Frontiers of Architectural Research (2017) 6, 96–107

Available online at www.sciencedirect.com

Frontiers of Architectural Research

www.elsevier.com/locate/foar

RESEARCH ARTICLE

Visual perception on the architectural

elements of the built heritage
of a historic temple town: A case study
of Kumbakonam, India
K. Kiruthigan, K. Thirumaran

Department of Architecture, National Institute of Technology, Tiruchirappalli, Tamilnadu 620015, India

Received 1 September 2016; received in revised form 22 October 2016; accepted 26 October 2016

KEYWORDS Abstract
Visual perception; This paper aims to investigate the changes in the virtual perception on the built heritage at the
Built heritage; traditional core settlement of Kumbakonam Town at Tamilnadu and to analyze their implica-
Ordinal regression; tions. Specifically, the major objectives of the study are (1) to identify the architectural
Architectural ele- elements that manifest the built heritage of Kumbakonam Town and (2) to assess the
ments;
contributions of these elements to the changes in the visual perception of the town. To
Historic town
achieve these objectives, this study adopts an empirical model that analyzes the architectural
elements of the buildings in the study area. Direct observations and documentations of 373
buildings are collected to analyze those elements that contribute to the changes in the visual
perception on the built heritage of Kumbakonam Town. An ordinary regression model is used to
examine the characteristics of the built heritage across the chariot processional route of the
town. Several architectural elements, including pilasters, horizontal cornices, arched windows,
and ornamental parapets, improve the image of the town. These empirical findings support the
policy framework that enhances the visual perception of Kumbakonam Town.
& 2016 The Authors. Production and hosting by Elsevier B.V. This is an open access article under
the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

Born out of their deep-rooted heritage, historic towns all

n
Corresponding author.
E-mail address: kiruthiga22@gmail.com (K. Kiruthiga).
Peer review under responsibility of Southeast University.
over the world have evolved into spaces for immortals. Built
heritage is more than stone and mortar; these spaces
represent the lifestyles, cultures, and traditions that are
being passed across generations. Most historic towns exhibit

http://dx.doi.org/10.1016/j.foar.2016.10.002
2095-2635/& 2016 The Authors. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Visual perception on the architectural elements
townscape qualities and a great scope for learning the
planning concepts (Gast, 2007). The physical elements of
these towns, including their streetscapes and skylines, add
to their beauty, their history, and the sense of belonging of
their inhabitants (Lynch, 1960). Over the years, the history
of these towns has become a symbol of their image, which
generally reflects the elements of imperial architecture.
Historical towns have confronted new challenges and
changes over the last few decades. Their monuments,
sacred precincts, processional streets, and public squares
are all endowed with moonstones and have remained in
obscurity. Increasing the number of ubiquitous buildings and
0 500
Color Index
Town Boundary
National Highways
Street Roads
Railway Line
Water body
Figure 1 (a) Tamilnadu State map showing Thanjavur District; (b) Thanjavur District map showing Kumbakonam Town;
(c) Kumbakonam Town map.
97
insensitively modifying the heritage buildings beyond recog-
nition can make these historic towns look identical. Sig-
nificantly changing these towns and their heritage only
decay their transformation (Feilden and Jokilehto, 1998).
The historical town of Kumbakonam is no exception. This
town has colorful gopurams that point to the sky and serve
as a reminder that Kumbakonam was previously considered
a center of medieval South Indian power (Nanda, 1996). The
visual perception of this town is reflected in its architec-
tural elements. Silva (2006) argued that given the limited
visual perception of historic towns, particularly on their
physical elements, special preservation programs must be
0 100
0 2 4
98 K. Kiruthiga, K. Thirumaran

launched to improve the image of these areas. Therefore,
an empirical study on the historic town of Kumbakonam
must be performed to understand the significance of the
architectural elements of their built heritage.
This paper aims to examine the significance of the visual
perception on the built heritage of Kumbakonam Town. The
main objectives of this paper are (1) to identify the
architectural elements that manifest the built heritage of
Kumbakonam Town and (2) to assess the contributions of
these elements to the changes in the visual perception of the
town. Direct observations and documentations of 373 build-
ings are collected to analyze the architectural elements that
may contribute to the visual perception of the town. An
ordinal regression model is adopted to achieve the afore-
mentioned objectives. This study also discusses various
conceptions of visual perception, forecasts its significance,
and suggests improvements in the current policy framework.
This paper is divided into six segments. Section 2 describes
Kumbakonam Town and its heritage settings. Section 3
presents the study area, the conceptual framework, and
the methodology for the empirical study. Section 4 presents
the empirical study, highlights the significance of the ordinal
regression model, and discusses the dependent and indepen-
dent variables. Section 5 presents and discusses the results.
Section 6 concludes the paper.
The urban configuration of the town, such as its temples,
tanks, civic buildings, religious institutions, Brahmin elite,
arts, and crafts, contributes to the urban Tamil culture. The
urban fabric of the town, including its temples, chattrams,
matams, agraharams, and paditorais, has made Kumbako-
nam one of the best surviving ancient Tamil towns. Given its
complex evolution, Kumbakonam Town has achieved an
esteemed position among other temple towns in India
(Nanda, 1996) (Figure 3).
The town has become a popular pilgrimage center over
the past centuries. The two holy tanks of Kumbakonam,
namely, Mahamaham tank (Figure 4) and Pottramarai, are
recognized as the most sacred and beautiful tanks in India
and have been associated with Hindu mythology (Ayyar and
Jagadisa, 1920). Possibly the best sacred and ancient
temple town in India, Kumbakonam is one of the few
remaining old settlements in the country where ancient
rituals and traditions remain intact (Ring et al., 1996). The
infrastructure of Kumbakonam Town can be traced back to
its Chola origins over a thousand years ago. The 15–17th
centuries Vijayanagara Nayaka compilations as well as the
18th century Maratha intervention elegantly manifest the

2. Kumbakonam town

2.1. Profile

Located 101570 N latitude and 791230 longitude, Kumbakonam

Town is a river-edge settlement in Tamilnadu, India.
Figure 1 shows a map of the town, its districts, and location.
The town is bounded by two rivers, namely, the Cauvery
River to the north and the Arasalar River to the south.
Kumbakonam Town has a gentle slope from the north to the
south. With more than 188 temples scattered throughout
the area (Conversion of City, 2007), Kumbakonam Town also
serves as a religious center in the region. The town is
famous for its Mahamaham festival, which is celebrated Figure 2 Government Arts College, Kumbakonam.
once every 12 years.
Kumbakonam Town has large-scale agricultural, trading,
and commercial activities. The town has a large silk weaving
segment and is famous for its Thirupuvanam sarees. Since
the ninth century, Kumbakonam Town has been famous for
its crafts, including textiles, sculptures, bronze casting,
brassware, and gold jewelry, which also contribute to its
commercial prosperity (Hunter, 1908). The town has also
become a center of scholarship and was previously called
the “Cambridge of South India” (Nanda, 1996). The bril-
liance of its scholars has been widely acknowledged since
the 20th century. One of the oldest educational institutions
in Tamilnadu established in 1867 is also located in Kumba-
konam Town (Figure 2).

2.2. Heritage and historic settings

Kumbakonam Town occupies a unique position in Hindu

mythology because of its invaluable inscriptions, iconogra-
phy, and religious architecture. The town highly articulates
cosmic geography and idiosyncrasy in its urban structure. Figure 3 Porter Town Hall, Kumbakonam.
Visual perception on the architectural elements
history and heritage of the town (Kathleen, 1981). Cholas,
Pandyas, Krishnadevaraya, and Nayakas ruled Kumbakonam
Town until the British colonization, and these periods have
considerably affected the heritage settings of the town.
2.3. Mahamaham festival
The Mahamaham Festival in Kumbakonam Town is similar to
the Kumbha Mela Festival in North India. This festival
happens once every 12 years in either February or March.
The last festival was celebrated on 22 February 2016
(Figure 5), during which 0.1 million people participated in
the holy bath. The date of the festival is decided based on
certain astrological positions, such as full moons, the
conjunction of the sun in Aquarius with Jupiter, and the
conjunction of the moon in Leo (the Mahamaham star),
which is considered very auspicious (Bansal, 2008). The
origins of this festival can be traced back to the 13th
century. Figure 6 shows one of the earliest photos of the
festival. During Nayaka, the Mahamaham tank was recon-
structed with 16 elegant pavilions and steps descending into
the water (Vriddhagirisan, 1995). The Mahamaham tank is
Figure 4 Mahamaham Tank with AbhiMukheswarar Temple.
Figure 5 Mahamaham Festival on 22nd Feb 2016.
99
currently located at the town center, has 21 wells, and has
an area extending to 6.2 acres. Among these wells, nine
represent the Holy River that flows across India, while the
other wells are named after their associated deities. The
divinity of the tank has been associated with the complex
mythology and creation myth of the town. Hindus strongly
believe that taking a bath in this tank during the Mahama-
ham Festival will absolve them of their sins (Gajarani,
2004).
2.4. Land use
Kumbakonam is a special grade municipal town and the
second largest town in the Thanjavur district in terms of its
administrative status. The town comprises 45 wards and
covers an area of 12.58 sq. km. The robust characteristics of
this town have been regularly associated with the similarity
in its land use or development patterns (Figure 7). Table 1
presents the land use pattern in Kumbakonam Town. The
land use in this area is generally residential, with the
commercial activities primarily located in the town center
and along the highways, Bazaar Street, TSR Big Street,
Nageshwara Koil Street, and Mahamaham tank precincts.
The public and semi-public zones are located in the tradi-
tional core of the town along with temples, tanks, schools,
and the town hall. The government offices and institution
buildings are located to the east, where new buildings are
also being developed.
Architectural style, scale, mass, overall pattern, texture,
building type, use, activity, and inhabitants that can be
seen throughout Kumbakonam reflect the physical charac-
teristics of the town. Most of the organized residential
neighborhoods demonstrate the similar development pat-
terns that can be seen in the traditional core of the town,
including their street lights, lot sizes, building façades, and
building scales. The architectural style, the similarity in
land use patterns, and the arrangement of these character-
istics in terms of their harmony, homogeneity, and order
altogether improve the image of Kumbakonam Town.
100 K. Kiruthiga, K. Thirumaran

3. Methods and data
3.1. Study area
The heritage areas of Kumbakonam Town comprise temples
and sacred temple tanks. The temples include the Adi
Kumbeswarar Temple, Sarangapani Temple, Someshwara
Temple, Nageswaran Temple, Kasi Vishwanath Temple, and
Chakrapani Temple, while the sacred tanks include the
Porthamarai Theertham and Mahamaham tanks (with the
latter being famous for its Mahamaham Festival held every
12 years). Nanda (1996) showed that the sacred and royal
configuration of Kumbakonam, as reflected in its royal
temples and holy tanks, has made this town one of the
earliest settlements of the Chola kings.
The study area (Figure 8) connects two sacred tanks
(Mahamaham tank and Porthamarai Theertham) and acts as
the major commercial corridor of the town. This stretch
also serves as a processional route where chariots travel
through to reach prominent temples. The major develop-
ments and land use patterns in Kumbakonam have been
observed along this stretch.

3.2. Conceptual framework

The rulers of Kumbakonam have conceived, or at least

Figure 6 Thousands of people take the holy dip at the
Mahamaham tank at Kumbakonam in Tamil Nadu, India
during the Mahamaham Festival, known as the Kumbamela of
South India. Mahamaham is performed once in 12 years
(source: Wiele and Klein, Madras).
developed, an architectural paradigm that visually expresses
their ambitions of creating an extensive and powerful state.
Their structures are unprecedented in terms of scale and
architectural clarity; the larger examples may be character-
ized as truly imperial in style (Champakalakshmi, 1979).
Therefore, the image of Kumbakonam and its elements
reflects the character of the entire town. These elements
have been undergoing a rapid transitional and disappearing
process since the past few decades. Figure 9 shows the
variations in the level of visual perception on the architec-
tural elements of the built heritage (BHVP ) in each street.
Such continuous change in the townscape is inevitable.
Figure 10 shows the conceptual framework and the
identified architectural elements. The visual perception of

Legend
Residential
Commercial
Educational
Industrial
Public and semi public 0 2 4
Water body

Figure 7 Land use in Kumbakonam town.

Visual perception on the architectural elements 101

towns is subjected to various components. This study 3.3. Methodology

proposes a conceptual framework by defining the visual
perception level on the architectural elements of the built
heritage at the traditional processional route of
Kumbakonam Town.
This study reviews how some architectural elements,
such as entrance gates or doors, lean-to rooves, horizontal
cornices, pilasters, arched windows, ornamental parapets,
and pot tiled roofing, influence the visual perception on
the built heritage of Kumbakonam Town (Figure 11).
The architectural elements of the buildings in all 12
selected streets are documented.
A total of 373 buildings were documented from the study
area, which comprises 12 streets. Each building was photo-
graphed for reference. The architectural elements of these
elements were identified and considered as primary sources
of information. The identified architectural elements were
used as variables, and the study was followed with an
empirical analysis. To check if the variables were related,
the null hypothesis and the assumed relationship between
the independent (X1–X7) and dependent (BH VP ) variables
were rejected. The analysis results describe the contribu-
tion of each variable to the effects of the dependent
variable. The empirical analysis is presented conceptually
in Figure 12.

Table 1 Land use in Kumbakonam Town.

4. Empirical analysis
S.no Land use type Area (in Ha) % of total

1 Residential 403.68 32.09 4.1. Variables

2 Commercial 34.61 2.75
3 Industrial 15.27 1.21 Seven variables were used to describe the contributions of
4 Public and semi public 81.48 6.48 the architectural elements to the visual perception of
5 Roads 125.44 9.97 Kumbakonam Town. The direct observations and documenta-
6 Vacant 17.18 1.37 tions of 373 buildings in 12 selected streets were used in the
7 Miscellaneous 17.75 1.41 analysis. Both dependent and independent variables were
8 Non-Urban areas 562.61 44.72 used to analyze the visual perception of Kumbakonam Town.
Total 1258.00 100

Legend Pie chart represents the Land use in Study Area

Residential 7% 5% 10%
Mixed Residential
Commercial Mixed
Institutional 78% Commercial
Temple Institutional
Water body

1 Potramarai North Street 7 Nageshwaran East Street

2 Potramarai South Street 8 Post Office Road
3 Potramarai East street 9 Mahamahakulam North street
4 Potramarai West Street 10 Mahamahakulam West Street
5 Thanjavur Main Road 11 Mahamahakulam East Road 0 200 400
6 Nageshwaran North Street 12 Mahamahakulam South Street m

Figure 8 Land use of the study area in Kumbakonam Town along with the streets name.
102 K. Kiruthiga, K. Thirumaran

4.2. Dependent variable where BH VP denotes the level of visual perception on the
built heritage in the processional route of Kumbakonam
BH VP depicts the visual perception level on the built Town, BHeritage denotes the number of buildings with a
heritage of Kumbakonam Town. This dependent variable heritage value located along this route, and BTotal denotes
can be calculated as follows: the total number of buildings along this route.
The dependent variable was expressed as an ordinal
BHeritage  100
BH VP ¼ ; ð1Þ number to account for the standard deviation of the mean
BTotal of all observations. The selected streets were classified as
follows according to their visual perception level (Figure 13):
Low
VISUAL PERCEPTION ON BUILT HERITAGE (BH VP)

Medium High
90
 Category 1: Streets with low visual perception.
80
 Category 2: Streets with medium visual perception.
70  Category 3: Streets with high visual perception.
60

50 4.3. Independent variables used in the regression

40
model
30

20 Entrance Gate or Door (X1): A decorated entrance, gate,

10 or door that marks the transition from the public exterior
to the private interior space (Ernest, 2008).
0
1 2 3 4 5 6 7 8 9 10 11 12 Lean-To Roof (X2): A hipped roof with a gable end and
STREETS IN THE SELECTED STUDY AREA lean-to projection. This roof is added to an existing
building to address a new need (Haneman, 1984).
Figure 9 Graph showing the level of visual perception of the
Horizontal Cornices (X3): A molded or decorated pro-
study area of Kumbakonam Town.
jection that forms the crowning feature of the top of a

KUMBAKONAM

ENTRANCE
GATE or DOOR

VISUAL PERCEPTION
ON BUILT HERITAGE LEAN TO ROOF

CORNICES
(horizontal)

ARCHITECTURAL

ELEMENTS PILASTERS

ARCHED
WINDOWS

ORNAMENTAL
PARAPETS

POT TILES
ROOFING

Figure 10 Conceptual framework of visual perception of Kumbakonam Town on its Built Heritage.
Visual perception on the architectural elements 103

Figure 11 Building images with architectural elements. (a) and (b) at Mahamahakulam South Street; (c) and (d) at Mahamaha-
kulam West Street; (e) and (f) at Mahamahakulam East Street; (g) Pottramarai North Street.

building wall or other architectural element by extending
any similar crowning and projecting element. Cornices
also refer to any projection on a wall that throws
rainwater off the face of a building.
Pilasters (X4): A partial pier or column, often with a base
and capital, that is slightly projected and embedded on a
flat wall (Ernest, 2008).
Arched Windows (X5): The bottom half of an “arched” or
“radius” window is rectangular, while its top is an arch or
a half circle. Arched windows bring a sense of softness to
a room and juxtapose the straight corners of rectangular
windows and walls for a contrasting effect.
Ornamented Parapets (X6): A low protective wall or
railing that is placed as an architectural element above
104 K. Kiruthiga, K. Thirumaran

DATA COLLECTION

IDENTIFYING THE ELEMENTS

ASSIGNING THE VARIABLES

EMPIRICAL MODEL SELECTION

Rejecting the Null

HYPOTHESIS TESTING
Hypothesis

If p<5% then significant

CHI SQUARE TEST association, If p>5% then
no significant association.

INTERPRETE THE RESULTS

Figure 12 Schematic flow chart of the empirical analysis.

Legend
Building with High Heritage value
Building with Medium Heritage value
Building with Low Heritage value

1 Potramarai North Street 7 Nageshwaran East Street

2 Potramarai South Street 8 Post Office Road
3 Potramarai East street 9 Mahamahakulam North street
4 Potramarai West Street 10 Mahamahakulam West Street
5 Thanjavur Main Road 11 Mahamahakulam East Road
6 Nageshwaran North Street 12 Mahamahakulam South Street
0 200 400
m

Figure 13 Level of visual perception in the study area of Kumbakonam Town.

Visual perception on the architectural elements
cornices and along the edge of a raised platform,
terrace, bridge, roof, or balcony. These parapets are
often decorated and incorporated into the existing
design (Ernest, 2008).
Pot tiles roofing (X7): The external covering on the top
of the building that usually comprises wooden rafters,
pot tiles, or countryside tiles placed on pitched slopes or
vertical built-up membranes for flat roofs.
4.4. Selection of empirical model
The relationship between the architectural elements and
visual perception of Kumbakonam Town is assessed using an
ordinal regression model in SPSS. An ordinal regression is
performed to predict the dependent variable with
“ordered” multiple categories and independent variables.
The regression facilitates the interaction of dependent
variables (with multiple ordered levels) with one or more
independent variables (Norusis, 2004). Previous studies have
examined how the sense of a place can be judged based on
its location. For instance, Boereback (2012) assessed the
relationship between shopping locations and sense of place
in the Netherlands, while Minetos and Polyzos (2010)
discussed some applications related to change in land use
and urban sprawl. Both of these studies applied ordinal and
multinomial regression models.
Ordinal logistic regression can forecast or reveal how the
changes in the independent variables can affect the dependent
variable (Frank and Harrell, 2001). This approach has been used
to analyze visual perception levels (low, medium, or high) when
these levels cannot be represented by a continuous variable or
proxy. The ordinal logistic regression model can be constructed
using five types of link functions based on the acceleration of
the cumulative probability of the response variable (Norusis,
2004). Given that the dependent

variable

 in this study has an
equal category, the Logit link link γ j is applied.
The general ordinal regression models in the Logit link
function can be expressed as follows:
! considered important because the change in likelihood demon-

 γj
link γ j ¼ log ð2Þ
1γ j
and
!   variables. As illustrated by the pseudo R-square values in Cox
γj PðY i rjÞ
log ¼ log ¼ αj ðβ1 X 1i þ:::þ βk X ki Þ;
1γ j PðY i 4jÞ
 
γ
where log 1 j γ is the Logit link, the link index j denotes
j rejects the null hypothesis H0, thereby indicating that the
the visual perception level (low, medium, or high), γ j
denotes the cumulative response probability up to and
including Y = j, Yi denotes the response variable for observa-
tions i= 1…n, k denotes the predictor variable associated
with the observed changes in the dependent variable, α is
the intercept of the regression equation or threshold for
each cumulative probability, and βk is the coefficient of the
predictor variables or locations of the model. The threshold
α and the regression coefficient βk are unknown parameters
to be estimated using the maximum likelihood algorithm.
105
Table 2 Test of parallel lines.a
Model  2 log Chi- df Sig.
likelihood square
Null hypothesis 21.758 – – –
General 11.893b 9.865c 7 0.000
The null hypothesis states that the location parameters
(slope coefficients) are the same across response categories.
a
Link function: Logit.
b
The log-likelihood value cannot be further increased after
maximum number of step-halving.
c
The Chi-Square statistic is computed based on the log-
likelihood value of the last iteration of the general model.
Validity of the test is uncertain.
4.5. Model fitting information
In statistics, the parallel lines assumption must be observed
when using the ordinal logistic regression model. This
assumption posits that the regression coefficients are the
same for all corresponding outcome categories. If the parallel
lines test fails, then the multinomial regression model can be
used as an alternative model. Table 2 presents the parallel
lines test in which the assumptions of the regression coeffi-
cients are the same for all the outcome categories. The chi-
square in this model is statistically significant (0.2540.05).
Therefore, the assumption of parallelism cannot be rejected.
We maintain the null hypothesis that the local parameters
are the same across different response categories.
Table 3 shows the Pearson's correlation coefficients and
goodness-of-fit indices. A large number of empty cells was
obtained because several continuous dependent variables were
used (the output of the model warned that 24 or 66.7% of the
cells had zero frequencies). Therefore, we should not depend
on these variables. Given the very low significance level in the
overall model test, the intercept-only model does not function
well with the predictors. However, when some cells have small
observed and predicted counts, the overall model test is
strates a chi-square distribution (Norusis, 2004). The pseudo
R-square measures indicate the success of a model by describ-
ing the variations in the data, which in turn signify the strength
of the association between the dependent and independent
and Snell (0.354), Nagelkerke (0.437), and McFadden (0.262),
ð3Þ
the values of pseudo R-square measures are always lower than
the R-square of a linear regression model (Norusis, 2004).
Therefore, the selected model for the regression analysis
model is statistically significant.
The confusion matrix in Table 4 shows the predictive ability of
the empirical model. In the first category of low visual percep-
tion, the model correctly assigned 100% of the cases. In the
second category of medium visual perception, the model
correctly assigned 3 out of 4 or 66.7% of the cases. In the last
category of high visual perception, the model correctly assigned
1 out of 2 or 50% of the cases. Therefore, the confusion matrix
that is obtained by the cross-tabulated rating of the predicted
response category is very high at the first and second categories.
106 K. Kiruthiga, K. Thirumaran

Table 3 Model fitting information.a

Model  2 log Likelihood Chi-square df Sig.

Intercept only 24.274

Final 21.758 2.516 7 0.000

Pseudo R-square Goodness-of-fit

Cox and Snell 0.354 Chi-square df Sig.
Nagelkerke 0.437 Pearson 46.139 30 0.000
McFadden 0.262 Deviance 37.714 30 1.000
a
Link Function: Logit.

Table 4 Confusion matrix.

Level of Visual Perception Predicted response category Total

Low Medium High

Low N 6 0 0 6
% 100%a 0% 0% 100%
Medium N 1 3 0 4
% 33.3% 66.7% 0% 100%
High N 0 1 1 2
% 0% 50% 50% 100%
Total N 7 4 1 12
% 58.3% 33.3% 8.4% 100%
a
Percentage of correctly assigned cases by the model.

Table 5 Parameter estimates.a

Estimate Std.error Wald df Sig. 95% confidence interval

Lower bound Upper bound

Threshold
BH VP [1] 5.952 0.8928 44.445 1 0.015 4.202112 7.701888
BH VP ½2] 8.312 0.9974 69.450 1 0.001 6.3570 10.2669
Location
X1 1.052 0.2419 18.912 1 0.012 0.5778 1.5261
X2 0.621 0.1428 26.915 1 0.001 0.34112 0.9008
X3 0.373 0.1231 9.1812 1 0.022 0.131724 0.6142
X4 0.466 0.1072 4.3470 1 0.036 0.255888 0.676112
X5 0.332 0.1527 2.1741 1 0.061 0.032708 0.631292
X6 0.741 0.1333 5.5588 1 0.029 0.479732 1.00226
X7 0.742 0.1630 4.5460 1 0.038 0.42252 1.06148
a
Link function: Logit.

5. Results and discussion
Table 5 presents the parameter estimates, standard errors,
wald, significance levels, and lower and upper bounds of the
95% confidence interval. The threshold estimate for
[BH VP =1.00] is the cutoff value between the low and medium
BH VP ; while the threshold estimate for [BH VP =2.00] is the
cutoff value between the medium and high BH VP . The estimate
represents the ordered log-odds (logit) regression coefficient.
The ordered logit coefficient indicates that for a one-unit
increase in the independent variable, the dependent variable
level is expected to change by its corresponding regression
coefficient in the ordered log-odds scale when the other
independent variables in the model are kept constant. The
positive signs of the regression coefficient indicate that those
streets with more architectural elements (X1, X2,.., X7) are
more likely to present a high visual perception level. The
significance level of the seven architectural elements, except
Visual perception on the architectural elements
for X5, has a positive regression coefficient and p-value (sig.) of
o0.05. Therefore, these architectural elements greatly affect
the visual perception on the built heritage of Kumbakonam
Town. The case summary of the model (Appendix A) describes
the visual perception levels of the 12 streets in the study area.
The low visual perception level of certain streets may be
attributed to their lack of built heritage or prevalence of
modern construction without any architectural elements.
6. Conclusions
The significance of architectural elements on the heritage of
Kumbakonam Town has been empirically analyzed. The influ-
ence of each architectural element on built heritage is assessed
using an ordinal logistic regression model. These architectural
elements are strongly associated with the visual perception of
the town. The presence and absence of an architectural
element correspond to a significant change in visual perception
level. The low visual perception of streets located in traditional
historic areas indicates that the characteristics of Kumbakonam
are gradually deteriorating. The development of urban infra-
structures, the dismantling of heritage structures, the changing
building use patterns, the adoption of new architecture styles,
and the lack of policy and guidelines for protecting built
heritage create a negative background for built heritage.
Considering its architectural heritage and cultural significance,
Kumbakonam Town requires a proper policy framework for
preserving and conserving its built heritage. Before constructing
such framework, the significance of the architectural elements
of the town must be investigated in detail. The findings of this
study may initiate further research into the architectural
elements of built heritage and improve the policy framework
for preserving the built heritage of Kumbakonam Town.
Appendix A
Case summary
N Marginal Percentage
BH VP Low 6 50%
Medium 4 33.3%
high 2 16.7%
Valid 12 100%
Missing 0
Total 12
The outcome variable in this regression is represented by
BH VP , which contains a numeric code for the level of visual
perception on built heritage. N denotes the number of
observations or streets in the study area, while the marginal
percentage denotes the proportion of valid observations in
each visual perception level (low, medium, or high) of BH VP .
107
References
Ayyar, Jagadisa, P.V., 1920. South Indian Shrines Illustrated. Asian
Educational Services, New Delhi (ISBN 81-206-0151-3).
Bansal, Sunita Pant, 2008. Hindu Pilgrimage: A Journey Through the
Holy Places of Hindus All Over India. Hindology Books, Delhi
(ISBN 978-81-223-0997-3).
Boereback, J.J., 2012. Sense of Place; Attachment to, Identity With
and Dependence of Shopping Locations (thesis). Eindhoven Uni-
versity of Technology, Real Estate Management and Development.
Champakalakshmi, R., 1979. Growth of urban centres in South
India: Kudamukkupalaiyarai, the twin city of the Cholas. Stud.
Hist. 1, 1–31.
Conversion of City, 2007. Corporate Plan to Business Plan for
Kumbakonam municipality-TNUIFSL (PDF) (Report). Commis-
sioner of Municipal Administration, Government of Tamil Nadu.
Retrieved 2012-11-16.
Ernest, Burden, 2008. Elements of Architectural Design: A Photo-
graphic Sourcebook, 2nd Ed.. ISBN: 978-0-471-37117-5.
Feilden, B.M., Jokilehto, J., 1998. Management Guidelines for
World Cultural Heritage Sites. ICCROM, Rome.
Frank, E., Harrell, 2001. Regression Modeling Strategies with
Applications to Linear Models, Logistic and Ordinal Regression,
and Survival Analysis. Springer Series in Statistics (ISBN 978-3-
319-19424-0).
Gajarani, S., 2004. History, Religion and Culture of India, 3. Isha
Books, New Delhi (ISBN 81-8205-061-8).
Gast, Klaus-Peter, 2007. Modern Traditions: Contemporary Architecture
in India. Birkhauser Basel (ISBN 3764382988, 9783764382988).
Gough, Kathleen, 1981. Rural Society in Southeast India. Cambridge
University Press.
Haneman, John Theodore, 1984. Pictorial Encyclopedia of Historic
Architectural Plans, Details and Elements: With 1880 Line
Drawings of Arches, Domes, Doorways, Facades, Gables, Win-
dows, etc. Dover Publications.
Hunter, Sir William Wilson, 1908. The Imperial Gazetteer of India
1908, 16. Clarendon Press.
Lynch, K., 1960. The Image of the City. The MIT Press, Cambridge,
MA.
Minetos, D., Polyzos, S., 2010. Deforestation processes in Greece: a
spatial analysis by using an ordinal regression model. For. Policy
Econ. 12, 457–472.
Nanda, V., 1996. Cambridge Kumbakonam Project: the ritual
topography of Kumbakonam. Arch. Des. 124, 30–34.
Norusis, M., 2004. SPSS 13.0 Advanced Statistical Procedure Com-
panion. Prentice Hall, New Jersey.
Ring, Trudy, Robert, M., Salkin, Boda, Sharon La, 1996. Interna-
tional Dictionary of Historic Places: Asia and Oceania, Volume
5 of International Dictionary of Historic Places. Taylore& Francis
(ISBN 1884964052).
Silva, K.D., 2006. Conserving Asian urban heritage: the need for re-
conceptualizing the value-based management approach. In:
Silapacharanan, Siriwan, Campbell, Jeffrey (Eds.), Asian
Approaches to Conservation. Chulalongkorn University, Bangkok,
Thailand, pp. 16–23.
Vriddhagirisan, V., 1995. Nayaks of Tanjore. Asian Educational
Services, New Delhi (ISBN 81-206-0996-4).