Shikhara in Hindu Temple Architecture

Reading parts without inventing a universal anatomy

Many northern towers can be discussed through a lower wall zone, the rising tower mass, a neck and a crown. Projecting offsets known in some scholarship as rathas may begin in the sanctuary plan and continue upward. Curving vertical bands, arch-like gavaksha motifs, miniature shrine forms and subordinate spirelets can articulate the exterior. An amalaka, commonly compared in form to a ribbed disk, crowns many examples; a kalasha or pot-like finial may rise above it. Not every monument has all these parts, and the names are not interchangeable across regions. ^{[3] [17] [28]}

The components also have different evidential status. A stone disk may be ancient, reset, renewed or displaced. A metal finial may have been replaced repeatedly while the tower below retained older courses. Carved miniature shrines can organise surface and mass without containing rooms. Sculpture may be inseparable from architectural composition and sacred meaning, yet carving alone does not reveal the concealed load path. An image of a deity, guardian, couple or river goddess is not texture to be detached from its iconographic and devotional setting. ^{[2] [13] [28]}

Material evidence matters. Some towers use carefully coursed stone; others combine facing, brick, rubble, mortar or plaster. A study of Paraśurāmeśvara shows why infill and imagery have to be read at the scale of a particular building. It would be unsafe to turn that finding into a general statement that all curving shikharas conceal the same core. Joints, bonded projections, cavities, course changes and repair scars must be recorded before a structural account is offered. ^{[13] [38]}

Early evidence and missing upper works

The history of shikhara cannot be narrated as a neat climb from low to tall. Early structural shrines frequently survive without complete upper works. Dashavatara Temple at Deogarh is an important early masonry sanctuary: its high platform, square shrine, subsidiary-shrine evidence and Vishnu sculptural programme make it central to accounts of temple development. Its tower survives imperfectly, however, so claims that it presents the first complete shikhara turn a fragmentary record into unwarranted certainty. ^[4]

Dashavatara Temple, Deogarh: the surviving sanctuary wall and incomplete upper structure, April 2015. Work2win, Dashavatara Temple, Deogarh: surviving upper works; CC BY-SA 4.0. JPEG prepared; no crop or retouch.

At Aihole, Badami and Pattadakal, rock-cut and built temples show builders testing plans, roof forms and architectural languages across the seventh and eighth centuries. Malegitti Shivalaya is especially useful because its Deccan setting and combination of features resist a tidy two-part map. Pattadakal contains both Hindu and Jain monuments and preserves evidence of patronage, including projects associated with royal women. Formal exchange here is historical activity, not a vague hybrid category. ^{[18] [19] [20]}

Central Indian evidence is equally varied. Work on the Lakshmana Temple at Sirpur and the idea of varata traces ingredients that later classifications might separate, while also warning that published plans can incorporate restoration. Kalachuri-period surveys document conventional rectilinear temples, circular examples and structures whose pinnacles are missing. Ashapuri’s dispersed fragments permit reconstruction because their relations can be studied, but every inferred junction should remain distinguishable from surviving material. ^{[14] [16] [17]}

Khajuraho: a group, not a single silhouette

The Khajuraho Group of Monuments is associated principally with Chandella patronage and renowned for the close integration of architecture and sculpture. Lakshmana and Kandariya Mahadeva help explain how halls, sanctuary, platform, main tower and subordinate forms create a coordinated whole. Yet “the Khajuraho shikhara” is too blunt a phrase: the group contains different dates, scales, dedications, conditions and compositional choices. ^{[3] [5]}

The main mass should also be read with its sacred programme. Exterior images, guardians and other figures occupy selected architectural zones; the sanctuary anchors worship and meaning inside. Treating sculpture as surface embellishment while treating the tower as pure engineering would split evidence that the makers and patrons brought together. Conversely, interpreting every ascending form as a cosmic mountain without a monument- and tradition-specific source replaces research with a familiar metaphor. ^{[1] [2]}

UNESCO’s property record and supporting documents add another layer: mapped components, boundaries, integrity, conservation and management. Periodic reporting captures condition and risk at a stated date. It does not make every observation current forever. A visitor should distinguish the monument’s historical phases from modern drainage, paths, repairs, barriers or temporary works, and should consult the responsible authority for present access. ^{[6] [7] [36]}

Odisha: tower, temple city and regional language

Bhubaneswar offers a dense record of Kalinga architecture rather than one isolated example. The Ekamra Kshetra dossier relates Lingaraja to water bodies, subsidiary temples, sacred precincts and processional routes. Odisha’s terminology helps a reader see the tower as a sequence of coordinated zones instead of merely a curve topped by a disk. That vocabulary also prevents central Indian type labels from becoming an all-purpose explanation. ^[25]

Parasurameswar Temple, Bhubaneswar: the sanctuary tower behind the low jagamohana roof, September 2013. Itikanta Mohapatra, Parasurameswar Temple, Bhubaneswar; CC BY-SA 3.0. JPEG prepared; no crop or retouch.

Lingaraja remains part of an active religious landscape. Government visitor material describes its principal architectural components, many subsidiary shrines and festival setting, but such a page is tourism-facing and time-sensitive. Dates, dimensions, entry conditions and current arrangements require corroboration and a date of observation. Heritage value does not override temple authority, and public access is not the same as religious permission. ^[26]

Paraśurāmeśvara and other earlier Odisha temples support close study of how masonry, infill, wall articulation, sacred imagery and rising form interact. Their evidence should not be collapsed into Lingaraja’s later scale. A regional sequence is made from many buildings, repairs, losses and inscriptions, not from arranging present-day silhouettes by height. ^{[13] [25]}

Konark makes the problem of absence unavoidable. The thirteenth-century Sun Temple complex retains its immense audience hall, sculptural programme and other fabric, but the principal sanctuary superstructure is lost. The surviving hall cannot simply stand in for that vanished tower. A British Museum lithograph made in 1848 records a fragment that then remained; it is valuable precisely as a dated, mediated view, not as a measured blueprint for a complete reconstruction. ^{[23] [24]}

Konark Sun Temple: the surviving Jagamohana, with scaffolding visible in March 2018. Sdbhadra, Konark Sun Temple: surviving Jagamohana; CC BY-SA 4.0. JPEG prepared; no crop or retouch.

Western and Bhumija cases

At Modhera in Gujarat, the temple, halls and stepped tank establish a rich spatial sequence while the sanctuary tower is no longer complete. Plan, foundations, wall fabric, scars and historical records therefore carry more weight than a modern imagined outline. The IGNCA account offers detailed monument documentation, while the Archaeological Survey of India supplies the protected-site and custodial frame. Neither licenses a seamless restoration of what is missing. ^{[21] [22]}

Ambarnath, Udayeshvara and other Bhumija monuments answer a different question: how a tower’s miniature forms are ordered. The upper-Deccan survey describes Ambarnath’s distinctive superstructure, sanctuary and hall, and records an eleventh-century consecration. Specialist analysis relates Udayeshvara to royal patronage and to textual discussions associated with Bhoja’s world. These examples show that geometry, patronage, texts and workshop knowledge belong in one history, while no single channel mechanically determines the built result. ^{[11] [27] [28]}

Udayeshvara Temple, Udayapur, Madhya Pradesh: rows of miniature tower forms on the sanctuary tower, March 2012. Deepesh9929, Udayeshvara Temple, Udayapur; CC BY-SA 3.0. JPEG prepared; no crop or retouch.

Stone carving can make the grid legible, but the grid is more than ornament. Repeated aedicules participate in the subdivision of the tower mass; vertical spines and horizontal rows produce an ordered field. A responsible diagram can abstract that organisational principle, but it should not copy a monument’s proportions, sacred figures or survey geometry. ^{[10] [28]}

Beyond a north–south binary

Kashmiri temple analysis demonstrates the value of a regional corpus: plans, roofed forms, aedicular elements and textual comparisons produce relationships that differ from central Indian expectations. Nepal’s Kathmandu Valley adds living urban ensembles, Buddhist and Hindu histories, distinctive local building traditions, earthquake damage and continuing restoration. Calling every upward sacred structure a shikhara would conceal more than it reveals. ^{[15] [37]}

The Deccan likewise records sustained exchange rather than an empty middle zone. Aihole, Badami and Pattadakal contain experiments that later labels divide between northern and southern families. Bhumija monuments extend across other central and western networks. Southern India, meanwhile, preserves vimana and component vocabularies that must be learned on their own terms. The accurate map is plural and historical. ^{[18] [19] [29]}

Later temples may select, combine or revive earlier forms. A contemporary tower can cite a medieval profile without sharing its masonry, workshop system, patronage or ritual history. The visual reference is real, but it does not erase the later date. The same principle applies to diaspora buildings: naming a source tradition requires evidence from designers, patrons, inscriptions or custodians rather than inference from outline alone. ^{[12] [32]}

Sacred images, patronage and makers

The tower is related to a consecrated interior. In Hindu worship, darshan concerns an encounter between deity and devotee; the architectural approach, thresholds and visibility participate in that experience. No diagram of an empty chamber can represent the full religious life of a sanctuary. The blank sanctuary in the accompanying illustration is an ethical boundary, not a claim that real sanctums are empty. ^{[1] [2]}

Sculptures placed on the exterior can identify deities, narratives, directions, patrons or protective roles, but only iconographic and site evidence justifies a particular reading. The page therefore does not invent a guardian, river goddess, couple, shrine image or inscription. It also avoids turning sacred figures into a decorative pattern used to explain tower geometry. ^{[2] [4]}

Lingaraja Temple, Bhubaneswar: miniature architectural forms and a projecting lion in the stone carving, January 2015. Mike Prince, www.flickr.com/people/70804987@N00, Lingaraja Temple: carved tower detail; CC BY 2.0. JPEG prepared; no crop or retouch.

Buildings have makers and commissioners. Inscriptions can record a patron, dedication, date, temple name, artisan or later work; they require responsible transcription and translation. Pattadakal, Khajuraho, Chola monuments and Udayeshvara each preserve different kinds of patronage evidence. Workshop practice appears through setting-out, carving, jointing, models, drawings and the choices visible in the finished fabric. “Anonymous craftsmen following tradition” is not an adequate explanation when named or material evidence exists. ^{[11] [12] [20] [30]}

Damage, rebuilding and conservation

A complete-looking shikhara may contain several time layers. Crowns are especially exposed: an amalaka can be reset, a finial renewed, and upper masonry rebuilt after lightning or water damage. Surfaces may be replastered or repainted; vegetation can open joints; salts can disrupt porous material; and incompatible repair can change both appearance and moisture movement. Diagnosis requires survey and monitoring, not a dramatic photograph. ^{[33] [34]}

The Markandeshwar Temple conservation record is a particularly clear time sequence. It documents lightning damage, partial tower collapse, earlier repair and an early photograph used in the modern assessment before Archaeological Survey of India intervention. The record shows why “original” and “restored” are insufficient as two simple categories: surviving older work, historic rebuilding, modern conservation and missing material may coexist. Its findings belong to that monument and date. ^[33]

Markanda temple group, Maharashtra, photographed in September 2008, before the ASI campaign begun in 2017. सुबोध कुलकर्णी, Markanda temple group, Maharashtra; CC BY-SA 4.0. JPEG prepared; no crop or retouch.

At a living temple, repair is also negotiated around worship, festivals, custodial authority and safe access. Scaffolding tells an observer that work or inspection may be occurring; it does not disclose the diagnosis or completion date. A national statement about annual conservation programmes cannot establish what happened to one tower. Current claims should identify the responsible body and date of checking. ^{[26] [34]}

Historic photographs and drawings are indispensable but imperfect. The Smithsonian finding aid maps photographic campaigns and captions across South Asian architecture; such archives can record features since lost. Yet viewpoints omit concealed sides, captions can inherit earlier mistakes, and a photograph cannot expose internal fabric. Comparing dated images with measured surveys, archaeology and the standing monument is stronger than trusting any one representation. ^{[24] [35]}

Reading the diagram

Original architectural-reading diagrams, not a measured temple and not to scale. They identify a sanctuary tower through plan position, distinguish optional and regional components, compare non-exhaustive organisations, and keep loss and repair visible. No deity, sacred symbol, inscription, copied carving, named monument geometry or third-party image is reproduced.

The illustration is an invented comparison, not a survey of a named temple. Its first panel places a blank sanctuary, possible circumambulatory zone, projecting wall offsets, attached hall and entrance in plan, then aligns an abstract tower above the sanctuary in elevation. It shows why a gateway or hall cannot be renamed from a distant view.

The second panel separates an invented wall or base zone, tower mass, vertical band, optional miniature spire, neck, optional amalaka and finial. Every optional label is marked as regional and variable. No arrow claims to calculate loads or thrust.

The third panel compares four organisations: one curving mass, a clustered mass, a grid of miniature forms and an eastern profile requiring regional vocabulary. These are non-chronological reading aids, not exhaustive species. The final panel marks surviving core, missing crown, later surface, reconstructed edge, documented repair, uncertainty and worship continuing beneath. Pattern, outline and dash repeat the colour distinctions.

Shikhara: Five Evidence Tests

Combine plan position, regional terminology, vertical organisation, fabric and building history before naming or classifying a sanctuary tower.