Kazi Khaleed Ashraf


Land, Water, and People in Bengal: Themes for a Deltaic Architecture

January, 1997

Published in Contemporary Architecture and City Form: The South Asian Paradigm, edited by Farooq Ameen (MARG Publications, Bombay, 1997).


Rice and Identity

Comprising of Bangladesh and most of West Bengal, the Bengal delta is crucially located between two geo-cultural matrices. On the one hand, it has always been considered part of a western orbit formed by the larger Indian culture, but one that extends towards Persia, Arabia, and ultimately Europe. This aspect is seen in most explicit levels of culture: language, liturgies, institutions, and the laws. On the other hand, at a more primordial or foundational level, there is an irreducible stratum that is linked to an eastern matrix, that of South-East Asia.1


Predominantly a water-based civilization, this eastern matrix is a world of moistness, fecundity, and lushness where cosmological and valorized concepts are generated from riverine dynamics and agricultural life patterns. This matrix is also one of “rice culture,” where rice is not something merely consumed but forms the basis of value-construction of a collective ethos and mythos, and of the articulation of self-identity. In the Bengal delta, rice cultivation is an existential occupation; the production of rice is “the production of a world-view.”


The Bengal delta is a pure chemistry of land and water; it is in fact land on lease from water. Formed by the silt, and then constantly shaped and reshaped by the rivers that themselves are perpetually shifting and changing, the land has a certain amorphous character. Advaita Malla Burman, in his novel A River Called Titash, captures the human equation of this land-water environment in two paradigmatic figures: the fisherman and the peasant, the man with the net (jaal) and the man with the plough (langol). Besides the life-sustaining occupation around rice and fish, the Bengal delta is characterized by the life-enhancing occupation around bamboo-weaving and clay-moulding, that is, basketry and pottery. The art of weaving may be seen in the objects of daily life, and in the fabrication of bamboo mats and other architectural elements. Clay kneading has led to the art of pottery, and to the magnificent product ion of terracotta art and brick construction. The fundamental fact is that, from wet-rice production to material fabrication and to cultivating dwelling, living in the delta still means a deep tie to the land and the climate.


Form Reveals Climate

“The Mother Tongue of Architecture” is a rather tantalizing title of an essay by Rabiul Husain, an architect who is more known as a poet. Concerned with the rehabilitation of “localness” in architecture, Husain invites an analogy between language and architecture: Is there, as in language, a “mother tongue” in architecture, an irreducible substratum that defines identity? If there is, where is that substratum? And what is the relationship between language as mother tongue and “mother tongue” in architecture?


Citing the current atmosphere of Euro-American-centric globalization and the loss of local cultures, Husain wonders if it is possible to recover the “mother tongue” of architecture if we have yielded, for example, to an English-centrism in language. Quatremere de Quincy, the eighteenth-century French theorist, would say that architecture and language originate from the same cultural junction. In Bangladesh specifically, the language over which people have laid down their lives (a nationalist symbolization) is however preceded by an architecture that is phenomenally more rooted to nature. There is possibly a rupture between architecture and language; while architecture still remains bound to place, site, and situation, that is, a matrix of nature-climate, language has assumed the contrary character of portability. Bengali is spoken with the same ease in Begumganj as in Bricklane and Brooklyn.


Climate then is more fundamental than language. It may not be simply that “form follows climate,” as Charles Correa suggests, but more significantly climate is the basis of how we see ourselves and how we see the world. The old Japanese notion of fudo offers an illuminating nexus of climate and architecture. Fudo, which literally means wind and earth, conceptually suggests the intertwining of the idea of climate (nature) and culture, a conjoined sense of the two. The singular notion of fudo challenges the Indo-European polarized sense of culture and nature, of songskriti and prokriti, the basis of the modern world-view.


Like fudo, the phenomenological concept of “lived-body” also questions the conceptual separation, of how the environment is distinguished as standing over and against the living body, while it actually may be considered as an indefinite extension of the “lived-body.” There is, as the French philosopher Maurice Merleau-Ponty sees it, a “reciprocal insertion and intertwining” of the “lived-body” and the environment. In living, the body not only lives itself but also the environment. The boundary of one overlaps with the other.


The Japanese thinker Tetsuro Watsuji, who has developed the notion of fudo in contemporary thoughts, qualifies the generalized terms space and environment by characterizing the environment primarily as climate. Thus the “intertwining” becomes specific. The phenomenological notion of how “we discover ourselves in space” is rephrased as how “we discover ourselves in climate,” that is, we find ourselves in a concrete climatic and geographic envelope. In this sense, nature-climate is the foundational aspect of culture. “Climate is the agent by which human life is objectivized, and it is here that man comprehends himself: there is self-discovery in climate.”6


Architecture may then be seen not merely as a shelter from climate, as if climate and architecture are confronting each other. Fundamentally speaking, architecture is the “presencing” of the intertwining of climate and culture; it is where climate is revealed. As Martin Heidegger was to say about the Greek temple, it portrays nothing, it simply is. By standing there on the rocky ground, the temple shows the ground’s spontaneous support, makes manifest the storm raging above, brings to radiance the light of the day and breadth of the sky, and makes visible the invisible space of the air.1


The Pavilion Paradigm

The most pervasive architectural presence in the delta is the pavilion structure: the rustic Bengali hut is a primal pavilion, a machine for living in a hot-humid milieu, so to speak. The hut is essentially a roof characterized by the uniquely swooping canopy or chhad that makes “visible” the presence of the intense sun and the torrential rain (chhad is cognate with chhatri, parasol, and chhaya, shadow). What holds the roof up is secondary; whether mud walls or bamboo posts, they are always placed well within the perimeter of the roof. But it is important that the enclosure be permeable, as in the ideal Laugerian structure on columns. The seventeenth-century Bundi painting Bangala Ragini is an illuminating depiction of the deltaic pavilion. Historical evidence of the idealization of the hut in Bengal – as a tiered shrine serving cosmological purposes – may be seen in representations in the eighth century Ashrafpur Bronze Caitya, and in the eleventh-century Buddhist manuscript Ashtasahasrika Prajnaparamita.9


Noting how the desert has shaped the Arab house and cosmology, the Egyptian architect Hassan Fathy writes: “Because the desert during the day is burning, glaring and productive of sand-storms, the Arab does not find any comfort in opening his house to nature at ground level. The only kindly aspect of nature for him is the sky, promising him coolness in the night and dwarfing even the expanse of the desert with the starry infinitude of the universe. So the Arab tries to shut out the desert with its suffocating sand and open his house on to the sahn or internal courtyard.” Fathy further cosmologizes the sahn, seeing its walls as props for the dome of the sky.10


Fathy’s glorified description of the Arab courtyard house is also illuminating for the Bengali pavilion house, for the two are typologically opposed. A pata painting from Murshidabad illustrates this well. Where the courtyard house puts up a blank wall to shut off the desert, the pavilion house dematerializes it’s walls, making them permeable thus diffusing the distinction between exterior and interior. And, where the courtyard house opens up its central space to the sky, the pavilion puts up a shadow-giving roof. Unlike the intention of shutting out the elements as in the interiorized courtyard-houses of hot-arid places, and the sealed-off spaces of colder climates, the pavilion is completely with the plein air.


The pavilion is a locus from where one looks outwards horizontally, or, as the Filipino architect Leandro Locsin would imply, one sees sunsets and moonlight. On a more sociological note, Anthony D. King points to the centrifugal nature of the pavilion house in which the movement of activity is “outward, on to the verandah, and further into the compound,” unlike the courtyard house, where it is centripetal, towards the courtyard.11 Terraces and verandahs, by extending this quality, become natural appurtenances of the pavilion.


To be with the plein air thus translates into a continuum of the external/natural world and the architectural space into establishing a visual, spatial, and sensorial interpenetration of “nature” and architecture. It would appear that the concept of fudo finds a suitable elaboration in a deltaic situation. It is for this reason the Bengali dwelling is always idealized with its setting. The territory of the house is much wider than the “room” under the roof. A khanar bachan (popular folk proverb) describes such a dwelling as having “Ducks to the East/Bamboo to the West/Banana to the North/Open to the South,” that is, a pond towards the east of the house, and orchards and gardens around it. This has been the perennial image of the Bengali landscape: modest-scaled detached structures among foliage and ponds.


Manifested diversely, the idea of the pavilion takes many forms. The bamboo-woven hut on stilts and the thatch hut on earth platform (bhita) are straightforward examples in which the plan-form may be rectangular or square. The solid centralized mass, constructed in brick or stone out of social needs and monumental urges, is a more arguable elaboration of the pavilion idea (not all centralized mass in the delta participates in the pavilion). For example, the brick-earth stupa with the symbolic Buddhist chhatri at the pinnacle is a sort of combination of the pavilion emblem and masonry mass in which the latter may aspire towards a pavilion condition in a more phenomenal way by the character of its openings.


From simple clustering to a deltaic city-form, the pavilion is also a key to complex organization. Clusters are formed by grouping pavilion “units” in a series or around an amorphously interiorized space. Deltaic morphology implies the disposition of isolated buildings in a fabric of paddyfields, gardens, orchards, lakes, and ponds. An understanding of deltaic city-form or urbanism lies, not in the dense labyrinthine fabric of cities like Jaipur and Lahore – the “city of figure-ground,” but in city-forms east of the Bengal delta, in the “rice culture” matrix, described by the Japanese architect Fumihiko Maki as the “city of figure-nature” in which the distinction between urban and rural morphology is not so oppositional. In summation, the deltaic pavilion may be seen as an independent form with a centrifugal spatiality creating a continuum with the surrounding. The nature of the pavilion is manifested in the parasol roof, the permeable wall, and the verandahs.


From Bangladar to Bungalows

Occurring between the fourteenth and sixteenth century, the Sultanate rule in Bengal was crucial and productive in engaging Bengali architectural themes, especially in restating the idea of the deltaic pavilion in a new mosque form. This particular architectural phenomenon reflects a conflict between a universalist Islam and a territorial ideal.


Adina Masjid (in Pandua, built 1375) points to the rather monumental failure of the large, interiorized courtyard model in the deltaic environment. In following a greater Islamic ideology, the inspiration for Adina must have been the Great Mosque in Damascus with its high vaulted space and large courtyard. The impressive Damascus model never continued through as neither huge congregation nor interiorized courtyards of a hot-dry condition was the reality in the waterlogged delta.


Sultanate architecture in Bengal, as other cultural aspects, grew from a territorial condition. Constructed in the regional brick and terracotta idiom, the architecture of the place was characterized by detached cubic volumes. The interiorized courtyard was dropped altogether. Many buildings instead “opened” their walls by multiple doors on all sides; in the case of a mosque, all sides had doors except the Qibla wall where niches represented the door. In some mosques, the symbolic dome itself was replaced by the monumentalized bangla roof.14


Sultanate architects learned their lessons from the common hut, and also, as Perween Hasan demonstrates, from contemporary temples that too developed from the hut.15 The transfer of the uniquely curved roof of the bangla hut, from bamboo and thatch to masonry, was a Sultanate contribution. And when the swooping chattri was not possible, a softly curving cornice still evoked the original hut. Most Sultanate mosques, and concurrent Hindu temples, employed this element as the single most important feature. The combination of the centralized cubic form and multiplication of the bangla roof would generate a distinct genre of temples.16 The Mughals, and then the Rajputs, fascinated by the bangla roof, would incorporate it as bangladar in their architectural repertoire.


Extending to the idea of the bungalow and the urban detached house, the deltaic pavilion continued to be a dominant theme in more recent times. In colonial Bengal, the bungalow began by monumentalizing the rural hut, but soon spurred by imperialist agendas took an unabashed neoclassical appearance. Modern architecture, by filtering out the stigma of colonial imageries, provided another twist to the bungalow. While the principles of “tropical architecture” gave a climatic rationale to the bungalow in hot-humid zones, modernist themes provided an abstract and asymmetrical configuration. The bungalow turned from a cuboid geometry into a lyrical composition of planes, the lessons of which could be found in the architecture of Frank Lloyd Wright.


The Parasol and the Jali Wall

In Le Corbusier’s architecture for India, the parasol was a central theme although conceived for hot-arid Ahmedabad and Chandigarh. The parasol is the dominant roof on the High Court, the entrance canopy for the Assembly, the symbolic chhatri in front of the governor’s chamber in the Secretariat, and a surrogate-dome on the unbuilt Governor's Palace and the Mill-Owners’ Building. The Shodhan Villa (1956) is a masterful essay of the parasol concept.


In Bangladesh, the earliest modernist realizations of the parasol idea are Constantin Doxiadis’s Teachers-Students’ Centre (1963) and Muzharul Islam’s NIPA building (1969), both for Dhaka University. Both projects are important for considering pavilion themes in a non-domestic realm. The inwardly-slanted parasol over the Teachers-Students’ Centre is however more gestural compared to the brooding presence of the monumental parasol of NIPA that hovers majestically on free-standing columns casting a deep shadow on the multiple functions tucked behind the ring of columns. In his own house (1969), Muzharul lslam conceived a free-floating concrete plane under which brick walls articulated complex layers of spaces. A more recent version is Bruno Lang Duan’s Buddhist monastery in Dhaka (1994), where a single concrete canopy shelters discrete brick volumes. In more abstract form, Uttam Kumar Saha has conceived a number of residences as three-dimensional compositions of largely horizontal planes.


When it comes to the wall, to be with the plein air, is not so natural as suggested in the elemental condition of the rural hut or the idealization in Bangala Ragini. More durable enclosures are necessitated by practical needs, economic status, or new urban conditions. While bamboo screens establish a degree of balance between the need for permeability and the necessity of enclosure, the paradox is heightened with a mud, brick, or concrete walls: how to make the wall both solid and porous.


The filigree screen at the Tomb of Salim Chisti in Fatehpur Sikri is a classic example of a permeable wall. Antonin Raymond’s dormitory at Pondicherry (1936) and Laurie Baker’s Centre for Development Studies at Trivandrum (1972) are more recent articulations of the jali wall. The jali was an important feature in Muzharul lslam’s Dhaka University Library (1954), Science Laboratories (1960), and other projects. The wooden blind (khorkhori) is another way of making a surface alternately open and closed, as in the Polytechnic projects (1966), designed jointly by Stanley Tigerman and Muzharul lslam. Louis Kahn conceived the double wall, in which an inner private wall of regular openings is preceded by an outer wall of large geometric openings: it would seem that the envelope of a taut geometric solid has been “eroded” by natural elements.


Timmy Aziz in his mother’s house (1993) takes up the issue more squarely but with some ingenuity. He literally perforated the solid brick wall and celebrated those perforations by introducing hollow glazed tiles with wirenet inserts.



Clusters and Courts

The delta has emerged with modernism also. In fact, the rationalist abstraction of modernism, as in Muzharul lslam’s essentialism, has filtered the lessons of the delta to a fundamental core (seen in another way, the delta has put modernism in its place), Deltaic themes are addressed in new building programmes without referring to historical motifs or traditional symbolisms.


In more recent times, themes of the pavilion and how pavilions are to be grouped together appear within a more traditional iconography and planning. The issue of grouping has been inseparable from the pavilion idea. Once the modernist monolith, in which diverse functions are subsumed within a single form, has been disintegrated into entities approaching autonomous pavilions, the next critical issue is how to group these together. Most recent projects evoke the traditional uthaan – the semi-enclosed courtyard in Bengali dwellings formed by isolated volumes where corners of the courtyard sort of “leak” to the outside.


Raziul Ahsan of CAPE, in his SOS Youth Village at Mirpur (1984), arranged clustered but disaggregated volumes, formed of vaults and slanted overhangs around semi-interiorized spaces. Uttam Kumar Saha, in his Children’s ViIIage at Khulna (1986), arranged crisp volumes around a stark courtyard. Diagram Architects (Saif Ul Haque and Jalal Ahmed) in a number of projects have created a rich grid of spaces by a seemingly disarrayed clustering of diverse pavilions jostling each other. Their concern is also about the traditional iconography of buildings, especially with pitched roofs, corbelled windows, and the relationship with water and vegetation.


In organizing his building volumes around an enclosed or semi-enclosed courtyard, Bashirul Haq has evolved a distinct typology: a solid mass at ground level, characterized by a brick structure, of arched openings and wooden shutters, with dark nooks and crannies, and a skeletal structure at the upper level, creating a partially canopied airy terrace from where one has a breath taking view of the endless paddyfields.


Raziul Ahsan has also displayed his concern for another aspect of the delta: building with the land. At the Children’s Village in Bogra (1994), he expanded and elaborated the bhita – the solid earth platform on which the house rises into varieties of land configurations: mounds, embankments, terraces, etc. In certain ways, Ahsan has taken a hesitant attitude towards the buildings themselves, keeping them rather straight-forward, while concentrating on and expanding this liminal condition, the grounding of the house.


Large-scale visions in recent times that would incorporate the form of landwater have so far been schematic. In a hypothetical plan for Dhaka (1987), Muzharul Islam suggested reconfiguring the land-water mesh that grids the city, and exploiting the seasonal exchange of water by land terraces with different levels and different uses17. Earlier, Louis Kahn suggested the creation of “land” by brick roads and arched bridges for a large wetland area north of the Capital Complex in Dhaka (1973). During drier seasons, the area beneath could be used for cultivation, while during the monsoon, the bridges would provide a safe and dry platform for building houses. Observing the nature of the deltaic topography and its basic architectural principle of “dig-and-mound” – earth excavated to create both a pond and a mound upon which rises the house – Kahn condensed the realities of the delta as “an architecture of the land.” In the delta, land could be reshaped to give architectural configurations, and conversely, buildings configured to approximate the form of land.


1. See Paul Wheatley, Nagora and Commondery: Origin of the Southeast Asian Urban Traditions (Chicago: University of Chicago, 1983), for a socio-cultural and urban contiguity in the area lying to the east of India and the south of China; and Sumet Jumsai, Nago: Cultural Origins in Siam and the West Pacific, Singapore: Oxford University Press, 1988, for an architectural commonality in the region between Burma and Indonesia. Curiously, in both studies, the western edge of the matrix stops in Bengal. The matrix actually dips down from Bengal towards Sri Lanka and most of South India.

2. See Emiko Ohnuki-Tierney, Rice as ldentity: Japanese Identities Through Time (Princeton: Princeton University Press, 1993).

3. See Richard Eaton, The Rise of Islam and the Bengal Frontier 1204- 1760, (University of California Press, 1993). Eaton suggests that the transformation of the (eastern) delta from forest- to-cultivation occurs with lslamization. But the cosmology of cultivation is still very much indigenous, for wet-rice production had started in the western Bengal delta by the sixth century BC.

4. See Rabiul Husain, Bangladesher Sthapatyasanskrti, (Dhaka: Sahitya Samabaya, 1987) (in Bengali).

5. See Maurice Merleau-Ponty, Phenomenology of Perception, (London: Routledge & Kegan Paul, 1962).

6. See Tetsuro Watsuji, Climate: A Philosophical Study (a translated version published by the Government of Japan in 1930).

7. See Martin Heidegger, “On the Origin of the Work of Art,” reprinted in Martin Heidegger: Basic Writings, (New York: Harper and Row, 1977).

8. Although the pavilion-type is quite widespread all across India, the deltaic one is distinct from the predominant usage in the north as “pleasure-pavilions” and kiosks for inclement weather. In the north, it is often an idealization, as depicted in so many lyrical scenes in miniature paintings. The deltaic pavilion has a rather “existential” necessity; it operates as a paradigm of dwelling, circulating with consistency from the common hut and rich dwellings to religious and public buildings.

9. See S. K. Saraswati, Architecture of Bengal (Vol.1), (Calcutta: Bharadwaj & Co., 1976, pp. 64-65).

10. See Hassan Fathy, “The Qa’a of the Cairene Arab House, Its Development and Some Usages for Its Design Concepts,” in Colloque International sur I'Histoire du Caire, 1969, Proceedings, (Cairo: Ministry of Culture, Arab Republic of Egypt, 1972), pp. 135-52.

11. See Nicholas Polites, The Architecture of Leandro V. Locsin, (New York: Weatherhill, 1977): and Anthony D. King, The Bungalow: The Production of a Global Culture, (London: Routledge & Kegan Paul, 1984), p. 35.

12. See Rafiuddin Ahmad, “Conflict and Contradictions in Bengali Islam: Problems of Change and Adjustment,” in Katherine P. Ewing, ed., Shariat and Ambiguity in South Asian Islam, (University of California Press, 1988).

13. The Great Mosque of Baghdad is a more likely model for Adina considering the importance of both the city and the mosque in the Islamic domain, although Richard Eaton argues that Adina is directly influenced by Ctesiphon. The common elements are the rectangular perimeter with the main hall and the courtyard, and the tripartite division of the hall with a central nave. See R. Ettinghausen and O. Grabar, The Art and Architecture of Islam 650-1250, (Harmondsworth: Penguin, 1987).

14. See Perween Hasan, “Sultanate Mosque and Continuity in Bengal Architecture,” Muqarnas G, An Annual on Islamic Art and Architecture, Leiden: E. J. Brill, 1989, pp. 58-74, for the Bengali characteristics of Sultanate mosques against a broader continuity of vernacular, historical, and “high” architecture that transcends specific religious expressions.

15. Ibid, for a correspondence between Bengali mosque forms and temples in Pagan.

16. See George Michell, editor, Brick Temples in Bengal: From the Archives of David McCutchion (Princeton: Princeton University Press, 1983).

17. Presented at a workshop on “Architecture and the City” in Dhaka (1987), organized by Chetana Architecture Research Group and sponsored by the Aga Khan Program for Islamic Architecture at Harvard and MIT.