal art began as a Persianate court style — Iranian masters trained the first Mughal ateliers, and Persian was the court’s administrative language from 1578. It diverged over the following century, most sharply under Shah Jahan, when a botanically observed single-plant motif replaced the continuous stylised scroll as the dominant floral form.
How do you identify a Mughal motif?
Look for a single upright flowering plant with an identifiable species, growing from a visible root or tuft, placed centrally within an architectural frame it does not fill. Poppies and tulips are the strongest indicators. Persian floral work is more likely to be a continuous interlaced scroll of idealised, unidentifiable blooms.
What is the difference between Persian and Indian floral patterns?
Persian floral pattern is generally stylised and geometrically resolved — the design is a system. Indian floral pattern after the Mughal period is generally observational — the design is a specimen. There is enormous overlap in the middle, particularly in carpets and in Deccani work.
Where did the Mughal floral style come from?
Three sources at once: Persian ornamental grammar inherited from Timurid and Safavid Iran, European botanical illustration brought by Jesuit missionaries in the early seventeenth century, and Indian flora and craftsmen. It is a genuine hybrid rather than a borrowing.
What is parchin kari?
Parchin kari is the Mughal stone inlay technique in which shaped pieces of semi-precious stone are set into carved recesses in marble. It developed at Shah Jahan’s court after Florentine pietre dure panels reached India, and it is the technique used for the floral inlay at the Taj Mahal.
Is paisley Persian or Indian?
The motif — boteh in Persian, ambi in India — originated in Iran, probably as a fusion of a floral spray with a cypress. It became strongly associated with Kashmir through the shawl trade, reached Europe in the nineteenth century, and was named after Paisley in Scotland, which mass-produced imitations in the 1860s. It is Persian in origin and Indian in the form most of the world recognises.
What is an arabesque?
An arabesque, or islimi in Persian usage, is a continuous, rhythmic, interlacing scroll of stylised vegetal forms. Its defining property is that it has no natural end — the stems loop back into the composition indefinitely.
Are Mughal motifs still made by hand in India?
Yes, in several centres — marble inlay in Agra, stone and wood carving in Rajasthan, and inlay and carving workshops across the north. The vocabulary survives where the workshops survive, which makes it more fragile than it looks.