At first glance, the long dark bands in this carpet might look like a bold geometric pattern. Look closer and fish appear inside them. The bands are watercourses. They divide planted plots, meet smaller channels and lead the eye past trees and flowers. A garden carpet asks you to read across its surface as though you were looking down on a planned landscape, even while its trees stand upright in profile.

Two surviving carpets held by the Metropolitan Museum of Art make that idea unusually clear. Both are attributed to western Iran, but they organize water and planting in different ways. One turns a single intersection into the heart of the design. The other repeats a garden unit along its length. Comparing them tells us more than treating chaharbagh, or “four garden,” as a fixed motif that every weaver copied.

Start with the water, not the flowers

A flower appears on countless Persian rugs. It does not, by itself, make one a garden carpet. In a garden layout, broad and narrow bands organize the whole field. They can be read as channels or paths separating areas of planting. A pool-like shape often marks a junction. Trees, shrubs, birds and flowers occupy the spaces between. The border holds this imagined landscape within a frame.

This is a change of viewpoint. The waterways and plots appear from above, but individual trees may be drawn from the side so their trunks and branches remain recognizable. The two perspectives coexist because the carpet is a designed image, not a surveyor's scale plan. Its geometry helps the eye move through a place that could never be seen from precisely one position.

Historic four-part garden plans in Kurdistan invite comparison with the woven divisions. Similar arrangements in gardens and carpets may point to a regional visual tradition. The surviving objects support that comparison. They do not, on their own, tell us who commissioned each carpet or prove that a particular weaver copied a particular garden.

A garden gathered around one center

The first Met carpet, made in Iran's Kurdistan region in the second half of the eighteenth century, gives the eye a central pool-like shape. Water-patterned areas and planted compartments spread around it. Rows of trees, flowering stems and small birds fill the surrounding ground. The trees are more stylized than architectural drawings, yet the contrast between broad channels and planted areas remains legible.

Trace the field from its outer edge inward. The border forms an enclosure. At each side, tree forms and flowering shapes occupy rectangular zones. Striped channels separate those zones from the center. The pool-like form is therefore more than a decorative medallion: it is the point at which the larger system becomes easy to read. At the same time, calling it a real basin in a documented garden would go beyond the evidence. We can describe its visual function without assigning it a known address.

The Met describes this object as a combination of a bird's-eye four-part garden and profile views of birds and trees. Its curators also call it more stylized than an earlier garden carpet now in Jaipur. That comparison matters because the geometry did not erase the garden. It altered how the garden was pictured. A reader can still distinguish the water-bearing bands from the planted spaces even when every leaf is not botanically precise.

This carpet has appeared in our discussion of Persian rug medallions, where its center helps explain how a basin-like form can organize a field. Its wider story lies beyond that central shape. The planted compartments, channels and edge work together, and the next carpet shows that a garden need not stop at one center.

When a garden repeats along the rug

A second Met carpet, dated to about 1800 and attributed to northwestern Iran or Kurdistan, is much longer than it is wide. Its wide central stream runs down the rug, crossed by narrower waterways. Fish make the central band hard to mistake for an ordinary stripe. Planted plots lie to either side, and three prominent pool-like shapes punctuate the length. In its current form, the design reads as two repetitions of a four-garden arrangement.

The full photograph rewards a slow read. Begin at either end and follow the fish through the middle. At a crossing, shift your eye sideways into the smaller channels. The planted rectangles contain trees drawn with different shapes and colour contrasts. Then return to the long stream and continue to the next intersection. The pattern gives the viewer a route as well as a repeated framework.

This changes the meaning of “four” in a useful way. Four-part division describes an organizing principle, not a command that the finished rug must contain exactly four identical boxes. A long carpet can extend the arrangement, repeat intersections and vary its planting. Its maker had to make the garden work within a rectangular textile of a particular size. A composition may be symmetrical without being a mechanical grid.

There are also differences between the two Met objects that should remain differences. The eighteenth-century rug concentrates its design around one dominant center. The circa-1800 rug leads the eye through a sequence. Both belong in a conversation about western Iranian garden carpets, but neither photograph proves that one was copied directly from the other. Their museum dates are estimates and their geographic attributions are broader than a documented workshop address.

What can a surviving carpet tell us about a lost garden?

Scholarship on historical gardens around Bijar and Sanandaj brings another kind of evidence into view. Regional plans and accounts describe water channels, divided planting and pavilions. Set beside woven examples, they raise the question of whether a local taste for ordered landscapes crossed between outdoor architecture and carpet design.

That is a productive comparison, but it needs careful language. A garden changes with pruning, seasons, repairs and later building. A carpet can travel far from its place of weaving. Some western Iranian garden carpets survive only in museum collections or older photographs, while some historic gardens survive in altered form or through plans. Similarity between two layouts is evidence of a shared visual idea, not an invoice linking a named patron, gardener and weaver.

Local patrons may have supported both garden building and garden-carpet commissions. That remains a historical interpretation rather than established provenance for the Met carpets. What we can see directly is narrower and stronger: the visual grammar of enclosing borders, watercourses, planted plots and crossing points. It is enough to make the relationship worth studying without claiming a continuous, unbroken line from an ancient garden to a specific surviving rug.

Nor should the design be flattened into a single religious explanation. Water, shade and planted order can invite ideas of pleasure, protection or paradise in Persian and Islamic culture. A particular rug's maker may also have been thinking about a known garden, a workshop pattern, a patron's taste or simply a compelling composition. No surviving statement gives us every intention behind these two examples. The visible forms should lead the reading; broader meanings can follow as possibilities supported by context.

Earlier garden carpets were not all alike

The western Iranian objects also have a longer history around them. The seventeenth-century Wagner Garden Carpet, attributed on technical grounds to Kirman in southeastern Iran, is full of birds, animals, fish and trees. Yet its channels do not create the perfectly balanced cross often used to explain chaharbagh. Two run lengthwise and meet a horizontal channel that does not reach both outer edges.

That older example is useful here as a caution, even though its loaned photograph is not one of the images reproduced in this article. The later Met carpets are associated with Kurdistan or northwestern Iran and each presents planting differently. They are not merely simplified copies of one “original” garden plan. Each makes choices about how much detail to give to a tree, how to mark flowing water and how to fit a landscape into its field.

A comparison of these objects is therefore about change as much as survival. The seventeenth-century example, the eighteenth-century centered carpet and the circa-1800 repeating carpet belong to a recognizable design family, yet the arrangement shifts. Museum objects preserve those shifts in a way a generic phrase such as “garden pattern” cannot.

How to recognize the idea without overreading it

If you encounter a carpet described as a garden rug, begin by testing the whole layout. Can you follow a watercourse across the field? Do distinct planted areas lie on either side? Does a pool or junction organize the paths? Are trees and flowers placed inside those zones, rather than merely scattered across the surface? These observations are more reliable than finding one tree or a single floral medallion.

Next, look at how the design handles perspective. A channel may be seen from above while a tree appears upright. That apparent inconsistency is part of the carpet's way of explaining a space. Then compare the border with the field: a strong enclosing frame can make the textile feel like a garden within walls, but it does not establish the exact architecture of a real garden.

Finally, keep description separate from attribution. A garden layout does not prove a rug was woven in Kurdistan, Kirman or any other place. Workshop records, structure, materials, provenance and specialist object study matter for origin. The two Met carpets are powerful examples because their photographs can be read alongside researched museum records, not because their motifs alone settle every historical question.

Seen this way, a garden carpet is neither a floor plan in wool nor a vague symbol of flowers. It is a woven proposal for how water, planting and movement might be held together. The two western Iranian survivors let us follow that proposal with our eyes, from one central pool to a long sequence of intersecting channels.

Sources

• Mohammad Ebrahim Zarei, “Reflections of Chaharbagh Garden in the Persian Carpets of the West of Iran with Emphasis on Examples from Kurdistan Province” (2011).
• The Metropolitan Museum of Art, Garden Carpet, object 22.100.128.
• The Metropolitan Museum of Art, Garden Carpet, object 67.156.
• Sheila Canby, “Eternal Springtime: A Persian Garden Carpet from the Burrell Collection” (2018).