iii A CLUSTER ANALYSIS COMPARISON OF SELECTED TRADITIONAL JAPANESE GARDENS, CLASSICAL CHINESE GARDENS AND MODERN CHINESE GARDENS By Dexin Chen A THESIS Submitted to Michigan State University i n partial fulfillment of the requirements f or the degree of Environmental Design - Master of Arts 2018 iii ABSTRACT A CLUSTER ANALYSIS COMPARISON OF SELECTED TRADITIONAL JAPANESE GARDENS, CLASSICAL CHINESE GARDENS AND MODERN CHINESE GARDENS By Dexin Chen Cluster analysis is a useful mathematical method that has been used to examine the differences or similar ities between gardens (Xu Y. , 2015) . This investigation aim s to test the differences and similarities between traditional Japanese gardens in Kyoto, the clas sical Chinese gardens in Suzhou , and the modern Chinese gardens in Xiamen , by comparing the design elements and design principles of th e gardens. A hundred and thirty - f our variables are selected base d up on a s personal experience in Kyoto, where the first seventy - five variables are adopted from earlier garden research by Yiwen Xu (Xu Y. , 2015) . After collecting the variables and applying the Principal Component Analysis by the software SAS , a group of cor responding eigenvalues are generated. According to the Principal Component Analysis , the first t wo principal components covered 63.81 percent of the sample variance . The first and second principal component together divided the gardens into three groups : the first principal component in dicated similarities between the traditional Japanese gardens and classical Chinese gardens; the second principal component i ndicated similarities between traditional Japane se gardens and modern Chinese gardens. T herefore, t he final result is a group of three two - dimensiona l scatter graphs, where each point represents a coordinate axes, and the dimensions are generated from the meaningful eigenvalues . K eywords: Landscape Architecture, Environmental Design, Historic Gardens, Contemporary Gardens, Garden Design, Aesthetic Principles, Cultural Context, Oriental Gardens . iii ACKNOWLEDGEMENTS I wish to thank my major advisor Dr. Jon Burley, FASLA, Professor, Michigan State University School of Planning, Design and Construction . Dr. Burley has helped me to gain a great deal of knowledge in Landscape Architecture , as well as answer s to some of life questions. I would also thank Dr. Patricia Machemer and Dr. Robert Schutzki , professors at Michigan State University, for serving as committee members and always offering helpful suggestions during my school program. In particular, I want to thank Yiwen Xu and Haoxuan Xu, students from Michigan State University, and Zhi Yue, professor of Nanjing Forestry University , for their patient guidance in the statistical research method. iv TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ .................. vi LIST OF FIGURES ................................ ................................ ................................ ............... vii CHAPTER 1: INTRODUCTION AND LITERATURE REVIEW ................................ .... 1 1.1 Introduction ................................ ................................ ................................ .................... 1 1.2 Literature Review of Classical Chinese Gardens and Modern Chinese Gardens ... 1 1.3 Literature Re view of Traditional Japanese Gardens ................................ ................. 3 1.3.1 A Brief History of Traditional Japanese Gardens ................................ ..................... 3 1.3.1.1 Nara period ................................ ................................ ................................ ........ 5 1.3.1.2 Heian period ................................ ................................ ................................ ...... 7 1.3.1.3 Kamakura period and Muromachi period ................................ ......................... 8 1.3.1.4 Momoyama period ................................ ................................ ........................... 10 1.3.1.5 Edo period & Meiji period ................................ ................................ ............... 10 1.3.2 Cultural Context and Aesthetic principles ................................ .............................. 11 1.3.2.1 Geography ................................ ................................ ................................ ........ 11 1.3.2.2 Zen Buddhism and Tea ................................ ................................ ..................... 13 1.3.2.3 Geomancy Rules ................................ ................................ ............................... 14 1.3.2.4 Imitation of Natural Landscapes ................................ ................................ ..... 16 1.3.3 Traditional Japanese garden styles and functions ................................ ................... 18 1.3.3.1 River - style garden (Yarimizu) ................................ ................................ .......... 18 1.3.3.2 Shinden style Garden (Palatial Architecture) ................................ .................. 19 1.3.3.3 Sho in - Zukuri Style Garden ................................ ................................ .............. 19 1.3.3.4 Zen Garden ................................ ................................ ................................ ...... 20 1.3.3.5 Tea Garden (R oji) ................................ ................................ ............................ 23 1.4 Design Rules and Major Features in Traditional Japanese Gardens ..................... 25 1.4.1 Borrowed scenery (Shakkei) ................................ ................................ ................... 25 1.4.2 Hide and Reveal ................................ ................................ ................................ ...... 26 1.4.3 Rock positions ................................ ................................ ................................ ......... 27 1.4.4 Water Features ( Sakuteiki ) ................................ ................................ ...................... 28 1.4.5 Clipped Bushes ................................ ................................ ................................ ....... 31 1.5 Conclusion ................................ ................................ ................................ .................... 32 CHAPTER 2: METHODOLOGY ................................ ................................ ........................ 35 2.1 Purpose of Study ................................ ................................ ................................ .......... 35 2.2 Study Sites ................................ ................................ ................................ .................... 35 2.2.1 Classical Chinese Gardens in Suzhou and Modern Chinese Gardens .................... 35 2.2.2 Traditional Japanese Gardens in Kyoto ................................ ................................ .. 36 2.2.2.1 Shisen - do ................................ ................................ ................................ .......... 36 2.2.2.2 Byodo - in ................................ ................................ ................................ ........... 38 2.2.2.3 Kiyomizu - dera ................................ ................................ ................................ .. 39 2.2.2.4 Daikaku - ji ................................ ................................ ................................ ........ 42 2.2.2.5 Tenryu - ji ................................ ................................ ................................ ........... 43 2.2.2.6 Ryoan - ji ................................ ................................ ................................ ............ 45 2.2.2.7 Kinkaku - ji/Rokuon - ji (The Golden Pavilion) ................................ ................... 47 2.2.2.8 Daitoku - ji ................................ ................................ ................................ ......... 48 2.2.3 Shrines in Kyoto ................................ ................................ ................................ ..... 50 2.2.3.1 Yasaka Shrine/Yasaka Jinsha ................................ ................................ .......... 50 v 2.2.3.2 Fushimi - Inari Taisha ................................ ................................ ....................... 51 2.3 Data Collection ................................ ................................ ................................ ............. 52 2.4 Analysis Techniques ................................ ................................ ................................ .... 56 CHAPTER 3: RESULTS ................................ ................................ ................................ ...... 60 CHAPTER 4: DISCUSSION ................................ ................................ ................................ 81 4.1 Cluster Analysis Method in Garden Comparison ................................ .................... 81 4.2 Comparison of Gardens and Garden Elements ................................ ........................ 81 4.3 Future Implication ................................ ................................ ................................ ....... 94 4.4 Limitations and Suggestion for Future Research ................................ ..................... 95 BIBLIOGRAPHY ................................ ................................ ................................ .................. 98 vi LIST OF TABLES Table 1: List of design elements and principles included in the eighteen selected sites. ........ 53 Table 2: List of design elements and principles in the three classical Chinese gardens and five modern Chinese gardens. ................................ ................................ ................................ . 60 Table 3: List of design elements and principles in the eight traditional Japanese gardens and two shrines. ................................ ................................ ................................ ...................... 65 Table 4: Means and standard deviation of the corresponding variables from the SAS software program. ................................ ................................ ................................ ........................... 69 Table 5: Principal Component Analysis eigenvalues of the covariance matrix from the SAS Software Program. ................................ ................................ ................................ ........... 73 Table 6: Principal Component Analysis eigenvalue coefficient for each variable from the SAS Software Program. ................................ ................................ ................................ ... 74 Table 7: Site scores in the first three principal components. ................................ ................... 78 Table 8: Corresponding initials of the selected sites as shown in scatter graphs. ................... 79 Table 9: List of th e design elements corresponding to the largest and the smallest coefficients of principal component 1. ................................ ................................ ................................ 82 Table 10: List of the design elements corresponding to th e largest and the smallest coefficients of principal component 2. ................................ ................................ ............. 83 Table 11: List of the design elements corresponding to the largest and the smallest coefficients of principal component 3. ................................ ................................ ............. 83 vii LIST OF FIGURES Figure 1: A garden setting on gravel paving area in Tenryu - ji temple. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ .......................... 4 Figure 2: A void, open gravel space adjacent to the revered hall in Yasaka Shrine. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ .............. 5 Byodo - in temple. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ................................ ................................ ................................ ............ 6 Imperial Palace. (Copyright ©2017 Dexin Chen all right reserved used by permission). 7 right reserved used by permission). ................................ ................................ ................... 9 Figure 6: Dry landscape named Honryutei in Enkou - ji, imitating a dragon flying among the clouds. (Copyright ©2017 Dexin Chen all right reserved used by permission). ............. 11 Figure 7: Naturalness in the pathway to Kasuga - taisha, a Shinto shrine in Nara. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ............ 12 Figure 8: Meditation by sitting in front of the Zen garden in Ryogen - in, a sub - temple of Daitoku - ji. (Copyright ©2017 Dexin Chen all right reserved used by permission). ....... 14 Figure 9: Rock settings and dry landscape in Daitoku - ji temple. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ................................ ... 17 Figure 10: River style garden in Kyoto Imperial Palace. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ................................ ................. 18 Figure 11: Moss featuring Zen garden with sitting area named Ryugintei in Ryogen - in, a sub - temple of Daitoku - ji. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ................................ ................................ ...................... 22 Figure 12: A tea house in a bamboo fenced roji located in Shisen - do. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ........................ 24 Figure 13: The mountain is borrowed as background in this view as Borrowed Scenery in Tenryu - ji temple. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ................................ ................................ ................................ .......... 26 Figure 14: The whole view is blocked by a wall, but a part of it is exposed. Photo taken in Kyoto Imperial Palace. (Copyright ©2017 Dexin Chen all right reserved used by permissio n). ................................ ................................ ................................ ...................... 27 Figure 15: Zen garden in Ryoan - ji temple. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ................................ ................................ ........ 28 viii Figure 16: An island with its reflection in the pond. Photo taken in Kyoto Imperial Palace. (Copyright ©2017 Dexin Chen all right reserved used by permission). ......................... 29 Figure 17: Stone hand wash basin in Tenryu - ji. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ................................ .......................... 31 Figure 18: Meditation in Shisen - do while viewing the dry landscape. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ........................ 37 Fig ure 19: Shishiodoshi in Shisen - do. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ................................ ................................ ................. 38 Figure 20: The phoenix hall of Byodo - in temple. (Copyrigh t ©2017 Dexin Chen all right reserved used by permission). ................................ ................................ .......................... 39 Figure 21: Pagoda and wisteria pergola in Kiyomizu - dera. (Copyright ©2017 Dexin Chen all right reserv ed used by permission). ................................ ................................ ................. 40 Figure 22: The stairs before getting to the gateway of Kiyomizu - dera, creating a sense of entering a sacred place. (Copyright ©2017 Dex in Chen all right reserved used by permission). ................................ ................................ ................................ ...................... 41 Figure 23: A view towards Osawa - no - ike pond. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ................................ .......................... 42 Figure 24: The lifted covered walkway in Daikaku - ji. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ................................ .......................... 43 Figure 25: The main hall facing to the pond of Tenryu - ji. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ................................ ................. 44 Figure 26: The raked sand design is merged into the view as foreground; the rock groupings suggest the pond edge as midground; the mountain behind is borrowed as background. (Copyright ©2017 Dexin Chen all right reserve d used by permission). ......................... 45 Figure 27: One of the fifteen rocks in the Zen garden of Ryoan - ji. The rectangular wall brings a sense of enclosure. (Copyright ©2017 Dex in Chen all right reserved used by permission). ................................ ................................ ................................ ...................... 46 Figure 28: A Torii is put in the pond in Ryoan - ji, suggesting the Shinto influence of this Buddhist temple. (Co pyright ©2017 Dexin Chen all right reserved used by permission). ................................ ................................ ................................ ................................ .......... 47 Figure 29: The golden pavilion and its reflection in Kinkaku - ji temple. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ........................ 48 - in, a sub - temple of Daitoku - ji. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ............ 49 - in, a sub - temple of Daikaku - ji. (Copyright ©2017 Dex in Chen all right reserved used by permission). .. 50 ix Figure 32: The main gate with viewing terrace in the Yasaka shrine. Chinese characters are used in the informat ive plaques. (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ................................ ................................ ................. 51 Figure 33: Thousands of Torii cover the outdoor pathway in Fushimi - Inari Taisha . (Copyright ©2017 Dexin Chen all right reserved used by permission). ................................ ............ 52 Figure 34 : A scatter graph of the correlation between the eighteen selected sites based o n the site scores of Principal Component 1 and Principal Component 2. ................................ . 79 Figure 35: A scatter graph of the correlation between the eighteen selected sites based o n the site scores of Principal Component 1 and Principal Component 3. ................................ . 80 Figure 36: A scatter graph of the correlation between the eighteen selected sites based on the site scores of Principal Component 2 and Principal Component 3. ................................ . 80 1 CHAPTER 1: INTRODUCTION AND LITERATURE REVIEW 1.1 Intr oduction The first impression that many people may have of traditional Japanese gardens is a space which has few flowers, and a sitting area comprised of only sand and rocks , which aim s to represent a peaceful and harmonious landscape in the spirit of Zen Buddhism. The arrangement of the rocks in traditional Japanese gardens has attracted many people to understand the principles of visual balance s in a continued and a limited space (Weiss, 2010) . The context and aesthetic principles of Japanese culture have influenced the evolution of the garden style that resulted in these garden landscape s, which can only be fully underst ood through deep thinking . S pend ing time walking and sitting in these gardens is often the best way to understand the spiritual concepts and the conventional rules that have spanned the evolution of Japanese gardens . Although traditional Japanese gardens and classical Chinese share many similar characteristics , few schola r s have used a scientific method to illustrate the similarities and differences between them. In order to comprehensively examine the similarities and differences of traditional Japanese gardens, classical Chinese gardens, and modern Chinese gardens , a s tatistic al method called Principal Component Analysis will be used . Therefore , this study can assist in developing a broader understandi ng of oriental garden designs . 1.2 Literature Review of Classical Chinese Gardens and Modern Chinese Gardens The classical Chinese gardens can be mainly classified as residential style in southern China and imperial style in northern China (Xu Y. , 2015) . Compar ed to the imperial gardens in Northern China, classical Chinese gardens in Suzhou are typically characterized by a relatively small size of land occupancy (Xu Y. , 2015) . During the Ming and Qing dynasty, Suzhou, a city in southern China, we re renowned because they were home to the leading poets and painters of the ancient era (Xu Y. , 2015) . 2 T he cultural ambience of Suzhou citizens foster ed the practice of classical residential gardens in southern China (Xu Y. , 2015) . Many classical Chinese gardens were destroyed during war s and revolutions ; however, some private residential gardens were well preserved in Suzhou, and UNESCO evaluated these spaces as representative of world cultural heritage (Liu, Burley, & Partin, 2014) . In the ancient era, the philosoph ies inherit in D aoism, Confucianism, Buddhism emphasized seclusion - which form ed the lifestyle s of educated Chinese men . This type of thinking was also widely employ ed in practices of reading, playing music, painting , and gardening (Liu, Burley, & Partin, 2014) . Chinese culture is also famous for a fondness for rocks, stones , and mountains, where rocks and mountains were considered aesthetic pleasure s and reminders of the flexible visual experience in daily life (Parkes, 2005) . M ountains are considered sacred and the most spectacular representatives o f earth ly power ; this type of energy is re - created in garden spaces by building artificial miniature mountains constructed of strangely shape d rocks (Parkes, 2005) . In classical Chinese gardens , a water feature is a significant design element , but great skill is needed to reproduce the metaphor of natural water and mountains in gardens (Liu, Burley, & Partin, 2014) . Names of Chinese gardens are commonly derived from romantic ideas with implied, short, hint ing or profound message s. F or example , names of classical gardens inclu de Master of Nets garden, Lingering garden , , which were all generated from philosop hical concepts (Liu, Burley, & Partin, 2014) . During the dynasty (618 - 906), a scholar - poet and prime minister, Li Deyu, built a famous rock garden near Luoyang, where he collected and arranged strangely shaped rocks from all over the country. The most incredible rocks in his garden came from Lake Tai (Tai Hu), which we re 3 to dignify them . F rom t hat po int on, inscriptions on rocks became common practice in China (Shih, 1962) . T he classical Chinese gardens are very similar in terms of the design elements and principles, while modern Chinese gardens are varied in terms of design elements and principles because of global exchange and the development of new materials and technology (Xu Y. , 2015) . For example, architectur al structures are highly valued in the design of classical Chine se gardens, but they are absent in modern Chinese gardens (Xu Y. , 2015) . In 2007, many modern gardens were built for the Garden EXPO held in Xiamen, China . Designed by many outstanding Chinese landscape designers , these gardens were an attempt to combin e classical Chinese and modern garden design elements in the modern era, which probably represent the contemporary aesthetics and considerations of garden design in China (Xu Y. , 2015 ) . Xu a nalyzed three private residential gardens in Suzhou, southern China , registered as World Heritage by UNESCO (Xu Y. , 2015) . Further information about (Xu Y. , 2015) . 1.3 Literature Review of Traditional Japanese Gardens 1.3.1 A Brief History of Traditional Japanese Gardens The history and culture of Japan has affected and produced the evolution of Japanese garden s . Japanese design principles that integrate natural landscapes and the changing seasons have also influenced Japanese garden design (Makowska, 2014) . Two considerations were involved in the creation of Japanese landscape form: one concern was to protect or isolate the inner structures against other activities, which created normal dwellings . The other concern came from the pursuit of an understanding of natural sequences and proce sses, which le d to the i nvention of formal arc hitecture and sanctified spaces (Treib & Herman, 2003) . 4 Figure 1 : A garden setting on gravel paving area in Tenryu - ji temple. (Copyright ©2017 Dexin Chen all right reserved used by permission). The Japanese gardens were derived from the sanctified space of the Shinto shrine (Treib & Herman, 1993) . The history of Japanese garden s c an be traced to the grand shrines at Ise in the fifth century, which represented t he traditions of both the Chinese and Japanese at that time (Treib & Herman, 1993) . For example, the red gateway (torii), which labels the place as a sacred space, evolved from the raised wooden storehouse of Japan; the geo metric order of the layout on the North - South axis was imported from ancient China (Treib & Herman, 1993) . The adjacent site around the shrines was covered with gravel ( Figure 2 ) , was retained as an element of garden planning , and became the formal entry court s (Treib & Herman, 1993) . In the shrines, the formal entry courts remained void; in the palaces, these courts were decorated with trees, rock s , ponds, and other ele ments that came together as a landscape (Treib & Herman, 1993) . 5 Figure 2 : A void, open gravel space adjacent to the revered hall in Yasaka Shrine . (Copyright © 2017 Dexin Chen all right r eserved used by permission). 1.3.1.1 Nara period Before the Heian perio d , Nara was the capital city of Japan in the seventh century (Treib & Herman, 1993) . In the seventh century, the an d soon became the government for most of the nation because of its splendid culture and military superpower (Brinkley & Kikuchi, 1915) Dynasty, Japanese emperor Kotoku imported the Chinese cultur e including religio us , archi tectur al , litera ry , and dress practices (Brinkley & Kikuchi, 1915) . Lavish Chinese palaces guided Japanese construction in the Nara period (Treib & Herman, 1993) . Buddhism was imported from China sometime during the sixth century, which enriched the Japanese civilization with its wo rldview (Treib 6 & Herman, 1993) . Furthermore, the philosophical ideas behind religious ritual s fostered new art forms in Japan (Treib & Herman, 1993) . Figure 3 of China i s shown in Byodo - in temple. (Copyright © 2017 Dexin Chen all right reserved used by permission). For over a thousand years, Japanese civilization used the garden and architectural styles from the ( Figure 3 ) (Treib & Herman, 1993) . The original gardens from the Nara period are all now gone ; however, the remaining paintings and texts from the Nara period recorded the rustic style landscapes and water features with islands and group ed rocks , all of which are reminiscent of the influence of Chinese prototypes (Treib & Herman, 1993) . In the Nara period, the Japanese nobility often used boat s in the Chinese style d ponds and compose d poetry. Recent excavations in Nara provided evidence of the river - styl e gardens of that period (Treib & Herman, 1993) . 7 1.3.1.2 Heian period Emperor Kammu moved the capital to Heian - kyo (now Kyoto) in 794, and from then on Kyoto became the capital city for almost a thousand years (Treib & Herman, 1993) . The (Treib & Herman, 1993) . T he palace ( Figure 4 ) was placed in the center of the city, and based on this plan, the Gosho (Imperial Palace) was a lso positioned in the center of the symmetrical city plan of Kyoto (Treib & Herman, 1993) . Japanese court life was based on imagin ing of court life in China (Treib & Herman, 1993) . The Tale of Genji (Genji Monogatari) was a book written in the early tenth century, which described how Chinese Style boating lakes were common in the court s during that time (Treib & Herman, 1993) . Nobility also adopted t ray landsca ping (Bonseki) as an artistic activit y during the Heian period (Treib & Herman, 1993) . Figure 4 : Architectural style strongly in Kyoto Imperial Palace . (Copyright ©2017 Dexin Chen all right reserved used by permission). Under comparatively peaceful political condition s , nobility had abundant time to amuse themselves by fashioning th eir gardens in the Heian period (Treib & Herman, 1993) . The Shinden style (Palatial style) of gardening design became the mainstream in the Heian period, and evolved into the Shoin - zuku ri style in the Kamakura period (Burley & Machemer, 2016). In the eleventh century, a book called Sakuteiki recorded the rules and layout pri nciples of Japanese gardens (Treib & Herman, 1993) . According to Sakuteiki (Takei 8 & Keane, 2008) , or Records of Garden Making , water should flow from the Blue Dragon to the White Tiger, where Blue Dragon symbolizes the east and White Tiger symbolizes the west in the geomancy concepts of the Four Guardian Gods . Using the Four Guardian Gods is a reflection of Japanese culture, which imported the geomancy rule s originating from central China (Takei & Keane, 2008) . Due to the reduced connection with China, Japanese civilization became more local, and garden styles transformed as a result (Treib & Herman, 1993) . For example, the river - style garden from the Nara period became more naturalized by using native plan t materials and imitating domestic country streams , rather than the Chinese p rototypes (Treib & Herman, 1993) . In the Heian period, Buddhist temples were often built with accompanying gardens (Makowska, 2014) . At first, garden s were outdoor space s for religious ceremonies and concerts. Gard ens were framed by and subordinate to the rectangular Buddhist monastery complexes built on a North - South axis. However, th e status of gardens and templ es was reversed, where temple s became subordinate in status to the garden s (Makowska, 2014) . 1.3.1.3 Kamakura period and Muromachi period In the Kamakura period, the samurai (the military class) rose in influence (Treib & Herman, 1993) . T heir leader , Shogun , became the regent of Japan (Treib & Herman, 1993) . In the twelfth century, the in Japan . The samurai favored this form of Chinese Buddhism, which later became known as Zen Buddhism (Treib & Herman, 1993) . As a result, the aesthetic of garden design in the Kamakura period was most affected by the spirit of samurai and Zen Buddhism, and an austere and symbolic style replaced the prevailing courtly style derived from the Heian period (Treib & Herman, 1993) . For example, the military class did not patronize the boating ponds for activities ; in fact, Zen philosophy influenced thought and symbolized gardens for 9 meditation became popular (Treib & Herman, 1993) . In addition, the Shinden and Shoin - zukuri style founded the basis of the Samurai (the military class) residential gardening style, which became the predecessor of the classical Japanese residential style (Burley & Machmer, 2016) . Zen culture reached its peak in the Muromachi period. Among the various artistic forms derived from the ritual of Zen, the tea ceremony (chanoyu) had the greatest influence on garden design , w hich f oster ed the growth in tea gardens (roji) (Treib & Herman, 1993) . The Chinese Sung (Song) dynasty had the greatest impact on Japanese arts during the fourteent h century (Treib & Herman, 1993) . Fig ure 5 : A erial view of right reserved used by permission). 10 1.3.1.4 Momoyama period After the Muromachi period, Toyotomi Hideyoshi controlled the country. The name of the Momoyama period was named after i n the Fushimi Momoyama district (Treib & Herman, 1993) . The aesthetic of refined poverty and simplicity was abandoned . In contrast to the previous period, the primary characteristic s of gardens in the Momoyama period were opulence and ornamentation (Treib & Herman, 1993) . Plant materials and rocks were highly cherished as priceless objects; garden workers would sometimes commit suicide or leave the district if the major plant or rock under their care was damaged (Treib & Herman, 1993) . Lesser lords devoted valued stones and plants (Treib & Herman, 1993) . T hus , Hideyoshi had a great collection of garden materials, and he ordered the gardener , Kentei , to modify the Sambo - in near his castle with th e materials (Treib & Herman, 1993) . Although the gardener spent almost twenty years trying to arrange the great amount of garden materials , the garden never achieved a harmonious transition between the various elements beca use there were too many (Treib & Herman, 1993) . The Sambo - in best represe nt s the gardens predominant during this period (Treib & Herman, 1993) . 1.3.1.5 Edo period & Meiji period The three imperial architec tural styles of the Edo period, the Katsura Villa, the Sento Gosho, and the Shugaku - in Villa best represent the aesthetic principles of th is period (Treib & Herman, 1993) . The popularity of the Stroll garden style r ose , and t he mainstream garden style s in this era were elegant and sophisticate d (Treib & Herman, 1993) . It was also popular to use t he display concept s of a pproa ch ing on the diagonal, hide and r eveal , a nd borrowed s cenery in garden design (Treib & Herman, 1993) . Because of the end of Japanese isolation during the Meiji period , Japanese traditional culture was heavily impacted by western culture 11 (Treib & Herman, 1993) . As a result, interest in traditional Japanese garde ns decreased during th is period (Treib & Herman, 1993) . Figure 6 : Dry landscape named Honryutei in Enko u - ji , imitating a dragon flying among the clouds . (Copyright ©2017 Dexin Chen all right reserved used by permission). 1.3.2 Cultural Context and Aesthetic principles 1.3.2.1 Geography Japan is an archipelago and this isolated physical location has a strong influence on Japanese culture . For example, the isolated physical location has contributed to unique Japanese internal views, which is expressed in thei r landscape design and planning (Burley & Machmer, 2016) . The native Japanese religion, Shintoism, evolved from animistic beliefs. for the spirits of Japanese ancestry and specific places (Treib & Herman, 1993) . One main concept of Shintoism is that if a place was considered sacred, it should be marked and separa ted from human residential area s (Treib & 12 Herman, 1993) . A Shinto shrine is a sacred place set for the specific spirit (Treib & Herman, 1993) . Compar ed to Chinese sacred places, Japanese sacred places tend to be implied , rather (Treib & Herman, 1993) . No clear boundar ies exist between the Shinto Shrine and other spaces ; instead a few gates (Torii), fences, straw ropes , and other signs are commonly placed inside Shinto shrines and a wall is constructed that can only be sensed , rather tha n viewed (Treib & Herman, 1993) . A recent geographical analysis in 2010 researched the 164 renowned Japanese gardens in the Kyoto Basin by using GIS data to find the link between garden locations and the geography of Kyoto (Ogata, Li, & Yamada, 2010) . The major finding from this research is that the rock garde ns (karesansui type gardens) are located in the alluvial fan; and water gardens (chisen type gardens) are clustered along the piedmont spring zones and the artificial canals of the irrigation system (Biwako sosui) constructed in 1890 (Ogata, Li, & Yamada, 2010) . These findings suggest that the hydrological and geological conditions have a large impact in determining the appropriate site for specific type s of Japanese garden s (Ogata, Li, & Yamada, 2010) . Figure 7 : Naturalness in the pathway to K asuga - taisha, a Shinto shrine in Nara . (Copyright ©2017 Dexin Chen all right reserved used by permission). 13 1.3.2.2 Zen Buddhism and Tea After the practice of drinking t ea was imported from China, the Japanese generated the t ea ceremony unde r the influence of Zen Buddhism (Burley & Machmer, 2016) . Zen Buddhism was a sect of Buddhi sm originat ing , which is also known as the sect of Buddhism (Treib & Herman, 1993) . The ritual of Zen and seated meditation ( Figure 8 ) fostered the tea ceremony, and led to the designs of tea gardens and Zen meditative gardens (Treib & He rman, 1993) . In fact, the Chinese also developed meditative gardens , but they were not the mainstream gardens in China , as they were in Japan (Burley & Machmer, 2016) . In about the 15th century, the Japanese had largely adopt ed the Chinese culture , including t ea ceremonies and Zen Buddhism . The Japanese had also created the concept of Wabi, or Wabi - Sabi, which became a unique Japanese aesthetic culture (Bullen, 2016) . Japanese garden design of the Muromachi period were impacted by the influence of a new school of painting which was derived from Zen Buddhis t practices (Stefan, Fora, Visoiu, Hernea, & Constantinescu, 2009) . A s a result, gardens were considered to be akin to the art of landscape paintings and thereby should be viewed statically , the way someone would view a painting (Stefan, Fora, Visoiu, Hernea, & Constantinescu, 2009) . Through the study o f literature and painting s from ancient Japan, it is apparent t hat Wabi is an abstract, nature - based aesthetic concept reflecting spi ritual and ethical ideas of Zen (Bullen, 2016) . B ecause of the Chinese preference for the sense of the and the interaction between Chinese and Japanese culture, the chanoyu (tea masters) developed the design principles of roji in the medieval period (Bullen, 2016) . Guided by the aesthetics of Wabi, the system of Chanoyu (Tea ceremony) was developed and the typical place set ting for tea activities was called chaya or sukiya (tea house), which was built inside of a ro ji (tea garden) (Bullen, 2016) . The setting of roji integrated Chinese philosophy and literary theory including Confucianism, Daoism (Taoism) and Buddhism , along w ith Zen 14 principles of simplicity and the exclusion of bright colors . R oji and the design principles presented wabi aesthetics (Bullen, 2016) . In general, Wabi could be us ed to describe the condition of a sense of loneliness, outside of society and materiality, focus ed on spiritual freedom, the evanescence of life, imperfection, impermanence, rustic beauty , and the elegant simplicity of natural elements (Juniper, 2003) . Figure 8 : Meditation by sitting in front of the Zen garden in Ryogen - in, a sub - temple of Daitoku - ji . (Copyright ©2017 Dexin Chen all right reserved used by permission). 1.3.2.3 Geomancy Rules Geomancy (Fengshui) also affected the design plans of traditional Japanese gardens. F or example, a stream that ran from east into a pond and emptied to the southwest of a garden usually satisf i ed the principle s of geomancy (Treib & Herman, 1993) . In the 11th century, a book called the Sakuteiki recorded the Taoist, Shinto and Buddhis t influences of gar den rules and layout p rinciples (Treib & Herman, 1993) . 15 The g eom ancy rules reflected in tr aditional Japanese garden designs originated in ancient China (Takei & Keane, 2008) . The geomancy rules act as an explanation of the (Takei & Keane, 2008) . There are three geomancy theories that guided the garden designs in Sakuteiki : Yin Yang (Theory of M utal Opposites), Wuxing (Theory of Five Phases), and Yi (Theory of changes) (Takei & Keane, 2008) . According to Sakuteiki , it was believed that the earth is the lord of water . T hus , the stream serves the mountain, and the sto counsellors; the mountain is complete when it cont ains stones (Takei & Keane, 2008) . In this example, the Five Phase theory is used as the prototype for garden design (Takei & Keane, 2008) . Geomancy has a long history of being widely applied in East Asia as an approach to explain natural phenomena and seek the balance between the human and the natural environment (Kalland, 1996) . Geomancy has also been used as guidance for urban planning, housing designs , and grave designs (Kalland, 1996) . Although geomancy was often related to superstition, geomantic principles can still be found in many architectur al an d pl anning practices (Kalland, 1996) . A study of Geomancy and town planning in 1996 found that Japanese villages are typically planned using the geomancy terms , including the flow of vital energy, yin and yang force s , the five basic elements of nature, and the pursuing of the mutual benefit of the human and nat ural environment s (Kalland, 1996) . Other than the geometric order of construction style and Buddhism, the idea of vertical disposition of rock clusters was also imported from China (Treib & Herman, 1993) . The Chinese attitude toward landscape was to treat the garden as a three - dimensional painting, and to view the landscape using ad of physical feelings (Treib & Herman, 1993) . Ther e fore, the main purpose of the design was to provide people with the impression of a landscape (Treib & Herman, 1993) . 16 1.3 .2.4 Imitation of Natural Landscapes According to Sakuteiki (Takei & Keane, 2008) , which was a garden design book written nearly a thousand years ago, the core of Japanese gardens is the art of setting stones (Takei & Keane, 2008) . At that time, w hen designing a garden, one was expected to aim to c reate a subtle atmosphere, reflect wild nature, and re - create the essence of the fa mous landscapes of Japan (Takei & Keane, 2008) . I t was believed that i f people br oke the conventional architecture rules, their households would become disorder ed (Takei & Keane, 2008) . For example, according to th e rules of Sakuteiki , stones taller than ninety centimeters should not be set close to buildings (Takei & Keane, 2008) . Aside from using a pond or stream as the major feature of a garden, another method of creating garden s included following the Dry Garden Style, which involved set ting stones in a small area without water feature s in a larger g a rden (Takei & Keane, 2008) . However , this garden style ( kara sanzui ) is somewhat different from the kare sansui in Zen temples ( Figure 9 ) and the residences of the warrior class in the later era, where designers imitated landscape s as expressed in ink paintings (Takei & Keane, 2008) . 17 Figure 9 : Rock settings and dry landscape in Daitoku - ji temple . (Copyright ©2017 Dexin Chen all right reserved used by permission). 18 1.3.3 Traditional Japanese garden styles and functions 1.3.3.1 River - style garden (Yarimizu) a river - style garden ( Figure 10 ) featured a riverway or stream with rock outcroppings to form a pond (Treib & Herman, 1993) . The earliest know n written book of Japanese gardens, the Sakuteiki , recorded the rules of how to construct a river - style garden: struck stones would make water change directions and dash ; it was considered preferable to p lace stones in a manner so that the feeling gradually changes (Treib & Herman, 1993) . The nobility of that period tended to own such gardens (Treib & Herman, 1993) . Figure 10 : River style garden in Kyoto Imperial Palace. (Copyright ©2017 Dexin Chen all right reserved used by permission). 19 1.3.3.2 Shinden style Garden (Palatial Architecture) Shinden style is a unique garden design style of Japanese palatial architecture which began in the Heian period. Shinden style architecture usually consist s of a group of symmetrical ly arranged rectangular buildings and pavilions, which are connected by the cove red corridors with courtyards or atriums, and the structures are placed in a geometric U - shape around a large central court (Burley & Machmer, 2016) . The style is very similar to the ancient Chinese and Korean palatial design practices. Another design conc ept of the Shinden style is using the enclosed bu ildings to define the landscape (Burley & Machmer, 2016) . The outdoor space , or the main court of the architecture , usually includes a stream or large boating pond as the major feature to mimic greater natur al landscape forms (Burley & Machmer, 2016) . The pond s later bec ame the cent er of the strolling gardens (Burley & Machmer, 2016) . 1.3.3.3 Shoin - Zukuri S tyle Garden The Kamakura period started after the H eian period, where the samurai and their leader , shogun , became the new noble elite (Burley & Machmer, 2016) . Due to the political change, the culture of simplicity and devotion to duty was promoted , while personal pleasure was discouraged (Burley & Machmer, 2016) . As a result, the archi tectur al style turned to small scale designs. The gardens and the buildings from the Kamakura period share the se common characteristics: small, quiet, and asymmetrical (Burley & Machmer, 2016) . Shoin - Zukuri style is a smaller asymmetrical residenti al design style that evolved from the large , symmetrical Shinden style (Burley & Machmer, 2016) . Shoin - Zukuri style is considered an important innovation of Japanese architectur al design because it uses the enclosed landscape to define the buildings and make the buildings thems elves features . This stands i n contrast to the Shinden style (Burley & Machmer, 2016) . 20 In the Shoin - Zukuri style, the main garden is usually seclu ded and the ponds are smaller than in the Shinden style ; ponds tend to include reflection ponds or fish ponds ; Shoji - screens are placed to blend the indoor and outdoor spaces ; and dwelling walls are designed as picture windows (Burley & Machmer, 2016) . Although it was a new design feature for the Japanese ton use screens , the Chinese started to use screens during the Zhou dynasty (Burley & Machmer, 2016) . The Dojinsai Tea Room within Ginkaku - ji in Kyoto is an example of Shoin - Zukuri style (Burley & Machmer, 2016) . 1.3.3.4 Zen G arden Buddhism, which is also known as Zen Buddhism , was introduced in Japan and became a par t of Japanese cultural practices (Burley & Machmer, 2016) . At that time, the military class (samurai) favored and accepted because of the devoti on to simplicity and duty inherent in Zen beliefs (Treib & Herman, 1993) . The principle of Zen focused on the philosophy in daily life, which pursues austerity, eternity , and simplicity (Treib & Herman, 1993) . Zen and its related aesthetics had a great influence on the Japanese arts including gardens, where garden s became place s merely to be viewed (Treib & Herman, 1993) . T he garden s influenced by Zen were set for a series of perspectives, and the viewpoints were tightly controlled (Treib & Herman, 1993) . One of the first gardens that present ed the Zen influences was included in Tenryu - ji, where the pond garden was created in the Heian style, and the vertical arrangement of rock clusters was derived from Chinese aesthetics of the Southern Sung dynasty (Treib & Herman, 1993) . Later gardens like Ryoan - ji, continually dim inished the number of garden materials until rock and gravel were highlighted as the primary features of the garden s (Treib & Herman, 1993) . In addition, Buddhist temples were often located near Shinto Shrines, 21 which implies that Buddhism and Shinto ism have a h armonious relationship in Japan (Burley & Machmer, 2016) . In the Muromachi period , the principle of a dry garden was developed for meditative use by providing symbolic space (Treib & Herman, 1993) . The dry gardens mimic the natural Ja panese landscape by contrasting with displayed as sand, gravel, pebbles, and rocks (Burley & Machmer, 2016) . The dry gardens provide a symbolic sense of archipelago through pure abstract views at a small scale (Burley & Machmer, 2016) . In the Edo period , the capital city was moved from Kyoto to Edo (no w Tokyo), where the natural landscape was very different (Burley & Machmer, 2016) . The rocks of the dry gardens were replaced by nicely trimmed shrubs because of the lack of rock formations in Edo (Burley & Machmer, 2016) . Most of the Zen gardens ( Figure 11 ) provide d a sense of enclosure through walls, and the sense of enclosure also came from the function of the gardens during a civil ly tumultuous period in Kyoto, where walls defined space s for tr anquility (Treib & Herman, 1993) . 22 Figure 11 : Moss featuring Zen garden with sitting area named Ryugintei in Ryogen - in, a sub - temple of Daitoku - ji. (Copyright ©2017 Dexin Chen all right reserved used by permission). Zen garden design includes a seating area that guides and limits people to sit down, view the area, and meditate , such as the roofed veranda in the Ryoan - ji garden (Treib & Herman, 1993) . By viewing the miniature landscape, people can stay and engage in self - reflection in the dry gardens and enjoy the spac e of simplicity and tranquility (Burley & Machmer, 2016) . In general, the experience is provided b y both the scenery and the mi nd (Treib & Herman, 1993) . The temple corridors serve as both walkways and sitting area s , and in this way, the perspective of view is changed and determined through the action of sitting, standing , or walking (Weiss, 2010) . T he se gardens are small; thus, people are able to easily realize the large visual difference s between viewpoints because the view changes as one moves (Weiss, 2010) . 23 1.3.3.5 Tea G arden (Roji) Importing tea from China to Japan possibly dates back to the fifth century (Treib & Herman, 1993) . The Japanese modified it with their own traditions and developed another unique Japanese c ultural practice , the t ea ceremony (Treib & Herman, 1993) . Chinese cultural philosophy and Chinese art i facts , including Chinese paintings , were imported to Japan and f oster ed the aesthetics of the tea ceremony (Bullen, 2016) . About 800 years ago, the Japanese Buddhist monks dra nk tea to stay concentrated through meditation and this act fos tered the t ea ceremony in Japan (Burley & Machmer, 2016) . In the sixteen th century, tea masters modified the ritual of the tea ceremony according to their preferences (Treib & Herman, 1993) . The Chanoyu masters regularly practiced the t ea ceremony , which was typically h eld in a teahouse (Chaya, sukiya) within a roji (tea garden) (Bullen, 2016) . Through a series of acts of tea making that follows sophisticate d rules, the tea ceremony allows people to share emotions and focus their minds (Sakuae & Reid, 2012) . Recent findings reflect that such a specific place for occupational engagement encourage s a pause in usual routines, engages learn in g skills, and fosters connection between people (Sakuae & Reid, 2012) . With the increasing popularity of tea drinking, tea tasting parties were held for introducing various tea species and tastes while testing t he sophistication of the guests (Treib & Herman, 1993) . A t ea garden ( Figure 12 ) is a specific place for tea a ceremony, where people can empty their mind for simplicity while making tea (Sakuae & Reid, 2012) . The size of roji varies greatly , but the setting of roji was standardize d in the Edo period, where a fence block ed roji from the outside world (Bullen, 2016) . T he composition of roji implies - (Bullen, 2016) . Roji includes : a small waiting house (machiai) , an outer garden (soto niwa) containing an outside waiting arbo u r (koshikake machiai), a f - 24 , and an inner garden (uchi niwa) , which contains a wa ter - basin and teahouse (Bullen, 2016) . Figure 12 : A tea house in a bamboo fenced roji located in Shisen - do . (Copyright ©2017 Dexin Chen all right reserved used by permission) . At fir st , the tea house was developed as a proper setting for a tea ceremony, and a dew path that surrounds the tea house was called the tea garden (roji) (Treib & Herman, 1993) . The tea garden function ed as a path that le d people to the tea house , rather than serving as an embellished gar den for people to admire (Treib & Herman, 1993) . In addition, the aesthetic of Wabi - S abi had a great impact on the tea garden and tea house design, which pursued the art of the simple l ife with insufficient materials (Bullen, 2016) . T ea garden s were used to differ entiate the space for the tea ceremony from othe r gardens in the garden complex 25 (Burley & Machmer, 2016) . In a way , the dew path itself provided a transitio n of the psyc he (Treib & Herman, 1993) . 1.4 Design Rules and Major Feature s in Traditional Japanese Gardens 1.4.1 Borrowed scenery (Shakkei) The concept of borrowed scenery ( Figure 13 ) was first used in the Tenryu - ji garden (Treib & Herman, 1993) . To fit the design site into the surrounding landscape without destroying the ru ral natural environment, the idea of the Tenryu - ji garden borrow ed from distant landscape element s and incorporate d them into the design of the site (Treib & Herman, 1993) . When a garden is created in the foreground of a building, us ing plants, walls or hedges to block the undesirable views in the middle ground ensures a smooth transition from foreground to background . For example, the two mountains, Arashiyama and Kameya ma, which are behind Tenryu - ji in Kyoto , are seen as part of the garden ( Treib & Herman, 1993) . In this case, the borrowed mountains integrate with the designated pond, rocks and plants harmoniously (Treib & Herman, 1993) . In the early Edo period, the borrowed scenery design was skillfully used in garden designs, including Shoden - ji, Entsu - ji and the upper garden of Shugaku - in (Treib & Herman, 1993) . 26 Figure 13 : The mountain is borrowed as background in this view as Borrowed S cenery in Tenryu - ji temple. (Copyright ©2017 Dexin Chen all right reserved used by permission). 1.4.2 Hide and Reveal The concept of h ide and reveal exist ed in the garden of Sento Gosho of the Edo period (Treib & Herman, 1993) . In Sento Gosho, the direction of the pathway first turned left to grant a view of a tree, and then turned right to grant a view of a group of rocks on the pond . Following the view of the pond, the path then a pproached stepping stones which encourages people look down . Finally, a building emerged in the front (Treib & Herman, 1993) . In addition, due to the different views of the four seasons, some desi gnated views could only be seen during a specific season, and the entire design intent c ould only be revealed over time (Treib & Herman, 1993) . 27 Figure 14 : The whole view is blocked by a wall , but a p art of it is exposed . Photo taken in Kyoto Imperial Palace. (Copyright ©2017 Dexin Chen all right reserved used by permission). 1.4.3 Rock position s The most attractive element of the dry landscapes, or positioned under a kind of traditional gardening rule, which aims at exhibiting the abstract concept of traditional Japanese culture and simplicity (Van Tonder, Lyons, & Ejima, 2002) . It is the un conscious perception that leads to the mystic al appeal of the garden (Van Tonder, Lyons, & Ejima, 2002) . One famous example is the Ryoan - ji rock garden ( Figure 15 ) , where the fifteen rocks were placed in a sophisticated composition (Treib & Herman, 1993) . The myth of rock arrangement is that one rock is always hidden when viewing the rock garden from any poin t of the seating area (veranda) (Treib & Herman, 1993) . 28 Although e ntering is prohibited in most of the Zen gardens , walking and sitting are integrated into the garden design (Weiss, 2010) . In Ryoan - ji, mobility became part of the myth of the rock arrangement in the Zen garden, because only fourteen out of fifteen rocks in total are visible from any perspective (Weiss, 2010) . T he incessant motions of people influence the meditation process (Weiss, 2010) . Figure 15 : Zen garden in Ryoan - ji temple. (Copyright ©2017 Dexin Chen all right reserved used by permission). According to Sakuteiki (Takei & Keane, 2008) , large stones should typically be set adjacent to a water fall, on the tip of an island or at the rear of a hill (Takei & Keane, 2008) . Solitary stones (hanere ishi) should be set on a stone shore, at the foot of a hill or at the tip of an island (Takei & Keane, 2008) . Furthermore, solitary stones can also be set in water with a foundation. F irst , one should set several large stones in a trian gular shape under the water surface. T hen , one should set the solitary stone in the center of th e triangle, and finally place foundation al stones around the solitary stone (Takei & Keane, 2008) . Most of the stones in the garden are set horizontally. (Takei & Keane, 2008) . 1.4.4 Water Features ( Sakuteiki ) According to Sakuteiki (Takei & Keane, 2008) , a water level should be established before construction (Takei & Keane, 2008) . In the design plan, the surface of the pond should 29 be set twelve to fifteen centimeters below the bottom edge of the veranda of the fish pavilion, so that the exposure percentage of a given stone can be controlled through the water level (Takei & Keane, 2008) . M oreover, the soil base under the stones must be strengthen ed with foundation stones. In this way, the stones are able to last for a long time and even if the pond is drained, the stones will look like they were well set (Takei & Keane, 2008) . This technique is found in Tenryu - ji : when the pond is drained during reparation, the foundation stones are revealed (Takei & Keane, 2008) . When designing islands in a pond , if set stones along the edge of the island are buil t , the stones will fail after the water is added (Takei & Keane, 2008) . One should set stones around the desired edge of the island, and gradually identify the shape of isla nd (Takei & Keane, 2008) . There are many styles of islands including Mountain Isle, Meadow Isle, Forest Isle, Rocky Shore Isle, Cloud Type, Mist Type, Cove Beach Type, Slender Stream, Tide Land , an d Pine Bark (Takei & Keane, 2008) . Figure 16 : An island with its reflection in the pond. Photo taken in Kyoto Imperial Palace. (Copyright ©2017 Dexin Chen all right reserved used by permission). 30 When a waterfall is established , one should choose a W aterfall stone which allows water to fall over and is rough , while harmoniously fitting the Bracketing stones (Takei & Keane, 2008) . If the major viewpoint is from the right side, set a pretty stone of appropriate height above the left Bracketing stone (Takei & Keane, 2008) . On the opposite side , set the stone above the right b racketing stone to achieve the best vi sual result (Takei & Keane, 2008) . The low waterfalls that expose the source of the water stream lack depth and become unimportant (Takei & Keane, 2008) . Waterfalls look best if they appear unexpectedly from narrow split s between half - hidden stones in shad e s (Takei & Keane, 2008) . During the of China, the Chinese always buil t wellsprings in residences to bring in coolness in the hot weather (Takei & Keane, 2008) . The wellsprings often appear ed in ed out of the mouth of animal sculptures (Takei & Keane, 2008) . A Buddhis t story in ancient India revealed that the Earth Goddess , was ordered to build a wellspring in the monastery garden, , which is called the S weet S pring (Takei & Keane, 2008) . In Japanese gardens, i t is better to s et a roof above the wellspring, a pipe inside the wellspring to push the water flow, and a small, slatted deck to let water pass through (Takei & Keane, 2008) . An artificial wellspring can be used to fill a well when a large water tank is constructed beside the well (Takei & Keane, 2008) . 31 Figure 17 : Stone hand wash basin in Tenry u - ji. (Copyright ©2017 Dexin Chen all right reserved used by permission). The Chinese classics books, Four Books/Five Classics and the Four Guardian Gods theory underlie the decision to have the water flow in different directions (Takei & Keane, 2008) . According to th e books , w ater should flow from East to South, an d then West (Takei & Keane, 2008) . The stones in the garden stream should not follow a uniform pattern or be gather ed in a crowd; they must be stuck deeply into the riverbed (Takei & Keane, 2008) . These stones can be divided into five types: Bottom Stones, Water - Splitting Stones, Foundation Stones, Crosswise stones, and Spil lway Stones (Takei & Keane, 2008) . 1.4.5 Clipped Bushes In the Edo period , b ushes were clipped to maintain a certain shape as a design feature (Treib & Herman, 1993) . Gardeners would selectively pluck the blossoms of the bushes to 32 balance the color that change d the space (Treib & Herman, 1993) . For example, the traditional dry garden in Konchi - in was composed of a large clipped shrub on a raked gravel plane , which was thought to be a (Treib & Herman, 1993) . In contrast to using topiary as living statue s in western traditions, another example of using clipped plant material as a design fe ature is found Daichi - ji, where the hedge was clipped to represent a dragon or wave and bec a me the major feature of the garden (Treib & Herman, 1993) . 1.5 Conclusion Over time, the traditional Japanese gardening construction and design system became a cultural product since its development in ancient Japan (Burley & Machmer , 2016) . The traditional Japanese gardens constructed during modern times essentially repeated the ideas embellished in the traditional design styles (Burley & Machmer, 2016) . In 1853, Japan opened ports to the west for trade and decided to promote its cultur al dissemination by exporting cultural products, such as the Japanes e gardening culture (G oto, Ristovska, & Fujii, The Japanese garden at Sonnenberg: the first traditional private Japanese garden in North America, 2014) . The Nihon Fakeiron (Japanese Landscape) written by Shiga Shigetaka, demonstrated the traditional Japanese elements that had become an inseparable part of the Japanese people and geography (Gavin, 2000) . Shiga highly valued the potential of promoting the Japanese landscape to raise national awareness and Japa world order (Gavin, 2000) . In fact, Japanese - style gardens have grown tenfold outside of Japan compared to the number of Chinese - style gardens constructed outside C hina since the 1860s (Gregory Kenneth Missingham Faculty of Architecture, 2007) . The primary reason for the large growth in Japanese gardens compared to Chinese gardens may be that access to Japanese gardens and the abili ty to learn traditional garden design s in Japan was easier than 33 entering the Chinese gardens in China or learning Chinese gardening designs (Gregory Kenneth Missingham Faculty of Architecture, 2007) . Some Chinese gardens were private gardens that we re not open to the public ; in contrast, many Japanese gardens were located in places for public worship , such as at Buddhist temple s and Shinto shrines (Gregory Kenneth Missingham Faculty of Architecture, 2007) . A lso, a series of educational aid s, including monasteries and universi ty programs were provided in Japan for foreigners interested in Japanese gardens ; but many Chinese craft traditions vanished with the fall of the imperial system (Gregory Kenneth Missingham Faculty of Architecture, 2007) . The spread of traditional Japanese gardens was aided in many aspects, while the construction style of Chinese gardens was outside of the knowledge of many western countries. In conclusion, the key principle of traditional Japanese gardens is the spatial relationship between structure and landscape, where the aesthetic and spaces are impli ed instead of stated (Treib & Herman, 1993) . The abstract view consisted of the mutual relationship , not the physical component (Treib & Herman, 1993) . It should be understood that a trad itional Japanese gar den required a great deal of maintenance , otherwise the views that provide d a sense of naturalness would rapidly disappear (Treib & Herman, 1993) . After perceiving and adopting many aspects of the ancient Chinese cultur al rules for gardening, including geomancy rules, aesthetics, and architectur al styles, Japanese gardens developed their own characteristics (Treib & Herman, 1993) . In the Edo period, the design principles and construction rules of the traditional Japanese garden were almost fully finalized (Treib & Herman, 1993) . Although the traditional Japanese gardens also follow ed a constructed design style, each garden tended to be unique (Treib & Herman, 1993) . The traditional gardens and ancient city planning reflected the culture a nd history of Japan (Treib & Herman, 1993) . Kyoto was the capital city in the foundation al periods of the Japanese garden style (Treib & 34 Herman, 1993) . Thus, an analysis of the ancient architectures in Kyoto is necessary to provide a complete overview of traditional Japanese gardens. In 2003, a study was conducted to evaluate of classical Chinese and traditional Japanese gardens using a series of survey s and scientific method s (Zhao, Matsumoto, Liu, Yuan, & Kawata, 2003) . The study from 2003 (Zhao, Matsumoto, Liu, Yuan, & Kawata, 2003) was based on a literal understanding of gardens, and the author did not look into the garden design process and design elements . Therefore, the current study aims to understand the design elements and principles using a statistical method to compare Chinese gardens and Japanese gardens . 35 CHAPTER 2: METHODOLOGY 2.1 Purpose of Study This research utilizes Principal Compone n t Analysis to gain a better un derstanding of the mutual relationship between oriental gardens, and to compare traditional Japanese gardens, classical Chinese gardens , and modern Chinese gardens . The primary goal of this research is to seek the similarities or differences between the selected gardens and develop theories as to what make s these gardens similar or different from the distinguishable garden elements . Moreover, th e distinguishable garden element s are considere d the primary indications of authenticity in traditional Japanese garden designs . Through integration of the results from the principal component analysis and literature review, the aesthetic evolution s of oriental garden designs can be suggest ed . Previous research using cluster analysis , generated through principal component analysis , includ es Yiwen gardens (Xu Y. , 2015) (Xu H. , 2017) . There primary goals of this research include : 1. Analyze the mutual relationship between traditional Japanese gardens and classical Chinese gardens ; 2. Analyze the mutual relationship between traditional Japanese gardens and modern Chinese gardens; 3. Discuss the major design elements that divide or bring together t he selected gardens. 2.2 Study Sites 2.2.1 Classical Chinese Gardens in Suzhou and Modern Chinese Gardens In this study, the three classical Chinese gardens in Suzhou and five modern Chinese as study objects (Xu Y. , 2015) . The three Chinese gardens are the Humble A s Garden, Master of Nets G arden , and Lingering Garden. The five modern Chinese gardens are the Bamboo Garden, Net. Wet . 36 Garden, Learning Garden, Sugar Cane Garden , and Landscape New Wave Garden. Detail ed information about each of these (Xu Y. , 2015) . 2.2.2 Traditional Japanese Gardens in Kyoto The Chinese had the greatest influence during the Nara period . A fter the Nara period, the Heian period began and the emperor Kammu moved the capital city from Nara to Kyoto. Kyoto was developed based on the symmetrical plan of the capital city of the it became the capital city of Japan for almost a thousand years. Comparing the it is easy to find the ir common characteristics: streets are planned in an order of importance ; the streets define d the rectangular blocks in a rectangular s ite plan; a rchitecture was oriented North - Sout h following geomancy principles; the imperial palace was located in the North of a broad central avenue that divided the city into Eastern and Western districts. The design concept emphasized the idea of imperial contr ol, hierarchy, and regent power (Treib & Herma n, 1993) . Kyoto is renowned for its traditional Japanese gardens . In this research, eight traditional Japanese gardens and two Shinto Shrines in Kyoto, Japan, are selected as the obje ct s of study. 2.2.2.1 Shisen - do I n the Edo period, Ishikawa Jozan escaped to Kyoto because of his opposition to military rule and built Shisen - do in 1636. During the forty years he lived in Shisen - do, Ishikawa studied tea, arts, philosophy , and garden design (Treib & Herman, 1993) . At first Shisen - do was designed to be a private villa, later it bec ame a Zen temple (Mehta & Tada, 2012) . Portraits of thirty - six Chinese and Japanese poets were displayed on the wall of the main building. The design of Shisen - do was called . T he main viewing 37 veranda , the raked sand area , the nicely trimmed bushes , and the natural vegetation in the background comp rise the major aspects of Shisen - do (Treib & Herman, 1993) . Shisen - do is also famous for its unique water device, Shishiodoshi , which is made of bamboo and generates sounds when motivated by the water stream to hit the rock (Mehta & Tada, 2012) . The original function of this device wa s to scare deer and other animals away from the crops (Fowler, 2015) . S hishi - odoshi can be considered an artificial technique to produce a soundmark or a rhythmic counterpoint to a background sou nd from constant falling water ; it is believed that give typical character s to the garden and remind people of the history of agriculture (Fowler, 2015) . Figure 18 : Meditation in Shisen - do while viewing the dry landscape. (Copyright ©2017 Dexin Chen all right reserved used by permission). 38 Figure 19 : Shishiodoshi in Shisen - do. (Copyr ight ©2017 Dexin Chen all right reserved used by permission). 2.2.2.2 Byodo - in In the Heian period, Fujiwara Yorimichi , a member of the imperial court, built a palace in the former villa of Fujiwara Michinaga, which was turned into a Buddhis t temple in 1052. Of all the building s , only the Phoenix Hall survived because it best represented the architectur al style of the Heian period . Functionally , the Phoenix Hall is used as a chapel of Amida Buddha. The main hall and the arcades on both sides were believed to be symbolic of a phoenix land ing or taking off . The entire setting of the garden was based on the imagination (Treib & Herman, 1993) . 39 Figure 20 : The phoenix hall of Byodo - in temple. (Copyright ©2017 Dexin Chen all right reserved used by permission). 2.2.2.3 Kiyomizu - dera In 780 , S akanoweno - Tamu - ramaro founded Kiyomizu - dera . Currently, t he temple present today is a re construction that Tokugawa Iyemitsu built in 1633. The main hall was built on a deep ravine with the support many columns , which refer r ed to the Shinden style . The Kiyomizu complex is a great example of Japanese spatial In this complex, the rigid, geometrical setting style of architecture was abandoned , while the unexpected appearance of the buildings arose from the merging with nature and created a splendid spatial composition (Treib & Herman, 1993) . 40 Figure 21 : Pagoda and wisteria pergola in Kiyomizu - dera. (Copyright ©2017 Dexin Chen all right reserved used by permission). 41 Figure 22 : The stairs before getting to the gateway of Kiyomizu - dera , creating a sense of entering a sacred place . (Copyri ght ©2017 Dexin Chen all right reserved used by permission). 42 2.2.2.4 Daikaku - ji In the early Heian period, emperor Saga R ikyu - in built the imperial villa in Kyoto , and Daikaku - ji originated from the imperial villa of E mperor Saga . Daitoku - ji is famous for the oldest artificial garden pond in Japan, Osawa - no - ike P ond , which has perimeter of about eight - hundred meters. Emperor Saga constructed the lake to imitate Lake Dongting in China. Figure 23 : A view towards Osawa - no - ike pond. (Copyright ©2017 Dexin Chen all right reserved used by permission). 43 Figure 24 : The lifted covered walkway in Daikaku - ji. (Copyright ©2017 Dexin Chen all right reserved used by permission). 2.2.2.5 Tenryu - ji Shogun Ashikaga Takauji built Tenryu - ji in about 1339 (Kamakura period). A famous garden designer and priest named Muso Kokushi designed the garden . Muso believed in the value of meditating while viewing gardens. The garden style was both Heian pond style and Chinese (Sung dynasty) style, which produced a sophisticate d landscape (Treib & Herman, 1993) . The unique design feature of the Tenryu - ji garden is a cluster of seven rocks in the pond and group of seven rocks is raised in the water , and it provide s strong vertical textures that contrast h with the horizontal pond surface. The composition of the seven rocks are highly valued. The reflection of the rocks in the pond lengthen s the 44 vertical shape of the rocks ; thus , the impression of t he rock composition is enhanced (Treib & Herman, 1993) . Figure 25 : The main hall facing to the pond of Tenryu - ji . (Copyright ©2017 Dexin Chen all right reserved used by permission). 45 Figure 26 : The raked sand design is merged into the view as foreground; the rock groupings suggest the pond edge as midground; the mountain behind is borrowed as background. (Copyright ©2017 Dexin Chen all right reserved used by permission). 2.2.2.6 Ryoan - ji Ryoan - ji rock garden was constructed in about 1500 ; although the original temple building was ruined by a fire in the (Treib & Herman, 1993) . Ryoan - ji i s famous for its dry garden and the sense of Zen and enclosure (Treib & Herman, 1993) . The most famous feature of the Ryoan - ji garden is the dry garden (karesansui) (Treib & Herman, 1993) . Enclosed by a wall, the dry landscape consists of fifteen rocks placed on a large flat plane of gravel , which reflects Zen Buddhis t aesthetic principles of simplicity and the Chinese influence of rock composition (Treib & Herman, 1993) . The design is entirely symbolic , where water and the landscape elements are represented only by rocks and raked gravel. In this dry garden, except for the moss that surrounds the rock pieces, none o f the 46 plant material is used (Treib & Herman, 1993) . Most of the Zen gardens emphasize the sense of enclosure, for example, the surrounding wall in the Ryoan - ji garden works as a visual boundar y against the gravel plane (Treib & Herman, 1993) . Figure 27 : One of the fifteen rocks in the Zen garden of Ryoan - ji. The rectangular wall brings a sense of enclosure. (Copyright ©2017 Dexin Chen all right reserved used by permission). 47 Figure 28 : A Torii is put in the pond in Ryoan - ji, suggesting the Shinto influence of this Buddhis t temple . (Copyright ©2017 Dexin Chen all right reserved used by permission). 2.2.2.7 Kinkaku - ji /Rokuon - ji (The Golden Pavilion) T he cultur e of the Chinese Sung (Song) dynasty dominated the Japanese arts in the fourteenth century with t he promotion of the third Ashikaga shogun , Yoshimitsu (Treib & Herman, 1993) . He or ganiz ed a group of arti st s, poets and Zen priests who h ad been to China to help build a palace complex , aimed to i mitate the Sung style a s well as the Heian Shinden style this palace complex became a Zen temple (Treib & Herman, 199 3) . F ew of the original buildings survived during the long history (Treib & Herman, 1993) . T he large boating pond, rock groupings and extensive plantings, and the pavilion comp ri sed the ma in view of - ji (Treib & Herman, 1993) . The pavilion was designed to exhibit the cha racters learned from the Chinese Sung style (Treib & Herman, 48 1993) . In 1950 , the original unpainted pavilion was destroyed by fire , but an exact copy was constructed s oon after the fire (Treib & Herman, 1993) . The exterior walls of the pavilion are painted in gold to match its name of the Golden Pavilion (Treib & Herman, 1993) . Figure 29 : The golden pavilion and its reflection in Kinkaku - ji temple. (Copyright ©2017 Dexin Chen all right reserved used by permission). 2.2.2.8 Daitoku - ji Daitoku - ji is one of th e largest Zen temple complex es in Kyoto , which contains many prayer halls and twenty - three sub - temples (Treib & Herman, 1993) . Daitoku - ji was founded in 1319 by the priest Shuho Myocho (later called Daito Kokushi), but later it was destroyed by fire and war (Treib & Herman, 1993) . The priest Ikkyu rebuilt Daitoku - ji in 147 4 (Treib & Herman, 1993) . Many of the sub - temples were built i n the sixteenth centu ry when the military government supported Daitoku - ji (Treib & Herman, 1993) . The layout of the entire 49 Daitoku - ji area exhibits the transformation from formal p lanning to informal planning , where a North - South axis that derive s from the traditional formal religious architectur al manner of Japan and China underlies the layout of the major buildings , and a cluster of the small sub - temples encircle s the main halls (Treib & Herman, 1993) . The central main halls are akin to the symbolic and visual center of the whole area ; in contrast , the sub - temples are treat ed as private center s of religious ritual (Treib & Herman, 1993) . T he sub - temple s express unique characters and Zen Buddhism in the garden design , because the garden is believed to be the most significant part of meditation in Zen practices (Treib & Herman, 1993) . Figure 30 : T he dry landscape - in, a sub - temple of Daitoku - ji . (Copyright ©2017 Dexin Chen all right reserved used by permission). 50 Figure 31 : Stepping stones and solitary stones - in, a sub - temple of Daikaku - ji . (Copyright ©2017 Dexin Chen all right reserved used by permission). 2.2.3 Shrines in Kyoto 2.2.3.1 Yasaka Shrine /Yasaka Jinsha In the ninth century, Fujiwara Mototsune turned his residence into the main shrine, and the architecture was restored in an imperial palace style . In 1654, the Yasaka shrine was reconstructed in old Shinden style according to the old prototype . T he main shrine, the small subordinated shrines , the two - storied gate ( which was built in the Kamakura period ), and the stone Torii ( whic h was one of the largest stone T orii in Japan ), were the main features in this area . 51 Figure 32 : The main gate with viewing terrace in the Yasaka shrine . Chinese characters are used in the informative plaque s. (Copyright ©2017 Dexin Chen all right reserved used by permission). 2.2.3.2 Fushimi - Inari Taisha The main hall of Fushimi - Inari Taisha was restored in 1499 in the nagare - zukuri style with an extra - long eave (Cali & Dougill, 2013) . The red - painted woodwork and the white stucco walls are typical character istics of Inari shrines (Cali & Dougill, 2013) . The five - thousand red Torii together compose a symbolic tunnel of a walking trail, which begins from the main hall and climbs up to the Inari M ountain (Cali & Dougill, 2013) . The pathway of the red Torii strongly contrast s with the green background of the natural environment, which provides a breathtaking scene (Cali & Dougill, 2013) . Thousands of f ox statu e s and stone 52 altars that are in scribed with the names of the gods or animistic belie fs are placed all over the site, reflect ing the traditional practices of Inari (Cali & Dougill, 2013) . Figure 33 : Thousands of Torii cover the outdoor pathway in Fushimi - Inari Taisha. (Copyright ©2017 Dexin Chen all right reserved used by permission). 2.3 Data Collection This research will compare the eight Japanese gardens and two shrines in Kyoto , Japan which are selected by the author, Dexin Chen , to the Chinese gardens that were selected by Yiwen Xu in a previous study (Xu Y. , 2015) . The first seventy - f our design 53 A Cluster Analysis Comparison of C lassical Chinese Gardens with M odern C hinese G ardens , the other sixty design elements are selected base d on the (Xu Y. , 2015) . Th us, a list of one hundred and thirty - f our design elements in total is generated , as described below ( Table 1 ) . Table 1 : List of design elements and principles included in the eighteen selected sites. 1. The Great Halls (ting tang) 2. Covered Stone Boat (fang) 3. Viewing Towers (lou ge) 4. Studies (shufang) 5. Covered Walkways (lang) 6. Pavilions (ting xie) 7. Viewing terrace 8. Black tile pavement 9. Brick paving 10. Cracked Ice Stone paving 11. Pebbles area 12. Mosaic pave with special pattern 13. Whitewashed walls 14. Grey Stone Walls 15. Openwork Brick walls 16. Curved top walls 17. Zigzag wall 18. Meandering walls 19. Bamboo paved pathway 20. Boardwalk 21. Curved Pathway 22. Straight Pathway 23. Zigzag Bridge 24. Semi - circular Bridge 25. Straight Bridge 26. Wall holes with symbolized shape 27. Lattice window 28. Moon Gate 29. Wood carvings 30. Glass carvings 31. Brick carvings 32. Reflecting Pond 33. Stream 34. Fish pond 35. Wetland 36. Island 37. Artificial mountains 54 Table 1 ( c on t 38. Sculptural rocks 39. Pond bank rocks 40. Taihu Rocks 41. Trees 42. Shrubs 43. Ground covers 44. Turf area 45. Pine 46. Bamboo 47. Plum 48. Magnolias 49. Camellia 50. Crepe myrtles 51. Sweet osmanthus 52. Peony 53. Willow 54. Lotus 55. Reed 56. Sugar cane 57. Moon 58. Clouds 59. Rain 60. Wind 61. Shadow 62. Originally private 63. Public 64. Located in suburban 65. Located in urban 66. Design concept 67. Poem and painting concept 68. Naturalness 69. Varied spaces with visual devices 70. Borrowed scenery 71. Enframed scenery 72. Opposite scenery 73. Contrast 74. Deep implication 75. Abstract geometrical composition 76. Miniature of natural landscapes 77. Boating pond 78. Stepping stones 79. Yellow painted wall 80. Unpainted structures or el ements 81. Fall color of plant material 82. Metal components in wooden structures 83. Courtyard/atrium 55 Table 1 ( c on t 84. Wooden / bamboo fences and railings 85. Bench 86. Pagoda 87. Adjacent to burial site 88. Stairs 89. Geomancy rules/Fengshui 90. Stone lanterns 91. Wooden lanterns 92. Trellis/pergola 93. Obviously exposed drainage ditch/gutter 94. Signage stone 95. Elevation change 96. Plaque 97. Lifted base of building 98. Building partially painted in red 99. Strings implied enchantment boundary 100. Raked sand design 101. Stone hand wash basin 102. Well 103. Bell 104. Shishiodoshi 105. Grouping stones 106. Ophiopogon japonicas 107. Hydrangea 108. Japanese maple 109. Moss 110. Azalea 111. Trimmed plant material 112. Weeping form plant material 113. Multiple layers of entry sequence 114. Multiple layers of edges 115. Control view point 116. Foreground, Midground and Background (visual technique) 117. Large scale 118. Small scale 119. Located on the foot of mountain 120. Location adjacent to Shinto shrine 121. Curved roof 122. Chinese characters 123. Wabi - sabi 124. Religious property 125. Tea 126. Sense of sacred place 127. Animistic believes 128. Buddhism 129. Shinto 56 Table 1 ( c on t 130. Confucianism 131. Imperial background 132. Simplicity 133. Tranquility 134. Harmony 2.4 Analysis Techniques A cluster analysis is applied in this analysis focused on a garden comparison in order to identify the similarit ies or differences by grouping similar item s into categories. Following the same steps used in a statistical analysis software program called SAS (version: 9.4 TS Level 1M2 , X64_8PRO platform, English , Copyright© 2002 - 2012 by SAS Institute I nc., Cary, NC, USA. ) is us ed to build principal component analysis (PCA) as the first step of the cluster analysis (Xu Y. , 2015) . By standardizing the original variables, the weight of the variance can be calculated by PCA to help generate the final score s of each studied site (Xu Y. , 2015) . In a research of suitability overlay in Southern Michigan, the author s Jon Bryan Burley and Terry Brown applied principal component analysis to study the suita bility maps of fifteen suitability overlays . T he original data was eventually reduced into seven dimensions that covered 65% of the original data structure and still present ed the major characteristics of the fifteen suitability overlays (Burley & Brown, Constructing Interpretable Environments from Multidimensional Data: GIS Suitability Overlays and Principal Component Analysis, 1995) the complexity of combining suitability maps (Burley & Brown, Constructing Interpretable Environments from Multidimensional Data: GIS Suitability Overlays and Principal Component Analysis, 1995) . P rincipal component analysis can simplify the o riginal set of variables into smaller sets of the uncorrelated variables that correspond to the dis t ing uishable character istics and 57 combine the correlated variables as one varia nce because they are express ing the same character istics. Therefore, the output of PCA would represent most of the meaningful variance of the sample data and make the cluster analysis more efficient by reducing the insignificant dimensions (Xu Y. , 2015) . After inputting the data into SAS (version: 9.4 TS Level 1M2, X64_8PRO platform, English , Copyright© 2002 - 2012 by SAS Institute Inc., Cary, NC, USA. ) , a set of eigenvalues , corresponding to the proportion of eigenvalues, means, standard deviations and eigenvectors , which correspo nd to the variables , is created for PCA . Means and standard deviations can be used to compute the standard scores of the correspond ing variables. E ach eigenvalue represents a dimension of the sample data ; the eigenvalue that is greater than 1.0 would be defined as the representation of a meaningful dimension and used as a distinguishable principal component in the cluster analysis. The proportion of each eigenvalue among the whole data set represent s its level of significance in the variance of sample data (Xu Y. , 2015) . Usually, cumulative eigenvalues that cover more than sixty to eighty percent of sample data are preferred for further study. In this study, the principal component that correspond s to the highest eigenvalue is defined as the first principal component, the second principal component that correspond s to the highest eigenvalue is defined as the second principal component, and so on. When the meaningful principal components are separate d from the original set of principal components, it is typical to use the corresponding eigenv alue s to generate the eigenv ector coefficients . In the output of SAS (version: 9.4 TS Level 1M2, X64_8PRO platform, English , Copyright© 2002 - 2012 by SAS Institute Inc., Cary, NC, USA. ) , e igenv alue s can be used to ector coefficients can be used to reveal the degree of mutual relationship s between the variables 58 and principal components . The variable s are defined as more similar in terms of items when the number s of correspond ing eigenv alue coefficient s are closer (Xu Y. , 2015) . In this research, cluster analysis is conducted using the equation s adopted from study of Chinese garden comparison (Xu Y. , 2015) . The first step includes standardiz ing the variables using the equation shown below (Equation 1) (Xu Y. , 2015) : Standard score of a variable Where: X = Each Value of Variable = Mean of the Variable SD = Standard Deviation of the Variable In this equation, m ean is the average number of the whole sample data set; standard deviation is the number indicate d the sample data that tend to be close to the mean , so that it can quantify the variance of sample data. The second step involves computing the score of each site with standardized variable scores and principal component coefficients correspondi ng to the variables using the equation shown below (Equation 2) (Xu Y. , 2015) : Site Score + Where : = Each Value of Variable = Mean of the Variable = Standard Deviation of the Variable = Each Principal Component Coefficient Finally, the last step includes comput ing the site score of each selected site on a scatter graph in Excel (Microsoft ® Excel for Mac, version: 15.32) , where each site score of 59 one corresponding set of the principal components is located by one axis . For example, in a two - dimensional scatter graph, site scores of one principal component set should be located by the horizontal axis, while site scores of another principal component set should be located by the vertical axis. In this way, site scores of two principal components can be reviewed directly through the ir location on the scatter graph. When the points of the corresponding site scores are clustered together on this scatter graph, according to the definition of the principal components, the se corresponding sites are defined as more similar than other sites . Eigenvector coefficient s can be used to explain the visually detected linear combinations of variables in the scatter graphs ; the highest and lowest eigenvalue coefficients are considered most responsible to the coordinate s of the site scores. Then, the next steps involves address ing the corresponding variables of the highest and lowest eigenvalue coefficients, so that the explanation of the data variance can be suggested using these distinctive variables . Thus, a conclusion of the site comparison can be generated using the cluster analysis to address their similarities or differences in the design elements and principles (Xu Y. , 2015) . 60 CHAPTER 3: RESULTS In this study, the design elements and principles are shown in Table 1 as one hundred and thirty - four variables . N umbers are recorded in Table 2 and Table 3 to describe the site conditions by the previous investigation of the eighteen selected sites , including three class ical Chinese gardens, five modern Chinese gardens, eight traditional Japanese gardens and two shrines . In this case, one present zero absent In this study, four out of the one - hundred and thirty - four variables (design concept, contrast, deep implication, and simplicity) are defined as insignificant variables because they all scored in all eighteen selected sites . In other words, this findin g means that there is n o variance from these variables, and are therefore not considered in PCA. Table 2 : List of design elem ents and princip l e s in the three classical Chinese gardens and five modern Chinese gard ens . Classical Chinese Gardens Modern Chinese Gardens Design elements Humble Administrator's Garden Master of the Nets Garden Lingering Garden Bamboo Garden Net. Wet. Garden Learning Garden Sugar Cane Garden Landscape New Wave The great halls (ting tang) 1 1 1 0 0 0 0 0 Covered stone boat (fang) 1 0 0 0 0 0 0 0 Viewing towers (lou ge) 1 1 1 0 0 0 0 0 Studies (shufang) 1 1 1 0 0 0 0 0 Covered walkways (lang) 1 1 1 0 0 0 0 0 Pavilions (ting xie) 1 1 1 0 0 0 0 0 Viewing terrace 0 0 0 1 0 1 0 0 Black tile pavement 1 1 1 0 0 0 1 0 Brick paving 1 1 1 0 0 0 0 0 Cracked ice stone paving 1 1 1 0 0 1 0 0 Pebbles area 0 0 0 1 0 1 1 0 61 Table 2 (con Mosaic pave with special pattern 1 1 1 0 0 1 0 1 Whitewashed walls 1 1 1 1 0 1 0 0 Grey stone walls 1 0 0 1 0 0 1 0 Openwork brick walls 1 1 1 1 1 1 0 1 Curved top walls 1 0 1 0 0 0 0 0 Zigzag wall 1 1 1 1 0 0 0 0 Meandering walls 1 0 1 0 0 0 0 0 Bamboo paved pathway 0 0 0 0 1 0 0 0 Boardwalk 0 0 0 0 1 0 0 0 Curved Pathway 1 1 1 1 1 0 1 0 Straight Pathway 1 1 1 1 0 1 0 1 Zigzag Bridge 1 1 1 1 0 0 0 0 Semi - circular bridge 1 1 0 0 0 0 0 0 Straight Bridge 1 1 1 1 0 1 0 0 Wall holes with symbolized shape 1 1 1 1 0 0 0 0 Lattice window 1 1 1 0 0 0 0 0 Moon Gate 1 1 1 0 0 0 0 0 Wood carvings 1 1 1 0 0 0 0 0 Glass carvings 0 0 0 0 0 1 0 1 Brick carvings 1 1 1 0 0 0 0 0 Reflecting Pond 1 1 1 1 0 1 1 1 Stream 1 1 0 0 0 0 0 0 Fish pond 1 1 1 0 1 0 0 0 Wetland 0 0 0 1 1 0 0 0 Island 1 0 1 1 0 0 0 0 Artificial mountains 1 1 1 0 0 0 0 0 Sculptural rocks 1 1 1 1 1 1 0 1 Pond bank rocks 1 1 1 0 0 0 0 0 Taihu rocks 1 1 1 0 0 0 0 0 Trees 1 1 1 1 0 1 0 1 Shrubs 1 1 1 0 0 1 0 1 Ground covers 1 1 1 0 0 0 1 1 Turf area 1 0 0 1 0 0 0 1 62 Table 2 (cont Pine 1 1 1 0 0 1 0 0 Bamboo 1 1 1 1 0 1 0 1 Plum 1 1 1 0 0 0 0 0 Magnolias 1 1 1 0 0 0 0 0 Camellia 1 1 1 0 0 0 0 0 Crepe myrtles 1 1 1 0 0 0 0 0 Sweet osmanthus 1 1 1 0 0 1 0 0 Peony 1 1 1 0 0 0 0 0 Willow 1 1 1 0 0 0 0 0 Lotus 1 1 1 0 0 1 0 1 Reed 0 0 0 1 1 0 0 0 Sugar cane 0 0 0 0 0 0 1 0 Moon 1 1 0 0 0 0 0 0 Clouds 0 1 1 0 0 0 0 0 Rain 1 0 1 0 0 0 0 0 Wind 1 1 1 0 0 0 0 0 Shadow 1 0 0 0 1 1 0 0 Originally private 1 1 1 0 0 0 0 0 Public 0 0 0 1 1 1 1 1 Located in suburban 0 0 0 1 1 1 1 1 Located in urban 1 1 1 0 0 0 0 0 Design concept 1 1 1 1 1 1 1 1 Poem and painting concept 1 1 1 0 0 1 0 1 Naturalness 1 1 1 0 1 0 1 1 Varied spaces with visual devices 1 1 1 1 0 1 1 1 Borrowed scenery 1 1 1 0 1 0 0 0 Enframed scenery 1 1 1 1 1 1 0 1 Opposite scenery 1 1 1 1 0 1 0 1 Contrast 1 1 1 1 1 1 1 1 Deep implication 1 1 1 1 1 1 1 1 Abstract geometrical composition 0 0 0 1 1 1 1 1 Miniature of natural landscapes 1 1 1 0 0 0 0 0 63 Table 2 (cont Boating pond 1 0 1 0 0 0 0 0 Stepping stones 0 0 0 0 0 0 0 0 Yellow painted wall 0 0 0 0 0 0 0 0 Unpainted structures or elements 1 1 1 0 0 0 0 0 Fall color of plant material 1 1 1 0 0 0 0 0 Metal components in wooden structures 0 0 0 0 0 0 0 0 Courtyard/atrium 1 1 1 0 0 0 0 0 Wooden /bamboo fence/railings 1 0 0 1 0 1 0 1 Bench 0 0 0 1 0 1 0 1 Pagoda 0 0 0 0 0 0 0 0 Adjacent to burial site 0 0 0 0 0 0 0 0 Stairs 0 0 0 1 0 0 1 1 Geomancy rules/Fengshui 1 1 1 0 0 0 1 0 Stone lanterns 0 0 0 0 0 0 0 0 Wooden lanterns 0 0 0 0 0 0 0 0 Trellis/pergola 1 1 1 0 0 0 0 0 Obviously exposed drainage ditch/gutter 0 0 0 0 0 0 0 0 Signage stone 0 0 0 0 0 0 0 0 Elevation change 1 1 1 0 0 0 0 0 Plaque 1 1 1 0 0 0 0 0 Lifted base of building 1 1 1 0 0 0 0 0 Building partially painted in red 1 1 1 0 0 0 0 0 Strings implied enchantment boundary 0 0 0 0 0 0 0 0 Raked sand design 0 0 0 0 0 0 0 0 Stone hand wash basin 0 0 0 0 0 0 0 0 64 Table 2 Well 1 1 1 0 0 0 0 0 Bell 0 0 0 0 0 0 0 0 Shishiodoshi 0 0 0 0 0 0 0 0 Grouping stones 1 1 1 0 0 0 0 0 Ophiopogon japonicus 1 1 1 0 0 0 0 0 Hydrangea 0 0 0 0 0 0 0 0 Japanese Maple 1 1 1 0 0 0 0 0 Moss 0 0 0 0 0 0 0 0 Azalea 1 1 1 0 0 1 0 1 Trimmed plant material 1 1 1 1 0 1 0 1 Weeping form plant material 1 1 1 0 0 0 0 0 Multiple layers of entry sequence 1 1 1 0 0 0 0 0 Multiple layers of edges 1 1 1 0 0 0 0 0 Control view point 1 1 1 1 0 0 0 1 Foreground, Midground, and Background 1 1 1 0 0 0 0 0 Large scale 0 0 0 0 0 0 0 0 Small scale 1 1 1 1 1 1 1 1 Located on the foot of mountain 0 0 0 0 0 0 0 0 Location adjacent to Shinto Shrine 0 0 0 0 0 0 0 0 Curved roof 1 1 1 0 0 0 0 0 Chinese characters 1 1 1 0 0 0 0 0 Wabi - sabi 0 0 0 0 0 0 0 0 Religious property 0 0 0 0 0 0 0 0 Tea 1 1 1 0 0 0 0 0 Sense of sacred place 1 1 1 0 0 0 0 0 Animistic believes 1 1 1 0 0 0 0 0 Buddhism 1 0 1 0 0 0 0 0 Shinto 0 0 0 0 0 0 0 0 Confucianism 1 1 1 0 0 0 0 0 65 Table 2 (cont Imperial background 0 0 0 0 0 0 0 0 Simplicity 1 1 1 1 1 1 1 1 Tranquility 1 1 1 0 0 0 0 0 Harmony 1 1 1 0 0 0 0 0 Table 3 : List of design elements and princ iple s in the eight traditional Japanese gardens and two shrines . Traditional Japanese Garden Shrines Design elements Daitoku - ji Ryoan - ji Kinkaku - ji Tenryu - ji Shisen - do Kiyomizu - dera Daikaku - ji Byodo - in Yasaka Shrine Fushimi - Inari Taisha The great halls (ting tang) 1 1 1 1 1 1 1 1 1 1 Covered stone boat (fang) 0 0 0 0 0 0 0 0 0 0 Viewing towers (lou ge) 1 1 1 1 1 1 1 1 1 1 Studies (shufang) 1 1 1 1 1 1 1 1 1 1 Covered walkways (lang) 1 1 1 1 0 1 1 1 1 1 Pavilions (ting xie) 1 1 1 1 0 1 1 1 1 1 Viewing terrace 1 1 1 1 1 1 1 1 1 1 Black tile pavement 0 0 0 0 0 0 0 0 0 0 Brick paving 1 0 0 0 0 0 0 0 0 0 Cracked ice stone paving 1 1 0 1 1 1 0 0 0 0 Pebbles area 1 1 1 1 1 1 1 1 1 1 Mosaic pave with special pattern 0 0 0 0 0 0 0 0 0 0 Whitewashed walls 1 1 1 1 0 1 1 1 1 1 Grey stone walls 1 0 0 0 1 1 0 0 1 1 Openwork brick walls 0 0 0 0 0 0 0 0 0 0 Curved top walls 0 0 0 0 0 0 0 0 0 0 Zigzag wall 0 0 0 0 0 0 0 0 0 0 Meandering walls 0 0 0 0 0 0 0 0 0 0 Bamboo paved pathway 0 0 0 0 0 0 0 0 0 0 Boardwalk 0 0 0 0 0 0 0 0 0 0 66 Table 3 Curved Pathway 1 1 1 1 1 1 1 1 0 1 Straight Pathway 1 1 1 1 1 1 1 1 1 1 Zigzag Bridge 0 0 0 0 0 0 0 0 0 0 Semi - circular bridge 1 1 0 0 0 0 1 1 0 0 Straight Bridge 1 1 0 1 1 0 0 1 0 1 Wall holes with symbolized shape 0 0 0 0 0 0 0 0 0 0 Lattice window 1 1 1 1 1 1 1 1 1 1 Moon Gate 0 0 0 0 0 0 0 0 0 0 Wood carvings 1 1 1 1 0 1 1 1 1 1 Glass carvings 0 0 0 0 0 0 0 0 0 0 Brick carvings 0 0 0 0 0 0 0 1 0 0 Reflecting Pond 1 1 1 1 1 1 1 1 0 1 Stream 1 1 1 1 1 1 1 1 1 1 Fish pond 1 1 1 1 1 1 1 1 0 1 Wetland 0 0 0 0 0 0 0 0 0 0 Island 1 1 1 1 1 1 1 1 0 0 Artificial mountains 0 0 0 0 0 0 0 0 0 0 Sculptural rocks 1 1 1 1 1 1 1 1 1 1 Pond bank rocks 1 1 1 1 1 1 1 1 0 0 Taihu rocks 0 0 0 0 0 0 0 0 0 0 Trees 1 1 1 1 1 1 1 1 1 1 Shrubs 1 1 1 1 1 1 1 1 1 1 Ground covers 1 1 1 1 1 1 1 1 0 1 Turf area 0 0 0 0 0 0 0 1 0 0 Pine 1 1 1 1 1 1 1 1 1 1 Bamboo 1 1 1 1 1 1 1 1 1 1 Plum 1 1 1 1 1 1 1 1 1 1 Magnolias 1 1 1 1 1 1 1 1 1 1 Camellia 1 1 1 1 1 1 1 1 0 1 Crepe myrtles 1 1 1 1 1 1 1 1 1 1 Sweet osmanthus 1 1 1 1 1 1 1 1 1 1 Peony 0 1 0 1 0 0 0 0 0 0 Willow 0 0 0 0 0 0 0 0 0 0 Lotus 0 1 1 0 0 0 1 1 0 0 Reed 0 1 1 1 1 0 1 0 0 0 Sugar cane 0 0 0 0 0 0 0 0 0 0 Moon 0 0 0 0 1 1 1 0 1 1 Clouds 0 1 0 0 1 1 1 0 1 1 67 Table 3 (con Rain 0 1 0 0 1 1 1 1 0 1 Wind 0 0 0 0 1 0 1 0 0 0 Shadow 1 1 1 1 1 1 1 1 1 1 Originally private 1 1 1 1 1 1 1 1 0 0 Public 0 0 0 0 0 0 0 0 1 1 Located in suburban 0 1 1 1 1 1 1 1 0 1 Located in urban 1 0 0 0 0 0 0 0 1 0 Design concept 1 1 1 1 1 1 1 1 1 1 Poem and painting concept 1 1 1 1 1 1 1 1 0 0 Naturalness 1 1 1 1 1 1 1 1 1 1 Varied spaces with visual devices 1 1 1 1 1 1 1 1 1 1 Borrowed scenery 1 1 1 1 1 1 1 1 1 1 Enframed scenery 1 1 0 1 1 1 1 1 0 1 Opposite scenery 1 1 1 1 1 1 1 1 1 1 Contrast 1 1 1 1 1 1 1 1 1 1 Deep implication 1 1 1 1 1 1 1 1 1 1 Abstract geometrical composition 1 1 1 1 1 1 1 1 1 1 Miniature of natural landscapes 1 1 1 1 1 1 1 1 0 0 Boating pond 0 0 1 0 0 0 1 1 0 1 Stepping stones 1 1 1 1 1 0 1 1 0 0 Yellow painted wall 1 1 1 1 1 0 1 0 0 0 Unpainted structures or elements 1 1 1 1 1 1 1 1 1 1 Fall color of plant material 1 1 1 1 1 1 1 1 1 1 Metal components in wooden structures 1 1 1 1 1 1 1 1 1 1 Courtyard/atrium 1 1 1 1 1 0 1 1 0 1 Wooden /bamboo fence/railings 1 1 1 1 1 1 1 1 1 1 Bench 0 1 1 1 0 0 0 0 0 1 Pagoda 1 1 1 1 1 1 1 1 0 0 68 Table 3 (cont Adjacent to burial site 1 1 1 1 0 1 1 1 1 1 Stairs 1 1 1 1 1 1 1 1 1 1 Geomancy rules/Fengshui 1 1 1 1 1 1 1 1 1 1 Stone lanterns 1 1 1 1 1 1 1 1 1 1 Wooden lanterns 1 1 0 0 0 1 1 0 1 1 Trellis/pergola 0 1 1 1 0 1 0 1 0 0 Obviously exposed drainage ditch/gutter 1 1 1 1 1 1 1 1 1 1 Signage stone 1 1 1 1 1 1 1 1 1 1 Elevation change 1 1 1 1 1 1 1 1 1 1 Plaque 1 1 1 1 1 1 1 1 1 1 Lifted base of building 1 1 1 1 1 1 1 1 1 1 Building partially painted in red 1 0 0 0 0 1 1 1 1 1 Strings implied enchantment boundary 0 0 0 0 0 0 0 0 1 1 Raked sand design 1 1 1 1 1 0 1 0 0 0 Stone hand wash basin 1 1 1 1 1 1 1 1 1 1 Well 1 1 1 1 1 1 1 1 1 1 Bell 1 1 1 1 0 1 1 1 1 1 Shishiodoshi 0 0 0 0 1 0 0 0 0 0 Grouping stones 1 1 1 1 1 1 1 1 0 0 Ophiopogon japonicus 0 0 0 0 0 0 0 1 0 0 Hydrangea 1 1 1 1 1 1 1 1 1 1 Japanese Maple 1 1 1 1 1 1 1 1 1 1 Moss 1 1 1 1 1 1 1 1 0 1 Azalea 1 1 1 1 1 1 1 1 1 1 Trimmed plant material 1 1 1 1 1 1 1 1 1 1 Weeping form plant material 1 1 1 1 0 1 1 1 0 1 Multiple layers of entry sequence 1 1 1 1 1 1 1 1 1 1 Multiple layers of edges 1 1 1 1 1 1 1 1 1 1 Control view point 1 1 1 1 1 1 1 1 1 1 69 Table 3 (cont Foreground, Midground, and Background 1 1 1 1 1 1 1 1 1 1 Large scale 0 1 1 1 0 1 1 1 0 1 Small scale 1 0 0 0 1 0 0 0 1 0 Located on the foot of mountain 0 1 1 1 1 1 1 1 1 1 Location adjacent to Shinto Shrine 1 1 1 1 1 1 1 1 1 1 Curved roof 1 1 1 1 0 1 1 1 1 1 Chinese characters 1 1 1 1 1 1 1 1 1 1 Wabi - sabi 1 1 1 1 1 1 1 1 1 1 Religious property 1 1 1 1 1 1 1 1 1 1 Tea 1 1 1 1 1 0 1 1 0 0 Sense of sacred place 1 1 1 1 1 1 1 1 1 1 Animistic believes 1 1 1 1 1 1 1 1 1 1 Buddhism 1 1 1 1 1 1 1 1 0 0 Shinto 0 1 0 0 0 1 1 0 1 1 Confucianism 0 0 0 0 1 0 0 0 0 0 Imperial background 1 1 1 1 1 1 1 1 1 1 Simplicity 1 1 1 1 1 1 1 1 1 1 Tranquility 1 1 1 1 1 1 1 1 1 1 Harmony 1 1 1 1 1 1 1 1 1 1 To standardize the variables , combine Table 2 and Table 3 as one data set and p ut it into SAS (version: 9.4 TS Level 1M2, X64_8PRO platform, English , Copyright© 2002 - 2012 by SAS Institute Inc., Cary, NC, USA. ) to generate means and standard deviations . The output is listed in Table 4 . Table 4 : Means and standard deviation of the corresponding variables from the SAS software program . Design elements Mean St andar d Dev iation The great halls (ting tang) 0.7222222 0.4608886 Covered stone boat (fang) 0.0555556 0.2357023 Viewing towers (lou ge) 0.7222222 0.4608886 Studies (shufang) 0.7222222 0.4608886 70 Table 4 (cont Covered walkways (lang) 0.6666667 0.4850713 Pavilions (ting xie) 0.6666667 0.4850713 Viewing terrace 0.6666667 0.4850713 Black tile pavement 0.2222222 0.4277926 Brick paving 0.2222222 0.4277926 Cracked ice stone paving 0.5 0.5144958 Pebbles area 0.7222222 0.4608886 Mosaic pave with special pattern 0.2777778 0.4608886 Whitewashed walls 0.7777778 0.4277926 Grey stone walls 0.4444444 0.51131 Openwork brick walls 0.3888889 0.5016313 Curved top walls 0.1111111 0.3233808 Zigzag wall 0.2222222 0.4277926 Meandering walls 0.1111111 0.3233808 Bamboo paved pathway 0.0555556 0.2357023 Boardwalk 0.0555556 0.2357023 Curved Pathway 0.8333333 0.3834825 Straight Pathway 0.8888889 0.3233808 Zigzag Bridge 0.2222222 0.4277926 Semi - circular bridge 0.3333333 0.4850713 Straight Bridge 0.6111111 0.5016313 Wall holes with symbolized shape 0.2222222 0.4277926 Lattice window 0.7222222 0.4608886 Moon Gate 0.1666667 0.3834825 Wood carvings 0.6666667 0.4850713 Glass carvings 0.1111111 0.3233808 Brick carvings 0.2222222 0.4277926 Reflecting Pond 0.8888889 0.3233808 Stream 0.6666667 0.4850713 Fish pond 0.7222222 0.4608886 Wetland 0.1111111 0.3233808 Island 0.6111111 0.5016313 Artificial mountains 0.1666667 0.3834825 Sculptural rocks 0.9444444 0.2357023 Pond bank rocks 0.6111111 0.5016313 0.1666667 0.3834825 Trees 0.8888889 0.3233808 Shrubs 0.8333333 0.3834825 Ground covers 0.7777778 0.4277926 Turf area 0.2222222 0.4277926 Pine 0.7777778 0.4277926 71 Table 4 (cont Bamboo 0.8888889 0.3233808 Plum 0.7222222 0.4608886 Magnolias 0.7222222 0.4608886 Camellia 0.6666667 0.4850713 Crepe myrtles 0.7222222 0.4608886 Sweet osmanthus 0.7777778 0.4277926 Peony 0.2777778 0.4608886 Willow 0.1666667 0.3834825 Lotus 0.5 0.5144958 Reed 0.3888889 0.5016313 Sugar cane 0.0555556 0.2357023 Moon 0.3888889 0.5016313 Clouds 0.4444444 0.51131 Rain 0.4444444 0.51131 Wind 0.2777778 0.4608886 Shadow 0.7222222 0.4608886 Originally private 0.6111111 0.5016313 Public 0.3888889 0.5016313 Located in suburban 0.7222222 0.4608886 Located in urban 0.2777778 0.4608886 Design concept 1 0 Poem and painting concept 0.7222222 0.4608886 Naturalness 0.8888889 0.3233808 Varied spaces with visual devices 0.9444444 0.2357023 Borrowed scenery 0.7777778 0.4277926 Enframed scenery 0.8333333 0.3834825 Opposite scenery 0.8888889 0.3233808 Contrast 1 0 Deep implication 1 0 Abstract geometrical composition 0.8333333 0.3834825 Miniature of natural landscapes 0.6111111 0.5016313 Boating pond 0.3333333 0.4850713 Stepping stones 0.3888889 0.5016313 Yellow painted wall 0.3333333 0.4850713 Unpainted structures or elements 0.7222222 0.4608886 Fall color of plant material 0.7222222 0.4608886 Metal components in wooden structures 0.5555556 0.51131 Courtyard/atrium 0.6111111 0.5016313 Wooden /bamboo fence/railings 0.7777778 0.4277926 Bench 0.3888889 0.5016313 Pagoda 0.4444444 0.51131 72 Table 4 (cont Adjacent to burial site 0.5 0.5144958 Stairs 0.7222222 0.4608886 Geomancy rules/Fengshui 0.7777778 0.4277926 Stone lanterns 0.5555556 0.51131 Wooden lanterns 0.3333333 0.4850713 Trellis/pergola 0.4444444 0.51131 Obviously exposed drainage ditch/gutter 0.5555556 0.51131 Signage stone 0.5555556 0.51131 Elevation change 0.7222222 0.4608886 Plaque 0.7222222 0.4608886 Lifted base of building 0.7222222 0.4608886 Building partially painted in red 0.5 0.5144958 Strings implied enchantment boundary 0.1111111 0.3233808 Raked sand design 0.3333333 0.4850713 Stone hand wash basin 0.5555556 0.51131 Well 0.7222222 0.4608886 Bell 0.5 0.5144958 Shishiodoshi 0.0555556 0.2357023 Grouping stones 0.6111111 0.5016313 Ophiopogon japonicus 0.2222222 0.4277926 Hydrangea 0.5555556 0.51131 Japanese Maple 0.7222222 0.4608886 Moss 0.5 0.5144958 Azalea 0.8333333 0.3834825 Trimmed plant material 0.8888889 0.3233808 Weeping form plant material 0.6111111 0.5016313 Multiple layers of entry sequence 0.7222222 0.4608886 Multiple layers of edges 0.7222222 0.4608886 Control view point 0.8333333 0.3834825 Foreground, Midground, and Background 0.7222222 0.4608886 Large scale 0.3888889 0.5016313 Small scale 0.6111111 0.5016313 Located on the foot of mountain 0.5 0.5144958 Location adjacent to Shinto Shrine 0.5555556 0.51131 Curved roof 0.6666667 0.4850713 Chinese characters 0.7222222 0.4608886 Wabi - sabi 0.5555556 0.51131 Religious property 0.5555556 0.51131 Tea 0.5555556 0.51131 Sense of sacred place 0.7222222 0.4608886 Animistic believes 0.7222222 0.4608886 73 Table 4 ( Buddhism 0.5555556 0.51131 Shinto 0.2777778 0.4608886 Confucianism 0.2222222 0.4277926 Imperial background 0.5555556 0.51131 Simplicity 1 0 Tranquility 0.7222222 0.4608886 Harmony 0.7222222 0.4608886 In this analysis , after the means and standard deviations of all significant variables are calculated, this information is then used to generate principal component eigenvalues , which best re veal s the covariance of the sample data (Xu Y. , 2015) . A s shown in Table 5 , only t he first fif teen principal component eigenvalues are greater than 1.0, which means the y are most useful to explain the variance (Xu Y. , 2015) . The first three principal component eigenvalues cumulatively covered 63.81 percent of the variance in the sample data, where as the first principal component covered 4 3.2 percent of the variance. The fourth to the fif teenth principal components are greater than 1.0 , but their proportions are much smaller than the first three principal components. Therefore , only the first three principal component eigenvalues are selected for the calculating principal component coefficient in this study . Table 5 : Principal Component Analysis eigenvalues of the covariance matrix from the SAS Software Program. Eigenvalue Difference Proportion Cumulative PRIN 1 56.1662056 29.375329 0.432 0.432 PRIN 2 26.7908766 17.7501596 0.2061 0.6381 PRIN 3 9.040717 0.6477713 0.0695 0.7077 PRIN 4 8.3929457 3.4082898 0.0646 0.7722 PRIN 5 4.9846559 0.4160248 0.0383 0.8106 PRIN 6 4.5686311 0.7018803 0.0351 0.8457 PRIN 7 3.8667507 0.8990064 0.0297 0.8755 PRIN 8 2.9677444 0.5141437 0.0228 0.8983 PRIN 9 2.4536007 0.3388601 0.0189 0.9172 PRIN 10 2.1147406 0.2062816 0.0163 0.9334 PRIN 11 1.908459 0.3155289 0.0147 0.9481 PRIN 12 1.5929301 0.1353563 0.0123 0.9604 PRIN 13 1.4575738 0.0209401 0.0112 0.9716 74 Table PRIN 14 1.4366337 0.2932455 0.0111 0.9826 PRIN 15 1.1433882 0.2990471 0.0088 0.9914 PRIN 16 0.8443411 0.5745355 0.0065 0.9979 PRIN 17 0.2698056 0.2698056 0.0021 1 In the output of PCA as shown in Table 6 , none of the corresponding coefficient s of t he principal components is larger than 0.4 or smaller than - 0 .4 , which means the variance of the sample data is determined by many distinguishable variables together, instead of a fe w outstanding distinctive variable s . The variables with the highest and lowest numbers in the principal component coefficients are the most distinguishable variables and w ill be used later for suggesting the meaning of the dimensions in the final scatter graphs (Xu Y. , 2015) . The largest coefficient of principal component 1 is around 0.13, and the smallest coefficient is around - 0.08 . In the second and third principal components, the largest coefficients are around 0.15, and the smallest coefficients are around - 0.15 . The eigenvalue coefficients of the first three principal component analysis will be used later for calculating the site scores of the eighteen selected sites using Equation 2. Table 6 : Principal Component Analysis eigenvalue coefficient for each variable from the SAS Software Program. Design Elements and Principles Prin1 Prin2 Prin3 The great halls (ting tang) 0.130476 0.023033 - 0.03526 Covered stone boat (fang) 0.012793 0.106904 - 0.001937 Viewing towers (lou ge) 0.130476 0.023033 - 0.03526 Studies (shufang) 0.130476 0.023033 - 0.03526 Covered walkways (lang) 0.115325 0.032365 - 0.029995 Pavillions (ting xie) 0.115325 0.032365 - 0.029995 Viewing terrace 0.063079 - 0.133227 0.11619 Black tile pavement - 0.013109 0.160909 - 0.101168 Brick paving 0.031125 0.161275 - 0.025306 Cracked ice stone paving 0.04896 0.082795 0.06076 Pebbles area 0.037908 - 0.148157 0.048379 Mosaic pave with special pattern - 0.030621 0.152181 0.135471 Whitewashed walls 0.079295 0.03028 0.114183 Grey stone walls - 0.000197 - 0.021784 - 0.051829 Openwork brick walls - 0.078724 0.13065 0.081931 Curved top walls 0.016686 0.154538 - 0.014102 75 Table 6 (cont Zigzag wall - 0.009346 0.166475 0.034001 Meandering walls 0.016686 0.154538 - 0.014102 Bamboo paved pathway - 0.058808 - 0.014097 - 0.191336 Boardwalk - 0.058808 - 0.014097 - 0.191336 Curved Pathway 0.051514 0.026468 - 0.169117 Straight Pathway 0.083454 0.02159 0.244795 Zigzag Bridge - 0.009346 0.166475 0.034001 Semi - circular bridge 0.06263 0.057576 - 0.00583 Straight Bridge 0.036083 0.072854 0.11728 Wall holes with symbolized shape - 0.009346 0.166475 0.034001 Lattice window 0.130476 0.023033 - 0.03526 Moon Gate 0.019605 0.189043 - 0.024031 Wood carvings 0.115325 0.032365 - 0.029995 Glass carvings - 0.066891 - 0.007286 0.221572 Brick carvings 0.031835 0.162186 - 0.014781 Reflecting Pond 0.03706 0.034377 0.160483 Stream 0.119064 - 0.029195 - 0.025043 Fish pond 0.09633 0.032734 - 0.11836 Wetland - 0.078476 - 0.014226 - 0.065983 Island 0.080989 0.005885 0.042502 Artificial mountains 0.019605 0.189043 - 0.024031 Sculptural rocks 0.05569 0.015524 0.144519 Pond bank rocks 0.108178 0.053326 - 0.014145 Taihu 0.019605 0.189043 - 0.024031 Trees 0.083454 0.02159 0.244795 Shrubs 0.100406 0.021538 0.144469 Ground covers 0.079486 0.032404 - 0.02688 Turf area - 0.031626 0.046562 0.136078 Pine 0.11602 0.021383 0.054663 Bamboo 0.083454 0.02159 0.244795 Plum 0.130476 0.023033 - 0.03526 Magnolias 0.130476 0.023033 - 0.03526 Camellia 0.120103 0.037952 - 0.019487 Crepe myrtles 0.130476 0.023033 - 0.03526 Sweet osmanthus 0.11602 0.021383 0.054663 Peony 0.043471 0.12873 - 0.003882 Willow 0.019605 0.189043 - 0.024031 Lotus 0.021541 0.089944 0.131182 Reed 0.007588 - 0.073862 - 0.03434 Sugar cane - 0.05569 - 0.015524 - 0.144519 Moon 0.055001 0.026302 - 0.043892 Clouds 0.065671 0.012178 - 0.042809 Rain 0.075677 0.02657 - 0.015644 Wind 0.039999 0.131885 - 0.029707 Shadow 0.067843 - 0.089983 - 0.028109 Originally private 0.108178 0.053326 - 0.014145 76 Table 6 (cont Public - 0.108178 - 0.053326 0.014145 Located in suburban - 0.032962 - 0.132783 0.03824 Located in urban 0.032962 0.132783 - 0.03824 Design concept 0 0 0 Poem and painting concept 0.070807 0.052928 0.14007 Naturalness 0.06809 0.00849 - 0.196042 Varied spaces with visual devices 0.058808 0.014097 0.191336 Borrowed scenery 0.108169 0.017047 - 0.143409 Enframed scenery 0.014341 0.049622 0.103526 Opposite scenery 0.083454 0.02159 0.244795 Contrast 0 0 0 Deep implication 0 0 0 Abstract geometrical composition - 0.019605 - 0.189043 0.024031 Miniature of natural landscapes 0.108178 0.053326 - 0.014145 Boating pond 0.057645 0.050134 - 0.011622 Stepping stones 0.081895 - 0.077878 0.010753 Yellow painted wall 0.072115 - 0.07412 0.005157 Unpainted structures or elements 0.130476 0.023033 - 0.03526 Fall colour of plant material 0.130476 0.023033 - 0.03526 Metal components in wooden structures 0.102906 - 0.121021 - 0.01376 Courtyard/atrium 0.104842 0.050013 - 0.012316 Wooden /bamboo fence/railings 0.052559 - 0.094241 0.205522 Bench - 0.021705 - 0.065217 0.202628 Pagoda 0.091426 - 0.089466 0.004146 Adjacent to burial site 0.094117 - 0.110391 - 0.010368 Stairs 0.03655 - 0.146899 0.031848 Geomancy rules/Fengshui 0.109887 0.016261 - 0.117615 Stone lanterns 0.102906 - 0.121021 - 0.01376 Wooden lanterns 0.063244 - 0.081409 - 0.026879 Trellis/pergola 0.073442 0.082069 - 0.001974 Obviously exposed drainage ditch/gutter 0.102906 - 0.121021 - 0.01376 Signage stone 0.102906 - 0.121021 - 0.01376 Elevation change 0.130476 0.023033 - 0.03526 Plaque 0.130476 0.023033 - 0.03526 Lifted base of building 0.130476 0.023033 - 0.03526 Building partially painted in red 0.073225 0.070826 - 0.044972 Strings implied enchantment boundary 0.01815 - 0.049893 - 0.028311 Raked sand design 0.072115 - 0.07412 0.005157 Stone hand wash basin 0.102906 - 0.121021 - 0.01376 Well 0.130476 0.023033 - 0.03526 Bell 0.094117 - 0.110391 - 0.010368 Shishiodoshi 0.017794 - 0.021568 - 0.007219 Grouping stones 0.108178 0.053326 - 0.014145 Ophiopogon japonicus 0.031835 0.162186 - 0.014781 Hydrangea 0.102906 - 0.121021 - 0.01376 Japanese Maple 0.130476 0.023033 - 0.03526 77 Table 6 Moss 0.098621 - 0.105123 - 0.000461 Azalea 0.100406 0.021538 0.144469 Trimmed plant material 0.083454 0.02159 0.244795 Weeping form plant material 0.107777 0.046833 - 0.015451 Multiple layers of entry sequence 0.130476 0.023033 - 0.03526 Multiple layers of edges 0.130476 0.023033 - 0.03526 Control view point 0.097762 0.022034 0.103072 Foreground, Midground, and Background 0.130476 0.023033 - 0.03526 Large scale 0.081233 - 0.090702 0.006129 Small scale - 0.081233 0.090702 - 0.006129 Located on the foot of mountain 0.091001 - 0.113464 - 0.010545 Location adjacent to Shinto Shrine 0.102906 - 0.121021 - 0.01376 Curved roof 0.115325 0.032365 - 0.029995 Chinese characters 0.130476 0.023033 - 0.03526 Wabi - sabi 0.102906 - 0.121021 - 0.01376 Religious property 0.102906 - 0.121021 - 0.01376 Tea 0.095049 0.065378 - 0.007474 Sense of sacred place 0.130476 0.023033 - 0.03526 Animistic believes 0.130476 0.023033 - 0.03526 Buddhism 0.10198 0.008273 - 0.004773 Shinto 0.053985 - 0.078082 - 0.024796 Confucianism 0.027379 0.157579 - 0.025519 Imperial background 0.102906 - 0.121021 - 0.01376 Simplicity 0 0 0 Tranquility 0.130476 0.023033 - 0.03526 Harmony 0.130476 0.023033 - 0.03526 ite scores should be calculated with the variable values (from Table 2 and Table 3 ) , means (from Table 4 ) , standard deviations (from Table 4 ) and principal component coefficients (from Table 6 ) following E quation 2 (Xu Y. , 2015) . F or example, the S ite S core of Humble Garden in the first principal component is calculated as below (Equation 2) : Site Score Therefore, a table of site scores corresponding to the first three principal components is generated a nd shown in Table 7 . 78 Table 7 : Site scores in the first three principal components. Name of the Study Sites PRIN 1 PRIN 2 PRIN 3 Humble Administrator's Garden 2.878997851 11.47605414 - 0.070171415 Master of the Nets Garden 2.026469535 10.25661709 - 0.715462914 Lingering Garden 2.27319477 11.28467025 - 0.630696698 Bamboo Garden - 10.99618606 - 0.581932239 3.651821376 Net. Wet. Garden - 13.23501174 - 1.513331994 - 6.931290784 Learning Garden - 10.02816392 - 0.668559031 6.091662956 Sugar Cane Garden - 12.53322977 - 1.666472273 - 5.235318345 Landscape New Wave - 10.62584938 - 0.404545683 4.920699942 Daitoku - ji 5.534979673 - 1.595120163 - 0.247457737 Ryoan - ji 6.323252889 - 2.981789645 0.580402245 Kinkaku - ji 5.49021076 - 3.450533706 0.178561372 Tenryu - ji 5.628313554 - 3.013841964 0.560982113 Shisen - do 4.004667988 - 2.315315832 - 0.261490675 Kiyomizu - dera 5.410253068 - 3.041815066 - 0.503209831 Daikaku - ji 6.419067306 - 3.018152569 - 0.426469414 Byodo - in 5.824693684 - 1.417633315 0.444542243 Yasaka Shrine 1.791856137 - 3.549740329 - 1.044909662 Fushimi - Inari Taisha 3.812484892 - 3.798556125 - 0.362192459 To obtain a more sensitive result from the principal components, a three - dimensional model is built using three two - dimensional charts with each two of the three selected principal components. After all site scores of the first three principal components are calculated, these scores are arranged in three groups name d after the order of the principal components: the first and the second principal component; the first and the third principal component; the second and the third principal component. The site scores are located on Excel scatter graphs using the following approach: in the first group, I use the first principal component to locate the sites on the horizontal axis and the second principal component to locate the sites on the vertical axis, and I apply the same process to the other two groups. In this way, three scatter graphs are generated for the cluster an alysis as shown in Figure 34 , Figure 35 , and Figure 36 . 79 Table 8 : Corresponding initials of the selected sites as shown in scatter graphs. Name of the Study Sites Initials in Scatter Graphs Humble Administrator's Garden h asg Master of the Nets Garden m ast Lingering Garden l ing Bamboo Garden bamb Net. Wet. Garden n etw Learning Garden l ear Sugar Cane Garden suga Landscape New Wave land Daitoku - ji d ait Ryoan - ji ryoa Kinkaku - ji kink Tenryu - ji tenr Shisen - do s his Kiyomizu - dera kiyo Daikaku - ji daik Byodo - in byod Yasaka Shrine yasa Fushimi - Inari Taisha fush Figure 34 : A scatter graph of the correlation between the eighteen selected sites based on the site scores of Principal Component 1 and Principal Component 2. 80 Figure 35 : A scatter graph of the correlation between the eighteen selected sites based on the site scores of Principal Component 1 and Principal Component 3. Figure 36 : A scatter graph of the correlation between the eighteen selected sites based on the site scores of Principal Component 2 and Principal Component 3 . 81 CHAPTER 4: DISCUSSION 4.1 Cluster Analysis Method in Garden Comparison This analysis employs the s ame statistical method Cluster Analysis , which is used for site evaluation . Prior research has suggested that Cluster Analysis is a useful potential quantitative study method to compare gardens (Xu Y. , 2015) . The sample personal understandings and experience of the design principles a nd elements in oriental gardens. The collected data is then put into the statistical software program SAS (version: 9.4 TS Level 1M2, X64_8PRO platform, English , Copyright© 2002 - 2012 by SAS Institute Inc., Cary, NC, USA. ) , which provide d output data for ge nerating site scores that represent the dimensions of each site on each corresponding principal component . After the site scores are , Excel (Microsoft ® Excel for Mac, version: 15.32) is used for arranging the site scores into two - dimensional scatter graphs for final cluster analysis (Xu Y. , 2015) . 4.2 Comparison of Gardens and Garden Elements Scatter graphs of site scores correspond ing to the principal component eigenvalue coefficients can be used for identify ing clusters visually (Xu Y. , 2015) . After the three scatter graphs are generated by the site scores of the corresponding variables, the meanings of th e axes are suggested using the most distinctive variables, which are correlated to the eigenvalue coefficients of the first three principal components as shown in Table 6 . Table 9 , Table 10 and Table 11 list the most iconic variables corresponding to the design elements and principles based on the principal component scores of each variable. The meaning of the positive and negative parts of the axes in the three sca tter graphs are suggested by the 82 variables of the highest and lowest numbers in the principal component eigenvalue coefficients. As analyzed in the previous chapter, the most distinguishable design elements are determined by the largest and smallest coefficients of principal component s . In principal component 1, the distinguishable coefficients are around 0.13 and - 0.08; in principal component 2 and principal component 3, they are around 0.15 and - 0.15. The corresponding design e lements of the distinguishable coefficients are listed below in Table 9 , Table 10 , and Table 11 . Since the first two principal components cover 63.81 percent of the variance, thi s research focuse s on the site scores and eigenvalue coefficients of principal component 1 and principal component 2 for further analysis. According to Figure 34 , by analyzing the site scores of the first principal component on the horizontal axis and the site scores of the second principal component on the vertical axis, the selected sites are located on the scatter graph and the relationship between these gardens is reveal ed. Table 9 : List of the design elements corresponding to the largest and the smallest coefficient s of principal component 1. The design elements of largest coefficients The de sign elements of small est coefficients The great halls (ting tang) Openwork brick walls Viewing towers (lou ge) Wetland Studies (shufang) Public Lattice window Small scale Plum Magnolias Crepe myrtles Unpainted structures or elements Fall color of plant material Elevation change Plaque Lifted base of building Well Japanese Maple Multiple layers of entry sequence Multiple layers of edges 83 Table 9 (cont Foreground, Midground, and Background Chinese characters Sense of sacred place Animistic believes Tranquility Harmony Table 10 : List of the design elements corresponding to the largest and the smallest coefficient s of principal component 2. The design elements of largest coefficients The design elements of small est coefficients B lacktile pavement Viewing terrace Brick paving Pebbles area Mosaic pave with special pattern Located in suburban Openwork brick walls Abstract geometrical composition Curved top walls Metal components in wooden structures Zigzag wall Adjacent to burial site Meandering walls Stairs Zigzag Bridge Stone lanterns Wall holes with symbolized shape Obviou s ly exposed drainage ditch/gutter Moon Gate Signage stone Brick carvings Stone hand wash basin Artificial mountains Hydrangea Moss Willow Located on the foot of mountain Wind Location adjacent to Shinto Shrine Located in urban Wabi - sabi Ophiopogon japonicas Religious property Confucianism Imperial background Table 11 : List of the design elements corresponding to the largest and the smallest coefficient s of principal component 3. The design elements of largest coefficients The design elements of small est coefficients Mosaic pave with special pattern Black tile pavement Straight Pathway Bamboo paved pathway Glass carvings Boardwalk Reflecting Pond Curved Pathway Sculptural rocks Fish pond Trees Sugar cane 84 Table 11 (cont Shrubs Naturalness Turf area Borrowed scenery Bamboo Geomancy rules/Fengshui Lotus Poem and painting concept Varied spaces with visual devices Opposite scenery Wooden /bamboo fence/railings Bench Azalea Trimmed plant material Japan and China share many characteristics in terms of traditi onal landscape design. In 2003, research on the images of Japanese gardens and Chinese gardens analyzed the - through survey s and scientific method s (Zhao, Matsumoto, Liu, Yuan, & Kawata, 2003) . The study revealed that there are some specific garden design elements that - - water elements , - hous e, bower, stone elements , and shape features are thought to be the main factors of - (Zhao, Matsumoto, Liu, Yuan, & Kawata, 2003) . According to the study , the most common impression s among Japanese p eople of Japanese gardens are purity and stillness , while the primary character istics of Chinese gardens are thought to be mixture and dynami sm . In contrast, the primary impressions among Chinese people of Japanese gardens is include primary characteristics of Chinese gardens are thought to include artificialness and complex ity (Zhao, Matsumoto, Liu, Yuan, & Kawata, 2003) . In the cluster analysis performed in the aforementioned study , the distinguishable variables of principal component 2 suggest that in terms of garden design, the are more representative of - are more representative of - 85 I n the cluster analysis in this analysis , several design elements and principles are addressed as factors of the garden partitioning. In the first scatter graph (Figure 34) , classical Chinese gardens and traditional Japanese gardens share common character istics used to locate the pos itive dimension of principal component 1. According to the distinguishable variables listed in Table 9 , traditional Chinese garden settings, the favor of plant selection , tradit ional Chinese culture, and the sense of strong influences on both classical Chinese gardens and traditional Japanese gardens. I n Table 9 , i t is suggested that traditional Chinese and Japanese gardens are closely related t o life consumption , amenity and buildings, while modern Chinese gardens typically function as public space s . H owever, the first scatter graph ( Figure 34 ) illustrates that traditional Japanese gardens and modern Chinese gardens are located in the negative dimension because of the low score s in principal component 2, where the are considered as the major fac tor s ( wh refers to a void, gravel, sandy rocky space ) . The five modern Chinese gardens are similar in terms of statistics. Instead of a scatter ed dispersal in terms of location on the graph , the scores of these gardens are very close in my research and are graphically different from the (Xu Y. , 2015) . The primary reason for the difference in results might be that Xu generated a variable list of the design elements that best describe the unique feature s of modern Chinese gardens . After the other fifty - nine variables of traditional Japanese garden features are added , according to the site condition data recorded in Table 2 , it might be that of the latter fifty - nine variables makes the modern Chinese gardens scores similar , because they are lacking of many traditional Japanese garden design elements . In the first scatter graph ( Figure 34 ) , modern Chinese gardens are more similar to the traditional Japanese gardens when compar ed to the classical Chinese gardens, because both 86 groups of gardens are locat ed in the negative dimension of principal co mponent 2 . As of international culture exchange (Xu Y. , 2015) . In the 1920s, F rank Lloyd Wright read a book about Dao (Taoism) and the tea ceremony, named The Book of Tea, which was a gift from a Japanese ambassador (Nute, 2000) . Wright was amazed by and he learned about the - Laozi (Thompson, 2017) . This book deeply influenced the design concerns of Wright, and his design style greatly shifted . Formlessness, naturalness, implication, asymmetry, and al insights after the 1920s (Nute, 2000) . From the result s in the third scatter graph , the l iterature review in (Xu Y. , 2015) , and the historical significance of Frank Lloyd Wright and his designs , prior research has suggested that under the universal cultural exchange, modern Chinese gardens exhibi t the character istics of western modernism (Nute, 2000) . Moreover, t he geometrical concepts , including rectangles and curve - linear lines , are widely applied to contemporary landscape designs (Burley, The Design Concept: Intellectual Landscapes in Michigan, 2006) . Modern architecture and landscape design er s have turned away from the Beaux Arts traditions, and instead have searched for the possibilities in new patter ns and materials (Burley, The Design Concept: Intellectual Landscapes in Michigan, 2006) . Concepts are integrated i nto garden designs; with the development of design concepts, designers are creating space s that are greater than a collection of unrelated shape s and forms (Burley, The Design Concept: Intellectual Landscapes in Michigan, 2006) . For example, landscape designs from the old period includ e Bom Jesus do Monte in Portugal, Xiaoling Tomb in China, Vaux - le - Vicomte in France, Villa Lante in Italy, Tapada das Necessidades in Portugal, Stourhead and Stowe in United Kingdom . A ll of these landscape designs are valued as experiential projects , which means that visitors must walk through the 87 sites to discover all of the design features . Designers believe that each of these projects have appli ed different concepts as if applying philosophical guidance to generate the detail s of each design (Burley & Loures, Conceptual Precedent: Seven Landscape Architectural Historic Sites Revisited, 2010) . Using the definition s of the highest and lowest values of the eigenvalue coefficients , and according to the distinguishabl e variables listed in Table 9 and Table 10 , the dimension of principal component 1 is features and modern Chinese garden . In contrast, the dimens ion of principal component 2 is Classica l Chinese garden desig n elements, Traditional Japanese garden design elements According to Figure 34 , eighteen selected sites are clustered into three groups, where principal component 1 can be used to indicate the difference between the group of traditi onal oriental gardens and modern Chinese gardens; while principal component 2 can be used to separate traditional Japanese gardens and classical Chinese gardens. According to the definitions of principal component 1 and principal component 2, the distinguishable design elements listed in Table 9 and Table 10 can be used as a proposed garden design element list of authentic oriental garden styles. However, when referring to the distinctive variables of principal component 3 in Table 11 and the third scatter graph ( Figure 36 ), there is no clear indication of garden design prefer ence . Since the site scores of the first two principal components clearly illustrat e a pattern of clusters, this research will primarily discuss the selected sites using the fir st two principal components. In Figure 34 , three clusters of sites are suggested. On the horizontal axis, - ji, Byodo - in, Shisen - do, Kinkaku - ji, Ryoan - ji, Daikaku - ji, Kiyomizu - dera, Tenryu - ji, Fushimi - Inari - Taisha, and Yasaka Shrine are locat ed in the positive dimension, which are all classical Chinese gardens and traditional Japanese gardens . Bamboo Garden, Learning Garden, 88 Net.Wet. Garden, Sugar Cane Garden , and Landscape N ew Wave Garden are locat ed in the negative dimension, which are all modern Chinese gardens. Thus, the differences between modern and traditional garden s can be analyzed using principal component 1, which refers to sign elements, oriental garden plant materials, Chinese culture, sense of sacred space , As discussed , the variables that ha ve the highest and lowest values of the eigenvalue coefficients are considered distinguishable variables. Thus, the distinguishable variables of principal component 1 are listed in Table 9 ; the distinguishable variables of principal component 2 are listed in Table 10 . The se varia bles are the major factors for the sites scoring high or low. As a result, the variables separate the sites into positive and negative dimensions. In the first scatter graph (Figure 34) , the thr ee classical Chinese gardens have very similar scores ; th e five modern Chinese gardens are also very close to each other ; and the scores of the eight traditional Japanese gardens and two shrines are also similar . These results suggest that the design elements and principles of classical Chinese gardens lack variance , which means they are very similar. The selected classical Chinese gardens follow the same conventional rules, from the distinguishable architectural style to the philosophic guidance , as previously described (Xu Y. , 2015) . Similar to the classical Chinese gardens , the traditional Japanese gardens and the modern Chinese garden s have also evolved their branches of conventional styles base d on the cluster s of gardens in the first result s graph ( Figure 34 ) , because clustering scores imply that the gardens are very alike. Traditional Japanese gardens and shrines e volved their own styles under the influence of the authentic Japanese aesthetics and religious concepts. According to the first scatter graph ( Figure 34 ) and the distinguishable elements of principal component 1 listed in Table 9 , traditional Japanese gardens are similar to the classical Chinese gardens in architectural form and plant selection . H owever, the traditional Japanese gardens differ from the classical 89 Chinese gardens construction features , including black tile pavement, zigzag walls, artificial mountains, and Taihu rocks listed in Table 10 . In addition, the result s demonstrate that the three cluster s of site s are different from one an other. Compar ed to the modern Chinese gardens, the architectur al structures , including halls, viewing towers , and lattice window are emphasized in the traditional Japanese gardens , classical Chinese gardens , and shrines . lack ; there are only simple structures in these gardens to fulfill the functional requirements of dividing space , such as the presence of walls (Xu Y. , 2015) . Moreover, according to Table 9 m odern Chinese gardens tend to be small in scale, open to the public and contain wetland . Possible reasons for the differences in the gardens may include the influence of the small number of sample garden s . T he small number of sample gardens may be due to the fact that the gardens and shrines in the ancient era were only affordable for the upper class, while the modern gardens are functionally designed to be public space s and exhibition s of ecological balance . Modern garden design may be represe ntative of educational value and naturalness in terms of design. Compar ed to the traditional oriental gardens, modern Chinese gardens are mainly created for public use . Modern Chinese gardens typically express the modern thoughts of the natural environment and attempt to get people to enjoy the beauty of wetland, even though wetlands were typically defined as bad place s to build gardens in the ancient era. Based on th e sample data ( Table 2 and Table 3 ), Chinese character istics are very often seen in traditional oriental gardens but are not distinguishable in modern Chinese gardens . T he reason might be that the ancie nt Chinese culture and aesthetics were integrated into Chinese and Japanese garden design during the ancient era, while the modern Chinese 90 gardens were mainly influenced by western culture. Because of the western cultural influence, perhaps the aesthetics of contemporary oriental society no longer focus ed on using poem design concepts to create a sense of nobility ; thus , the Chinese character istics are not seen in modern gardens as much as in traditional oriental gardens. In general, modern Chinese gardens respond to the need s of contemporary society and ecology . Modern Chinese gardens are designed for everyone to step in and enjoy their time in an open space, while providing an opportunity fo r people to come into closer contact with the natural environment. The plant materials also differentiate the selected modern Chinese gardens and traditional oriental gardens. A ccording to Table 9 , p lum ( Prunus sp. Linnaeus. ) , m agnolias ( Magnolia sp. L innaeus . ) , crepe myrtles ( Lagerstroemia sp. Linnaeus. ) , maple ( Acer sp. Linnaeus. ) and the intended assumption of the fall color of plants compose the main view of traditional oriental gardens, which impl ies the pursuit of naturalness in traditional oriental garden design and reflect the preference for plant species in ancient oriental countries. However, a typically use unitary plan t spec ies, and they do not focus on providing pleasant views during the four seasons , but strive to achieve simplicity under the influence of the modern western landscape aesthetics (Xu Y. , 2015) . It is reasonable to use unitary plant species in modern gardens . The contemporary aesthetic trends focus on very simple , but also very strong patterns t o match the architecture style that is d iagonal or asymmetrical . Furthermore, in order to provide a sense of urbanism , unitary plant species and forms are often used in outdoor space s with angular hardscape s . The other distinguishable variables shown in Table 9 reflect the essence of the traditional approach in creating traditional oriental gardens. In the ancient era, p laques are necessary for labeling the estate; well s are essential for coolness in the hot summer. However, modern Chinese gardens no longer serve as landmark s to honor a place; tap water is provided in urban area s . In t raditional oriental gardens, la yering can create a sense of ritual, which 91 makes the garden seem more official and high ranking. A s ense of sacred space is needed becau se people want to build gardens in a safe and comfortable place base d on geomancy rules and religious belie fs . Foreground, midground , and background are visual technique s that blend the artificial area and the natural environment for a sense of layering and richness . P erhaps these visual techniques are used in traditional oriental gardens to provide the best views with limited resources and cost , under the aesthetics of design ing the garden as an ink painting . Animistic belie fs occurred in the design of traditional Chinese and Japanese gardens , which can be used to divide the traditional oriental gardens and modern gardens . According to these distinguishable variables and the literature review, traditional oriental gardens were mainly designed for the mi nority . Traditional oriental gardens were fostered by the Chinese culture and strongly adhered to conventional rules. Traditional oriental gardens focus on the subtle correlation between elements and people. T he sense of tranquility and harmony mostly likely aims to provide people with a peaceful mood to start the day, which could be referred to as religious belie fs and traditional aesthetics. In Figure 35 , the horizontal axis represents principal component 1 and the vertical axis r epresents principal component 3 . T he Bamboo garden, Learning Garden, Landscape New Wave Garden, Sugar Cane Garden, Nets. Wet. Garden are located in the negative d imension of the horizontal axis. According to the distinctive variables listed in Table 9 and Table 11 , the result s suggest that the design elements of openwork brick walls, wetlands, public - use, and small scale have the largest impact on separating the modern Chinese gardens from the traditional oriental gardens. In this graph ( Figure 35 ) , the results suggest that t he five modern Chinese gardens are divided into two dimensions by principal component 3 , while the classical Chinese gardens and traditional Japanese gardens are very similar. In the cluster of the sites with positive scores on the horizontal axis ( Figure 35 ) , the - dera , 92 Daikaku - ji, Daitoku - ji, Byodo - in, Ryoan - ji, Tenryu - ji, Kinkaku - ji, Shisen - do , Yasaka shrine and Fshimi - Inari Taisha are locat ed close to the zero number of the vertical ax is, indicating that they are a ll traditional Japanese gardens, classical Chinese gardens , and shrines . Compar ed to the very high and very low scores of modern Chinese gardens in the vertical axis, the traditio nal oriental gardens and shrines are clustered around the zero point. In Figure 36 , the horizontal axis represents principal component 2 and the vertical axis represents principal component 3. According to the third scatter grap h ( Figure 36 ), Master of Nets dimension of the horizontal axis with very high scores, and they are all classical Chinese gardens. However, principal component 3 cannot be used to generate clear conclusions due to the unclear result from its corresponding distinguishable elements ( Table 11 ) , thus the discussion will not emphasize Figure 35 a nd Figure 36 . B y analyzing the distin guishable variables of principal component 2 show n in Table 10 , this result of the first scatter graph ( Figure 34 ) may imply that classical Chinese gardens can be identified by containing traditional garden elements under Chinese culture including blacktile pavement, brick paving, mosaic pave with special pattern, openbrick walls, zigzag walls, meandering walls, zigzag bridge, wall holes with symbolized shape, moon gate, brick carvings, artificial mountains, Taihu rocks, willow (Salix sp. Linnaeus.) , wind, Ophiopogon japonicas (L.f., Ker Gawl) , and Confucianism. According to Table 10 , p rincipal co mponent 2 also suggest s that traditional Japanese gardens and shrines can be identified with the practice of Zen meditation pebble area s and void gravel paving area s in Shinto shrines . Viewing terrace, metal components in wooden structures, and stairs are the most iconic architectur al design elements in traditional Japanese gardens and shrines . The aesthetics of traditional Japanese gardens and shrines is closer to Wabi - Sabi and abstract geometrical composition. The traditional Japanese gardens are 93 religious properties and tend to be located close to Shinto shrines and the foot of mountains . A ccording to the literature review, these religious properties are related to th e belie fs of the Japanese people that gardens should be set in a sacred place. Thus, the traditional Japanese gardens are also located near burial sites, because people wanted to be buried in sacred places. The fact that s tone hand wash basin appeared in traditional Japanese gardens and shrines may be because of the belief that people need to clean their hands before entering a sacred place. Hydrangea and moss are used very often in traditional Japanese gardens and shrines. Compar ed to the classical Chinese gardens, o ther distinguishable variables of traditional Japanese gardens and shrines are stone lanterns, obviously exposed drain age ditch es /gutter s , signage stone s and stone hand wash basin. The drainage ditch or gutter in classical Chinese gardens are usually hid de n . In classical Chinese gardens, paper lanterns are used under a roof; a plaque with text on it instead of a signage stone is used to label a place. In general, the evolution ary direction of the selected sites include: modern Chines e gardens are strongly related to public space ; traditional Japanese gardens and classical Chinese gardens are connect ed to life consumption and buildings; modern Chinese gardens and traditional Japanese gardens share similarit ies in terms of symbol ism, abstract geometrical composition, and pebble area s ; the design of classical Chinese gardens tend toward the richness of texture, boundaries, and artificial garden elements ; and traditional Japanese gardens tend toward more reductive designs , keep ing only the essentials , which is strong ly related to the implications of sacred space and giving up an o pulen t design style . Nevertheless, if the number of sample gardens is larger to include more sample data of the a hundred and thirty - four variables, the scatter graph and distinguishable elements of each principal component may be different. 94 4.3 Future Implication This research analyzed sev eral Chinese gardens and traditional Japanese gardens with a focus on their cultural context and typical design elements . In addition, this research provided suggestions to help identify Chinese gardens and Japanese gardens using their iconic cultural characteristic values. In addition, the statistic study method, principal c omponent a nalysis , makes the comparison of the se garden s efficient. Future research o n landscap e architecture may use the same study method to help evaluate and understand the design elements and principles of other sites . Moreover , in this research, the inner relationship of classical Chinese gardens, modern Chinese gardens , and traditional Japanese gardens was demonstrated in th e result s of the scientific method ology . Few stud ies have focused on garden correlation using a scientific method . , This research help s to provide the public with an understand ing of ancient Chinese culture and aesthetic principles of Japanese gardens using broader dimensions . Furthermore, t his research can be used as guidance for design elements and principles when planning for traditional Japanese garden s and classical Chinese garden s . To create an authentic traditional Japanese garden , according t o the literature review, one should define the function of the garden as well as the buildings, respect the geomancy rules and the religious ritual s , select the typical Japanese species for plant materials, and understand the spiritual pursu it and aesthetics of Japanese culture. P ersonal understandings of traditional Japanese gardens may vary , but the sense of Japanese authenticity should be created using the distingui shable design elements originat ing from the geological environment and the native culture practices in Japan. 95 4.4 Limitations and Suggestion for Future Research P rincipal component analysis has its limitations . For example, the final results may be different if extra objects are included in the same study ; in p rincipal c omponent a nalysis , the dataset is considered multiv ariate normal , which is hard to illustrate across the dataset . The stability of the results of one study are not definitive, and it takes numerous similar studies to make the results close to definitive (Burley, et al., 2009) . In t his study, the contribution s of the variables and the studied site s of different types are counted e quivalently in terms of the ir weights ; different result s would be generated if the classification and contribution of the gardens is reconsidered based on their types . In addition , the site scores of three classical Chinese gardens are very close in the t hree graphs of the cluster analysis, which may not be representat ive for all classical Chinese gardens. (Xu Y. , 2015) , t he small group of sample gardens might be the main factor towards this output. Thus, m ore classical Chinese gardens can be added in future research to enrich the sample data in principal component analysis , including other famous gardens in Suzhou. Another limitation is that the three - selected classical Chinese gardens are all located in Suzhou ; it would be better to select more garden s in the cities with historical heritages , , for future study to represent the overall style of classical Chinese gardens . Furthermore , following the same steps in (Xu Y. , 2015) , the selection of the variables in garden design is only l understandings and experience s using the literature review and site visit . Therefore, the data source in this research is subjective thinking . To avoid the effect of the author subjectivity , e xtra variables can be added to enhance the sample collection of future garden research based on this study . In addition, s ome variables can be subdivided in detail ; for 96 on. In this way, a more sensitive criteria list would be generated to better evaluate the sites and eventually gain a deeper insight into the various garden styles . In addition, the term plants of P runus genus in this research . S imilarly, in the Bambusoideae subfamily . Plant materials are classified by their general forms in this study . S ince the number of sample sites is not big , the differences between the similar species are ignored . To generate a more sensitive and accurate result, t hese variables can be classifi ed into more items of specific species in a larger data base for future studies . n agreement in terms of variable selection should be considered i n fu ture research (Xu Y. , 2015) . Finally , fu ture studies in the field of landscape architecture and garden design can e valuate other garden s of different cultur al context s in the world with the given equations and statistic al me thod s used in this study . dy of Chinese gardens (Xu Y. , 2015) , new consideration s related to oriental gardens are generated in this research. Furthermore, research based on a scientific method of patients with dementia and Japanese gardens discovered that viewing Japanese gardens has positive effects on lower ing heart rate and enhancing behavioral symptoms (Goto, et al., 2016) , which provides other direction to take further study on traditional Japanese gardens in the future . T raditional Japanese gardens attract people not only through their design , but also by the sense of nature that they produce (Goto, et al., 2016) . In addition, v arious a pproac hes to garden design include not only visual embellishment s , but also include sophisticate d and intended sound technique s to create refreshing sceneries in the traditional Japanese gardens (Fowler, 2015) . In addition to garden viewing, s oundscape and scent are potential areas for future research to gain a deeper understanding of Japanese gardens. Therefore, future studies of site comparison 97 can be conducted by enriching the variable collection and garden samples used in this research. In this study, the feasibility of using principal component analysis to discover the distinguishable garden design elements of classical Chinese gardens, modern Chinese gardens , and traditional Japanese gardens is display ed . Comparing gardens using scientific method is very meaningful to provide data driven support to analyze t he inner relationship of the selected gardens through the distinguishable garden design elements ; the results can be described using the scatter graph, where the distinguishable design elements are the major factor s mak ing these gardens similar and different. According to the discussion of the results, garden styles can be separated using a scientific method . The evolution directions of different garden styles can also be suggested using the statistical results . P rincipal component analysis is considered an effective and efficient methodology in this study of garden comparison. C omp aring the spread of Japanese gardens and Chinese gardens, the promotion of traditional garden design helped to spread native cultural export in the global exchange; gardens could serve as a useful art product of landscape to improve the in ternational influence of a country. 98 BIBLIOGRAPHY 99 BIBLIOGRAPHY Brinkley, F., & Kikuchi, D. (1915). A History of the Japanese People from the Earliest Times to the End of the Meiji Era. New York: Encyclopædia Britannica Co. Bullen, R. (2016, August 25). Chinese sources in the japanese tea garden. Studies in the History of Gardens & Designed Landscapes , 5 - 16. Burley, J. B. (2006). The Design Concept: Intellectual Landscapes in Michigan. The Michigan Landscape, 49 (12), 33 - 40. B urley, J. B., & Brown, T. J. (1995, December). Constructing Interpretable Environments from Multidimensional Data: GIS Suitability Overlays and Principal Component Analysis. 38 (4), p. 536. Burley, J. B., & Loures, L. (2010, November). Conceptual Precedent: Seven Landscape Architectural Historic Sites Revisited. WSEAS Transactions on Environment and Development, 6 (12), 783 - 792. Burley, J. B., & Machemer, P. (2016). From eye to heart: Exterior spaces explored and explained. United States of America: Cognella academic publishing. Burley, J. B., Singhal, V. B., Burley, C. J., Fasser, D., Churchward, C., Hellekson, D., & Raharizafy, I. (2009, July 15). Citation Analysis of Transportation Research Literature: A Multi - Dimensional Map of the Roadside Universe. Lands cape Research, 34 (4), 481 - 495. Cali, J., & Dougill, J. (2013). Shinto shrines : A guide to the sacred sites of japan's ancient religion. Fowler, M. (2015). Sound as a considered design param eter in the Japanese garden. Studies in the History of Gardens & Designed Landscapes, 35 (4), 312 - 327. Gavin, M. (2000, January). Nihon fukeiron (Japanese Landscape): nationalistic or imperialistic? Japan Forum, 12 (2), 219 - 231. Goto, S., Gianfagia, T. J., M unafo, J. P., Fujii, E., Shen, X., Sun, M., . . . Herrup, K. (2016). The Power of Traditional Design Techniques: The Effects of Viewing a Japanese Garden on Individuals With Cognitive Impairment. HERD: Health Environments Research & Design Journal, 10 (4), 74 - 86. Goto, S., Ristovska, L., & Fujii, E. (2014, October 27). The Japanese garden at Sonnenberg: the first traditional private Japanese garden in North America. Studies in the History of Gardens & Designed Landscapes, 34 (4), 295 - 308. Gregory Kenneth Miss ingham Faculty of Architecture. (2007). Japan 10±, China 1: A first attempt at explaining the numerical discrepancy between Japanese - style gardens outside Japan and Chinese - style gardens outside China. Landscape Research, 32 (2), 117 - 146. Juniper, A. (2003) . Wabi Sabi: the Japanese art of impermanence. Tuttle Publishing. 100 Kalland, A. (1996). Geomancy and town planning in a Japanese community. Ethnology, 35 (1), 16. Liu, C., Burley, J. B., & Partin, S. (2014, November). A metaphor for design: Fishing and Seclusion - the master - of - nets garden, Suzhou, China. The Michigan Landscapes , 32 - 33. Makowska, B. (2014). Intergrating Architecture and the Created Landscape in Japane se Gardens. International Journal of Arts & Sciences, 7 (5), 597 - 604. Mehta, G. K., & Tada, K. (2012). Japanese Gardens: Tranquility, Simplicity, Harmony. Tuttle Publishing. Nute, K. (2000). Frank Lloyd Wright and Japan: The role of traditional Japanese art and architecture in the work of Frank Lloyd Wright. Routledge. Ogata, T., Li, W., & Yamada, S. (2010). A Mapping and Geographical Analysis of Japanese Gardens in the Kyoto Basin. Japan Review , 213 - 221. Parkes, G. (2005, August 1). Thinking Rocks, Living S tones: Reflections on Chinese Lithophilia. Diogenes, 52 (3), 75 - 87. Sakuae, M., & Reid, D. (2012, August 1). Making Tea in Place: Experiences of Women Engaged in a Japanese Tea Ceremony. Journal of occupational science, 19 (3), 283 - 291. ). Handbook of the old shrines and temples and their treasures in Japan. Tokyo: Sansusha. Shih, H. - Y. (1962, February). Reviewed Work: Tu wan's Stone Catalogue of Cloudy Forest. The Journal of Asian Studies, 21 . Stefan, C., Fora, C., Visoiu, D., Hernea, C., & Constantinescu, A. (2009). Landscape painting in Japanese gardens. Journal of Horticulture, Forestry and Biotechnology, 13 , 195 - 196. Takei, J., & Keane, M. P. (2008). Sakuteiki: Visions of the Japanese Garden . Tuttle Publishing. design, and space. International Communication of Chinese Culture, 4 (1), 5 - 23. Treib, M., & Herman, R. (1993). A Guide to the Gardens of Kyoto. Sh ufunotomo Company, Ltd. Van Tonder, G. J., Lyons, M. J., & Ejima, Y. (2002, September 26). Perception psychology: Visual structure of a Japanese Zen garden. Nature, 419 (6905), 359 - 360. Weiss, A. S. (2010). The limits of metaphor: Ideology and Representatio n in the Zen garden. . Social Analysis: The International Journal of Social and Cultural Practice, 54 (2), 116 - 129. 101 Xu, H. (2017). Cross - Cultural Comparative Study of Burial Sites. Proquest Publishing. Xu, Y. (2015). A cluster analysis comparison of classic al Chinese gardens with modern Chinese gardens. Ann Arbor: Proquest Publishing. Zhao, C. - S., Matsumoto, N., Liu, T. - F., Yuan, Y. - Q., & Kawata, K. (2003). A Study on the Image of Landscape of Japanese and Chinese Gardens. Journal of Asian Architecture and B uilding Engineering, 2 , 69 - 76.