A new perspective on the evolution of “Kawara” roof tiles in Ryukyu: A multidisciplinary non-destructive analysis of roof tile transition at Shuri Castle, Ryukyu Islands, Japan

Hiroaki Aoyama; Kaishi Yamagiwa; Wataru Taira; Takeshi Kon

doi:10.1371/journal.pone.0277560

. 2022 Nov 16;17(11):e0277560. doi: 10.1371/journal.pone.0277560

A new perspective on the evolution of “Kawara” roof tiles in Ryukyu: A multidisciplinary non-destructive analysis of roof tile transition at Shuri Castle, Ryukyu Islands, Japan

Hiroaki Aoyama ^1,^2,^*,^#, Kaishi Yamagiwa ^3,^4,^5,^*,^#, Wataru Taira ^1,^2,^6,^#, Takeshi Kon ¹

Editor: Abbas Farmany⁷

PMCID: PMC9668144 PMID: 36383616

Abstract

A unique historical architecture was created at Shuri Castle (Shuri-jo) in the Ryukyu Islands by its “Kawara” roof tiles. After the 13th and 14th centuries, Kawara tiles were introduced to the Ryukyu Islands from several regions, including China, Korea, and mainland Japan, and evolved shapes and patterns that are unique to this island region. However, the transition of some internal features, such as the chemical components and microstructure, had not been analyzed. This study used a multi-faceted approach for such internal data and non-destructive quantitative methods to propose a new perspective on the evolution of historical Ryukyuan Kawara. We analyzed two styles of Ryukyuan Kawara from the 13th to 15th centuries and found that the material processing and firing conditions of the two styles were very similar, even though it had been suggested that they had different origins. A quantitative analysis of tiles from the 16th to 19th centuries revealed a transition in color tone to red, leading to the modern traditional Ryukyuan tiles; traces of changes in firing conditions were also found along with this transition. Finally, the study revealed that the evolution of Ryukyuan Kawara consisted of changed factors, e.g. surface color, and unchanged factors, e.g. paste density. Previous archaeological studies mainly focused on changing external characteristics, such as form and pattern; however, our analysis showed that the internal features changed, while the elemental composition and paste density remained constant from the appearance of the roof tiles until the 19th century. We propose that this is related to different responses of individual factors to external stressors, such as the social context, which may be common to other archaeological artifacts as well. Our study provides a new perspective on the evolution of Ryukyuan Kawara and presents a different discussion of and methods for the chronological study of material culture.

Introduction

The Ryukyu Islands in the southwestern Japan archipelago have a unique history compared to surrounding regions, such as mainland Japan (e.g., Kyushu, Honshu), Taiwan, and China (Fig 1A). Shuri Castle (Shuri-jo) on Okinawa Island is representative of the architecture of the Ryukyu Islands and their culture and was used as a fort, castle, or center of rule mainly during the Ryukyu Kingdom period (AD 1429–1879). It is now registered as one of the World Heritage Sites of this region [1]. The architecture of Shuri Castle is characterized by “Aka-gawara” (means red roof tile), a material associated with traditional houses in Okinawa (Fig 1B). The Castle burned down multiple times, including in a tragic fire in 2019, and each time it was rebuilt using a large number of tiles. The ancient roof tiles at Shuri Castle therefore embody the history of tile culture in the Ryukyu Islands.

Fig 1 — (A) Location of Shuri Castle (*Shuri-jo*) on Okinawa Island, the main island in the Ryukyu Islands, which are located between mainland Japan (Kyushu, Honshu, etc.), eastern Eurasia (southeastern China), and Taiwan. (B) Exterior view of the Main Hall and two typical types of roof tile: Semi-cylindrical “Maru-gawara” and flat “Hira-gawara” tiles. The Maru-gawara at the eaves are called “Nokimaru-gawara”, and the Hira-gawara at the eaves are called “Nokihira-gawara” (or “Tekisui-gawara”). The various patterns decorating these eave-end tiles often serve as indicators of their age.

Kawara roof tiles in Japan are a building material designed mainly to keep out rain, sunlight, or wind and are generally known to be more durable and fire-resistant than shingle or thatch roofing. In East Asia, ancient roof tiles made of ceramics were notably developed after the Warring States period in China (ca. fifth to third century BC) and were introduced into mainland Japan through the Korean peninsula until AD 538 (in the Kofun period) [2]. Initially, the roof tiles were mainly used for certain Buddhist temples, spread widely throughout various regions of mainland Japan [3], and eventually became a common roofing material for houses in Japan. In the Ryukyu Islands, Kawara was introduced after about the 13th or 14th century AD, when multiple types of roof tiles appeared with several origins assumed to be mainland Japan, the Korean peninsula, or mainland China. The ancient roof tiles, as a typical artifact, were evidence of the prosperous international relations between the Ryukyu Islands and its surrounding countries.

Ryukyuan ancient roof tiles have been broadly classified into three styles, “Kourai”, “Yamato”, and “Minchō”, depending on their external characteristics, such as design and form. This classification is based on chronological order and the presumed origins of culture and technology (S1A Fig in S2 Appendix). The oldest Ryukyuan roof tiles appeared at least as early as the 13th to 14th centuries AD (This period corresponds to the Gusuku period in 10th to 14^th century of Ryukyu Islands) and are presumed to have been used in Gusuku, the large stone-walled castles and fortresses that are representative of this period. The oldest tiles are classified into two styles: Kourai and Yamato. The Kourai style is associated with Korea’s Goryeo Kingdom (AD 918–1392) due to external characteristics, such as design, shape, paddling traces, imprinting, and inscriptions on their surface [4–6]; specifically, they are assumed to be connected to refugees from the failure of the “Sambyeolcho” Rebellion in Goryeo (AD 1273) [7]. The appearance of the Yamato-style tile differs from that of the Kourai; based on external characteristics, such as design, form, and traces, it is presumed to have originated in Japan, possibly Kyushu, between the Kamakura (AD 1185–1333) and Muromachi (AD 1336–1572) periods. Although the two styles overlap chronologically, some scholars consider the Kourai to be older, as the Yamato incorporates techniques that resemble the Kourai traces [6,8,9].

In the 14th and 15th centuries AD, the Okinawa Island was unified into the Ryukyu Kingdom and became part of the Chinese dynasty’s tributary system. A new style of roof tile (Minchō) appeared between the 16th and 17th centuries whose appearance was clearly different from previous styles, and its technical relationship with these earlier types is not well understood. Although the Ryukyu Kingdom was invaded and became a vassal state under the Japanese Satsuma feudal domain after AD 1609, some previous researchers have asserted that the new style of roof tile reflected cultural influences from the Ming dynasty, based on a comparison with roof tiles in the Chinese dynasty, for example, as described in the Chinese historical record “Tiangong Kaiwu” (published in AD 1637) [8,10].

However, there are many aspects of the Ryukyuan Kawara that have not been explained. These tiles, considered to have been introduced from surrounding regions, developed into unique tiles in the Ryukyu Islands, but the background of their uniqueness remains unclear. For example, one characteristic of Ryukyuan tiles is that they transitioned from gray to red without the use of glaze [11], but the reason for this is not known, although some hypotheses have been proposed. In addition, few changes other than in external characteristics have been analyzed. Most previous studies have focused on external features to show the chronology of Ryukyuan roof tiles, but limited consideration has been given to internal factors, such as raw material characteristics, paste condition, or internal structure. In the study of cultural evolution, various functional and stylistic factors influence artifacts, and the evolution of material culture cannot necessarily be discussed based on a single factor. Internal features are important factors when considering changes in the properties and functions of roof tiles and their production. However, as there is no effective method of analyzing tile characteristics, which are difficult to observe or quantify, quantitative or statistical analyses of these features are lacking. For example, the color changes considered characteristic of Ryukyuan tiles have not been subjected to a quantitative analysis, and paste analysis through powder X-ray diffraction or petrographic observation has been limited to a few roof tiles due to the destruction of the materials; therefore, a statistical analysis was not possible.

In this study, we analyzed the internal features of the ancient Ryukyuan roof tiles at Shuri Castle quantitatively and statistically using non-destructive methods, an X-ray fluorescence microscope, a computed tomography (CT) scanner, and a digital image scanner. The results were then compared with the chronological transition of roof tiles based on their external characteristics, and finally, the process of roof tile transition in the Ryukyu Islands, its factors, and the mechanism of evolution were discussed.

Materials and methods

Site information

Shuri Castle is viewed as a special architectural resource due to the large number of tiles of various ages (including the three styles mentioned in this article) that have been excavated there. Although the Castle’s origins remain unclear, some historical and archaeological research has suggested that the main palace was constructed in the mid-15th century (“Ankoku-zan Juka-boku no Kihi”: inscription on stone monument built in 1427). It is presumed to have had a tile roof at this time because of the large number of tiles excavated [12–14]. This main palace was destroyed by fire at least three times (AD 1453, 1660, 1709) (“Kyuyou”, a historical record written in AD 1743–1745). Records show that it was rebuilt with a single roof after the first fire in AD 1453 and that it was tiled when rebuilt after the second fire in AD 1670. [12,15–17] (S1B Fig in S2 Appendix) In addition, records of renovations for damage caused by storms, earthquakes, and deterioration exist, and it is presumed that tiles were produced and replaced each time [15,18]. Previous studies of ancient roof tiles have shown no significant difference in the tiles used at Shuri Castle and other sites [19]. Therefore, the changes in roof tiles identified at the Castle can be considered to generally reflect the historical transition of Ryukyuan roof tiles.

Materials

As a sample, we used a total of 39 ancient and modern roof tiles from Shuri Castle (S1 Table in S1 Appendix) that comprised three ancient styles, including Kourai (KOR) (S2 Fig in S2 Appendix), Yamato (YMO) (S3 Fig in S2 Appendix), and Minchō (MCO) (S4 and S5 Figs in S2 Appendix), as well as modern roof tile samples (MDN) (S6 Fig in S2 Appendix). All the ancient samples were from the Ryukyu University Museum (Fujukan) and had been collected on the old Shuri campus of the University of the Ryukyus, which was on the site of Shuri Castle during 1950–1979. The modern samples were extracted from the debris of the Main Hall, which burned down during the Shuri Castle fire on October 31, 2019; these were associated with the recent reconstruction of Shuri Castle after 1989 and made by modern roof tile manufacturing technology.

Although the ancient roof tiles in our sample were not scientifically chronologically dated (e.g., using radiocarbon dating techniques), they can be classified according to their style and dated based on their external characteristics. We classified all the ancient roof tiles into three basic types: KOR, YMO, and MCO. In addition, based on the chronological classification established by Uehara [6], we separated MCO into two phases: Period I (early 17th century) and Period II, III, and IV (late 17th to mid-19th century). Some scholars have argued that MCO may have appeared in the 16th century [20]; however, only tile types from the early 17th century onward have been definitively established.

Image scanning and color distance analysis

Roof tile images were collected using the scan function of the bizhub C368 multifunction printer (Konica Minolta, Japan). All images were scanned under the same conditions as the Kodak color control patches (Kodak, USA). The scanned images were trimmed to a 1 cm-square region without shadows or patterns. The average color information in two models, RGB and Lab, was quantified from the trimmed images using GIMP 2.10.20 software (http://www.gimp.org/). Traditional reddish colors were used from the DIC Color Guide’s “Traditional Colors of Nippon” series (DIC Graphics Corporation, Tokyo, Japan).

Color distance was measured according to CIEDE2000, which takes into account human perception, as suggested by the International Commission on Illumination [21]. Color distances were calculated using the delta_e_cie2000 function in the Python module (python-colormath) [22].

The color distance data were clustered using the hclust function of the stat package in R. The heatmap was drawn using the heatmap.2 function of the gplots package in R. A multidimensional scaling (MDS) analysis was performed using the cmdscale function of the stat package in R and drawn using the dpyr, tidyverse, and ggpubr packages in R. Box and jitter plots were performed using Past software (ver. 4.03). The Steel-Dwass test was performed using the psDCFlig function of the NSM3 package in R.

X-ray fluorescence microscopic analysis

The analysis in this study improved on our previous study [23,24]. An XGT-7200 XRF microscope (Horiba, Japan) was used to analyze the chemical components of the roof tiles, enabling measurement of the elemental distribution on the tile surfaces using spectral analysis. X-rays were generated at 30 kV/1 mA; further, the mass ratio of elements on a measuring point was calculated using 10-μm irradiation diameters over a preset time of 60 s. We performed this analysis with 231 multipoints set at intervals of 1 mm in 20 mm × 10 mm squares for the surface of each roof tile. The X-ray sources were based on rhodium (Rh); each chemical composition was calculated as an oxide based on fundamental parameter methods, and all the elemental ratios of the spectral analysis were described in terms of percentage data for mass.

The average values were calculated based on the Dirichlet distribution using 231 measuring points, as described in our previous study [24]. A principal component analysis (PCA) was performed using the prcomp function of the stat package in R. Linear discriminant analysis effect size (LEfSe) was performed using the tool within the Galaxy web application [25]. A p-value of < 0.05 was considered significant in both Kruskal–Wallis and pairwise Wilcoxon tests, and a score of ≥ 2.0 was considered the threshold on the logarithmic linear discriminant analysis (LDA) score for discriminative features.

X-ray computed tomography (CT)

X-ray CT images of the tile were acquired using a Nikon XT H225 (Nikon, Tokyo, Japan). As the dimensions and shapes differed greatly for each sample, the radiographic conditions could not be standardized. The radiographic conditions for voltage, current, exposure time, and the use of metal filters were changed while maintaining the brightness distribution of each image within a range of 10,000–60,000. A few samples did not meet the brightness requirements due to shape issues. In those cases, we checked the noise in the 3D reconstructed image and used the data if there appeared to be no major problems. The process for the difference in resolution between samples is described in the supplementary methods in S2 Appendix.

Pores identification, counting, and measurement

CT Pro 5.4 software (Nikon, Tokyo, Japan) was used to perform 3D image reconstruction and image correction (i.e., beam hardening and noise reduction) was used as needed. All pores were identified, counted, and measured, according to standard protocols, by the Porosity/Inclusion Analysis Module in VG Studio MAX 3.4 (Volume Graphics GmbH, Heidelberg, Germany), but no thresholding by defect probability values were used in this process. Using this software, we obtained data on the number, volume, and character values of pores (i.e., defect probability value, compactness, sphericity) and volume of material paste. The software package NYSOL version 2.4.2 (NYSOL Corporation, Osaka, Japan) [26], was used to perform data wrangling, such as unification of detection limits and thresholding defect probability values (see Supplementary Materials and Methods in S2 Appendix).

Eight values for internal microstructure features

The following eight values were calculated to represent the internal microstructural features of each sample: (1) Density-S: weight density of the sample (g/cm³); (2) Density-M: weight density excluding pores (i.e., weight density of material paste); (3) Median-vol: median pore volume (g/cm³); (4) Mode-vol: most frequent pore volume (g/cm³); (5) Dispersion-vol: dispersion of pore volume; (6) RLV: ratio of large pores to total pore volume; (7) Num-density: pore number density (n/cm³); and (8) RLN: ratio of large pores to total number of pores. For details of each value, see the supplementary methods in S2 Appendix.

Consideration of data reliability

We checked whether pore data acquisition and data wrangling were successful. First, violin plots were drawn to determine the distribution of pore volumes for each sample (S7 Fig in S2 Appendix). In MCO020, the distribution (especially the lowest value) appeared to be very different from that of the other samples. This suggested the possibility of outlier generation during CT imaging and data filtering. Therefore, we performed outlier tests (Q test) for Median-vol for each style using Dixon’s method. These were performed in R with the dixon.test function in the {outliers} package. MCO020 (Median-vol: 0.71 mm³) was detected as an outlier in the Minchō style (Q = 0.94, p < 2.2E-16), but there were no significant outliers in other styles. In the case of MCO020, its physical features (it was very large and had a complex shape) and analytical factors, such as the CT imaging resolution, resulted in abnormal data. Therefore, we removed the MCO020 data in subsequent analyses.

Data analysis

Spearman’s correlation analysis was performed using the cor.test function and stats package in R. The Steel Dwass test using the Monte Carlo method (with 10,000 iterations) was performed using the pSDCFlig function and the NSM3 package in R. The Wilcoxon rank sum test was performed using the wilcox.test function and stats package in R. The exact Wilcoxon rank sum test was performed using the wilcox.exact function and exactRankTest package in R. Nonmetric multidimensional scaling (NMDS) was performed using the metaMDS function of the vegan package in R.

Results

Quantitative color differences analysis

To analyze the color differences in roof tiles, we quantified the color based on the information on the roof tile surface acquired by the image scanner (S2 Table in S1 Appendix). Hence, we measured the color distance between each roof tile using the CIEDE2000 method, which reflects differences in color perception (S3 Table in S1 Appendix). The result of clustering based on the color distance matrix showed two large reddish and grayish clusters, each containing two or three subclusters (Fig 2A). All the ancient roof tile styles were dispersed across multiple subclusters, especially the Minchō style, which was included within all subclusters and concentrated in two subclusters (SC 1A and SC 2C). In contrast, the Yamato and Kourai styles were only included in the grayish cluster, and both styles were distributed in multiple subclusters (SC 2A, SC 2B, and SC 2C). The modern tiles formed a single subcluster (SC 1B). We showed the relative distance to each roof tile using MDS analysis based on color distance (Fig 2B). The Minchō was widely distributed, compared with the other styles, and some of the tiles overlapped with other roof tile styles. These results suggest that Minchō had greater color diversity than other styles, but the grayish color of Kourai and Yamato differed slightly from that of Minchō.

Fig 2 — (A) Heatmap of roof tiles based on color distance. A light color indicates the color proximity based on the color distance of each roof tile. The left column shows clusters by dendrogram. Three ancient and one modern roof tile were classified into two clusters, including five subclusters: SC. The right column represents scanned images of each roof tile. MCO: Minchō style, KOR: Kourai style, YMO: Yamato style, MDN: Modern. (B) MDS of the roof tiles based on the color distance matrix: Classical MDS plot derived from the color distance matrix of each roof tile. The four polygons represent the convex hull of each roof tile style. (C) Box plot of color distance from six reddish colors. The points indicate the color distance between each color and the roof tiles that were classified in the red color cluster in Minchō style. The six reddish colors are five traditional Japanese reddish colors and a standard red color. “Bengara-iro” had a significantly smaller color distance from the roof tile color, compared with the other five colors (p < 0.001: Steel-Dwass test). (D) Box plot of color distance from Bengara-iro in both Minchō-style phases. The points indicate the color distance between Bengara-iro and the roof tiles in three groups: Early and late phases of Minchō style and the modern tile. The late Minchō phase had a significantly smaller color distance from Bengara-iro, compared with the early phase (p < 0.05: Steel-Dwass test).

Next, we focused on the reddish color of some Minchō roof tiles that are usually called Aka-gawara. There are multiple traditional reddish colors in Japan. Therefore, we analyzed the color distance between the roof tiles of the reddish cluster (Cluster 1) and each reddish color for five traditional Japanese reddish colors (“Hi-iro”, “Shoujyouhi”, Bengara-iro, “Beni-iro”, and “Syu-iro”) and a standard red color (Kodak color control patches) using the CIEDE2000 method (S4 Table in S1 Appendix). The results indicated that the reddish color of roof tiles was significantly similar to Bengara-iro significantly in these colors (Fig 2C). Furthermore, the color distance between the Bengara-iro and Minchō roof tiles was significantly less in the late phase, compared with the early phase (Fig 2D). These results suggest that the roof tiles in Cluster 1 was significantly similar to Bengara-iro, not the standard red color, and that the roof tile color became redder in the late phase of the Minchō style.

Comparative analysis of chemical components

We measured the ratio of 11 elemental components of the roof tiles using a non-destructive method (S5 Table in S1 Appendix). According to the results of the PCA, each style of roof tile was not clearly separated from the others, but there were certain trends in their components that can be explained by three biplot vectors, SiO₂, Al₂O₃, and Fe₂O₃ (Fig 3A). Furthermore, the results of the nonmetric multidimensional scaling (NMDS) analysis indicated that the characteristics of elemental compositions in the ancient roof tiles overlapped with each other but not with the modern tile (Fig 3B). Moreover, the two Minchō-style phases overlapped (Fig 3C). Additionally, to detect the feature elements that explain the differences between the ancient roof tile styles in each elemental composition, we performed a linear discriminant analysis effect size (LEfSe). The results indicated that the Minchō style was featured by two elements, Fe₂O₃ and K₂O, and the Kourai style was characterized by SiO₂ within each style (Fig 3D). Furthermore, the comparison among the Minchō-style tiles indicated that the early phase (Period Ⅰ) was featured by two elements, Al₂O₃ and Fe₂O₃, while the late phase (Period II and III) was characterized by SrO (Fig 3E). These results suggest that the ancient roof tiles were not clearly formed into independent clusters based on the elemental compositions. In contrast, the differences in roof tiles between styles or phases could be explained by specific elements in the roof tile paste.

Fig 3 — (A) PCA of the roof tiles based on chemical components. The roof tiles were plotted by PCA to show the contribution rate of each element among the roof tiles. The ellipses were 95% confidence ellipses assuming a multivariate t-distribution. (B) NMDS of the roof tiles based on chemical components. The NMDS plot was derived from the chemical components of each roof tile. The ellipses were 95% confidence ellipses assuming a multivariate t-distribution. (C) NMDS of the Minchō-style roof tiles based on chemical components. The ellipses were 95% confidence ellipses assuming a multivariate t-distribution. (D) LEfSe of the chemical components in three ancient roof tile styles. The LEfSe revealed a list of features that enabled discrimination between the three ancient roof tile styles. The horizonal bar graph shows the LDA scores for the Kourai and Minchō styles. The histograms show the presence of three featured elements in the ancient roof tiles. The horizontal straight lines in the panels indicate the group means, and the dotted lines indicate the group medians. (E) LEfSe of the chemical components in two Minchō-style phases. The LEfSe revealed a list of features that enabled discrimination between the three ancient roof tile styles. The horizonal bar graph shows the LDA scores for the early and late phases. The histograms show the presence of three featured elements in the Minchō-style roof tiles. The horizontal straight lines in the panels indicate the group means, and the dotted lines indicate the group medians.

Physical feature analysis for the internal microstructure of the roof tiles

We measured the internal structures of the roof tiles non-destructively using a CT scan and obtained eight values indicating the features (S6 Table in S1 Appendix, S8 Fig in S2 Appendix). An MDS analysis was performed using these values to check the similarity of the internal microstructure among styles (Fig 4A). The MDS plots overlapped among the three ancient roof tile styles (Korai, Yamato, and Minchō), indicating that the internal microstructure was similar. In contrast, the modern roof tiles had a different internal microstructure from the other styles, with the center of its 95% probability ellipse being away from the distributions of the other styles.

We then performed Kruskal-Wallis tests to check whether there were significant differences among the styles for each of the eight values (Fig 4B). Significant differences were found in seven values, the exception being Num-density. Next, post hoc tests were performed using the Steel-Dwass multiple comparison procedure to check which style combinations showed significant differences (S7 Table in S1 Appendix). As with the MDS results, most of the significant differences were found between the modern tiles and the other styles. All values except Mode-vol were significantly different between the modern tiles and the Minchō style, while there were significant differences between the modern tiles and the Yamato style for four values and for two values for the Korai style. On Density-S, the modern tile was significantly different from all the other styles.

When focusing on the three ancient styles, there were significant differences between Minchō and Korai for Median-vol and between Minchō and Yamato for Mode-vol. This means that the Minchō tiles tended to have larger pores than the Korai and Yamato tiles. Contrastingly, there was no significant difference in Density-M among the three.

In addition, to investigate the differences between the Minchō-style phases, we compared the early phase (Period I) with the late phase (Periods II and III). An MDS analysis was performed using eight values, which showed that the distribution of points completely overlapped between the early and late phases, and the internal microstructure was similar (Fig 4C). Exact Wilcoxon rank sum tests were also performed, but no significant differences were found for any of the values (S8 Table in S1 Appendix).

Correlation analysis of the factors contributing to the color change of the ancient roof tiles

To clarify the chemical components that affected the color change to reddish in Minchō tiles, we analyzed the correlation between the quantities of elements and the color distance from Bengara-iro in each roof tile (S9 Table in S1 Appendix, S9A Fig in S2 Appendix). The results indicated a significant correlation between the color distance and four elements: SiO₂, SO₃, K₂O, and SrO. In particular, the amount of SO₃ showed the most significant negative correlation (rho = −0.59, p = 0.0058). Additionally, we performed an LEfSe analysis to determine the elemental features to explain the differences between the reddish and grayish color clusters in the Minchō-style tiles (S10 Fig in S2 Appendix). The results indicated that the reddish cluster was characterized by SO₃, and the grayish cluster was featured by SiO₂. This suggests that the change in color to reddish in the Minchō tiles was the result of an increase in sulfur content.

Next, to clarify the microstructural properties related to the change in color to reddish in the Minchō tiles, we analyzed the correlation between the properties of the microstructure and the color distance from Bengara-iro in each roof tile. The results showed that five properties, Mode-vol, Median-vol, Num-density, RLV, and RLN, were significantly correlated with the color distance from Bengara-iro (S10 Table in S1 Appendix, S9B Fig in S2 Appendix). We also analyzed whether there were differences in the properties between the two Minchō reddish and grayish color clusters. The results indicated that the pore volume mode was significantly different (S11 Table in S1 Appendix, p = 0.032). This result suggests that the roof tile pores became larger and more numerous as the tile color moved closer to Bengara-iro, with the pore size being especially strongly related to the change.

Discussion

1. Transition from Kourai and Yamato to Early Minchō style

For the Kourai and Yamato roof tiles, which coexisted in the 14th to 15th centuries, there were external differences (shape, surface pattern, description, and imprinting) between the two styles, and it has been assumed that these differences are due to their different cultural origins and technological channels [6,9,20]. In contrast with these assumptions, however, our analysis did not detect any significant differences between the two styles in color tone, paste material characteristics, or internal structure (Figs 2B, 3B and 4A). In particular, there was no significant difference in paste characteristics between the two styles based on the overall elemental components and internal structures, suggesting that there was no significant difference in the processing of materials, firing conditions, or product quality of the two styles.

Roof tiles are not usually made from natural clay: the clay is processed by removing foreign bodies (e.g., large stones) and adding tempers to prepare the paste. Because there were no obvious foreign substances on the fractured surfaces of any of the samples, the tiles were thought to be refined. In the Urasoe Castle ruins, which are located on Okinawa Island and are older than those of Shuri Castle (ca. 13th to 15th century AD), there was no significant difference in the elemental composition of the Kourai- and Yamato-style tiles [27], and this finding was confirmed at Shuri Castle.

Because no clear traces of firing or production facilities for 13th- to 15th-century roof tiles have been found in the Ryukyu Islands, it has not been possible to conclude whether they were produced in this region or were imported. However, our results suggest that, at the very least, the technical background of the production of these two roof tile styles at Shuri Castle (perhaps also at Urasoe Castle) had similarities that can be detected in their color, paste characteristics, and internal structure. This might mean that both the Kourai and Yamato roof tiles were made in the Ryukyus.

The Minchō style is estimated to have been established around the 16th century AD [28], but the cultural or technological connections with previous styles, such as Kourai and Yamato, have not been clarified. Although some preceding studies have shown that the external characteristics of Minchō tiles are significantly different compared to previous styles [6,28], our results confirmed changes in Minchō color, raw materials, and firing temperatures.

First, our results for the gray tiles show that Minchō tiles have a different coloration from Kourai and Yamato tiles. Previous studies [6,29] had considered the gray tiles of all styles to be based on the same reduction flame, but the analysis of the color tone indicated variations in the reduction flame firing.

Second, significant differences in the elemental characteristics of the paste, such as Fe₂O₃, K₂O, and SiO₂, were found, mainly between the Kourai and Minchō styles. In contrast, the overall elemental components of Minchō and previous styles were similar; therefore, it is unlikely that the processing method, such as refining the material or adding admixtures, changed significantly. The differences in the characteristics of the paste found between the Kourai and Minchō styles may reflect differences in the origin or composition of the raw clay.

Third, we detected an increase in the size of the internal pores from previous styles to the Minchō (Fig 4B). It has been reported that the average pore radius increases with higher firing temperatures up to approximately 1000–1050°C [30], suggesting that the Minchō used a higher firing temperature.

However, the elemental composition of the paste and the density of the tiles did not change significantly from previous styles to the Minchō (Figs 3A, 3B and 4B), suggesting marginal change in the processing method for the paste and the quality of the tiles. There are two possible explanations for this finding. The first is that the soil conditions and product quality required for tile production were the same, regardless of the cultural or technological background. The second is that the commonality of material processing and product quality indicates technical continuity from the Kourai and Yamato to the Minchō. Our results suggest a new possibility—that the Minchō was influenced by Chinese patterns and forms but inherited techniques related to material processing and tile quality from the Kourai and Yamato. Although this possibility needs to be tested through comparisons with tiles from surrounding areas, it is a new possibility that emerged only by looking at tiles from a complex data perspective.

2. Unique color transition after Minchō style and its background

The Minchō-style roof tiles that appeared from the 16th to the 17th centuries had many chronological variations and were recognized as having developed their uniqueness in the Ryukyu Islands. In particular, the changes in patterns, shapes, and colors were considered characteristic. Of these, the change in color, from grayish to reddish, was recognized as a major turning point leading to the modern red roof tile (Aka-gawara), and previous studies have attributed this to changing firing conditions from reduction to oxidation [6,29,31]. However, it is thought that the changes did not all occur at once, because when temporally classified, the changes in the various roof tile patterns overlap [12]. Our results support this explanation of the color transition. In contrast, the quantitative analysis revealed that the color in the later phase, which was recognized as a single color in Minchō, resembled Bengara-iro more closely (Fig 2D). This also suggests that the change to Bengara-iro was gradual.

The proximity to Bengara-iro was positively correlated with the amount of sulfur (S9 Table in S1 Appendix and S10 Fig in S2 Appendix). Because the sulfur content in the two Minchō-style phases was not significantly different, the amount of sulfur in the paste correlated with the color, not the pattern. It has been reported that sulfur vapor at high temperatures in oxidation firing conditions reacts with hematite on the pottery surface to generate ferrous sulfide (FeS) [32]. In our case, sulfur that was indirectly present under the firing conditions vaporized and most likely sulfurized some of the iron on the roof tile surface during the oxidation firing. Firewood may contain trace amounts of sulfur, depending on species and growth environment [33] and is considered a candidate source of sulfur during burning. In addition, the firing temperature may have increased because the pore size of the reddish cluster of roof tiles was significantly larger than that of the grayish cluster (S11 Table in S1 Appendix). Therefore, there are three possibilities for the progression of color tone toward Bengara-iro: 1) the oxidation of iron increased in the firing conditions with the rise in temperature; 2) the type of wood used and/or the amount of sulfur it contained changed; or 3) a combination of both. The detailed mechanisms will require additional research in the future. Interestingly, the historical and archeological data strongly suggest that the transition related to colors and elements was caused by changes in firing conditions and firewood. For example, the political consolidation of the ceramic production areas in Tsuboya (Naha, Okinawa) occurred in AD 1682 [6] and resulted in a change in kiln type from flat kilns to climbing kilns. In addition, there was an increase in demand for timber during the late 17th and early 18th centuries because of an increase in public works projects, such as the construction of the royal palace at Shuri Castle; industrial factors (i.e., shipbuilding and sugar production); and the firewood needs associated with population growth [34]. Limited firewood resources are considered to have contributed to the conversion to oxidative firing [6], and it is possible that the firewood itself changed.

The Bengara-iro colored Minchō-style roof tiles endured in the Ryukyu Islands until the 19th century. During this period, black smoked roof tiles named “Ibushi-gawara” were produced for use in symbolic castles in Japan (e.g., Edo Castle) [3], and glazed roof tiles, in colors such as yellow and green, were produced in China (e.g., Forbidden City) [35]. Although Ryukyu had the opportunity to import techniques from Japan and China, it did not use these techniques but developed its own unglazed reddish roof tiles. The reasons for this remain unclear, but as Ryukyu was in the complicated situation of being under the control of Satsuma while having a feudal relationship with China [36], it is possible that they did not prioritize the incorporation of a single country’s technology. This suggests that the background to the appearance of colored roof tiles in Ryukyu encompasses not only the production technology and fuel resources, but also the socio-political background.

3. New perspective on the evolution of Kawara in Ryukyu

In this study, we analyzed the roof tiles from Shuri Castle and detected differences and similarities in color, material characteristics, and internal structure between past styles. Based on changes in each factor, we classified them into three transition patterns (Fig 5). Transition pattern 1 based on the external characteristics comprised the most variables. Transition pattern 2 consisted of three factors (color tone, elemental characteristics of the paste, and pore size), which had similarities in the Kourai and Yamato styles and differed in specific phases of the Minchō style. And transition pattern 3 consisted of two factors (whole chemical composition and paste density), which had similarities among all ancient roof tile styles. Transition patterns 2 and 3, newly revealed by our analysis, differed from the existing tile transition (transition pattern 1).

Fig 5 — Differences in appearance factors, such as patterns, forms, and traces between tile styles, from previous studies are shown (transition pattern 1). These differences are the basis for differentiating the roof tile style or period. Our analysis showed no significant difference between the Kourai and Yamato for color, specific elements (SE) of paste, and pores in paste; in contrast, after the emergence of the Minchō style, these factors were detected to have changed, along with the appearance characteristics (transition pattern 2). No significant difference in elemental composition (EC) or paste density were found between any of the ancient tile styles, only between them and the modern reconstructed tiles (transition pattern 3). Contrastingly, in the modern reconstructed tiles, the external factors did not change because they were designed as imitations, while internal factors such as SE, EC, pore, and paste density were confirmed to be significantly different.

We suggest two hypotheses to explain the detection of the three transition patterns the roof tiles as material culture (Fig 6). The conventional hypothesis based on previous studies is that each roof tile trait was subjected to individual stress, which resulted in the formation of multiple transition patterns (Fig 6A). This resembles the previous view of cultural evolution. For example, Dunnel suggested that an artifact has multiple traits classified into style and function and supposed different evolution mechanisms between them [37]. Boyd and Richerson [38], also focused on separate evolutionary processes of multiple traits of material culture and considered that evolution results from stress that encourages the selection of various individual traits. However, in these explanations of cultural evolution, external stress seems to affect each trait individually, yet it is difficult to envision such a situation in practice. For example, if we proposed that the evolution of transition pattern 2 was caused by social stressors such as fuel shortages, such stressors must not only have affected the factors related to transition pattern 2 (color tone, materials, internal structure), but also the whole of the roof tile culture.

Therefore, we propose another possibility based on the conventional hypothesis that an external social or environmental stress affected the “whole” of the roof tile culture but that only limited traits reacted to it, resulting in the formation of various transition patterns (Fig 6B). Although it is not different from the previous view of cultural evolution in that each factor evolves differently, it shows a new correspondence between traits and external stresses. In the case of transition pattern 2, it is considered that the entire roof tile culture was subjected to the social stress caused by fuel shortages, but some of the firing methods showed an especially strong reaction to it, which led to a transition in the color and pore size of the tiles. In other words, the roof tile artisans responded to this stress by changing their firing methods. In contrast, other factors did not respond to the stress, and therefore constructed different transition patterns.

The evolutional mechanism of roof tiles has only been revealed by studying the evolution of each factor from multiple perspectives. In some archaeological studies, artifacts including roof tiles were classified based on limited factors that were mainly external characteristics, and that classification was regarded as a historical indicator of human behavior and culture. In previous studies of external characteristics, ancient Ryukyuan roof tiles were considered to have been influenced by some roof tile cultures of mainland Japan, the Korean peninsula, and the Chinese dynasty. However, although Ryukyuan roof tiles were transformed into unique red tiles in the Ryukyu Islands and were passed on to today’s traditional culture, these cultural origins do not sufficiently explain the roof tiles’ evolution. We therefore provided a new perspective on the evolution of the Ryukyuan roof tile by proposing that it was caused by external stresses and various reactions of traits to such stressors. As a result of our multifaceted analysis, we conclude that the uniqueness of Ryukyuan roof tiles is the result of the accumulation of external stresses in Ryukyu society and the various responses and evolutions of each trait to such stressors. This is similar to the relationship between external stress and the response of each trait as well as the relationship between such traits and evolution described in the field of ecological developmental biology [39]. Many discussions of cultural evolution have applied biological evolutionary models, such as neutrality and adaptation, to the components of material culture, and our study also provides another way to think about and discuss the relationship between external stresses and responsiveness on material culture. This is a new perspective on the evolution of Ryukyuan roof tiles that could not have been achieved by discussing only the cultural origins and temporal changes based on specific factors such as appearance and form. We believe that this logic can be applied not only to Ryukyu tiles, but also to various other material cultures.

Conclusion

This study reveals for the first time the chronological transition of internal features such as elemental composition, color tone, and internal structure of the paste of ancient Ryukyu tiles from Shuri Castle. Previous archaeological studies have explained the history of the roof tiles based on changes in their external features, but the internal features show a different transition from these external features. It is also noteworthy that some internal features of the tiles remained largely unchanged for a very long time after the appearance of the tiles in the Ryukyu Islands. We propose that these factor-specific chronological transitions are the result of the different reactions of individual factors to external stresses based on environmental and sociopolitical contexts. In other words, the characteristics of Ryukyuan tiles did not all change at the same time, only the factors that responded to various external stresses. This shows the historical flexibility of Ryukyuan tile making, and we conclude that this represents the evolution of the roof tiles. This is a significant change from the conventional framework that views roof tile history in terms of a single transition such as typology, and it can be applied to research on various other material cultures. Our analytical and statistical methods, which were non-destructive, can also be applied to a variety of artifacts, such as pottery and ceramics, allowing us to analyze samples that include valuable materials. We believe that this will lead to new research and discussion of the evolution of material culture.

Supporting information

S1 Appendix. Supplementary tables S1-S11.

(XLSX)

Click here for additional data file.^{(72.7KB, xlsx)}

S2 Appendix. Supplementary methods and figures S1-S10.

(PDF)

Click here for additional data file.^{(1.4MB, pdf)}

Acknowledgments

This research was funded by a grant from the Shuri-jo Revival Research Project of the University of the Ryukyus. This research used the Ryukyu University Museum (Fujukan) collection and samples provided by the Damaged Roof Tile Utilization Project of Okinawa Prefecture. The XGT measurements were performed at the University of the Ryukyus Center for Research Advancement and Collaboration. X-ray CT measurements were performed at the Okinawa Industrial Technology Center. We thank T. Sasaki for managing the study’s resources and providing much information, advice, and help, M. Shimabukuro for help with financial operations, M. Ozaki and S. Yabu for technical assistance, T. Kawano for providing a picture of Shuri Castle, and Y. Tanahara, K. Hanashiro, and other members of the Okinawa Industrial Technology Center for providing information and advice.

Data Availability

All relevant data are within the paper and its Supporting Information files.

Funding Statement

This research was funded by a grant from the Shuri-jo Revival Research Project of the University of the Ryukyus. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1.ICOMOS. Gusuku Sites and Related Properties of the Kingdom of Ryukyu. 2000. [Cited 2021 June 27]. In: World Heritage List [Internet]. Paris: World Heritage Centre 1992. Available from: https://whc.unesco.org/en/list/972/.
2.Esaka T, Serizawa C, Sakadume S. [New Japanese Archaeological Dictionarly] Shin, Nihon Kouko-gaku Jiten (in Japanese). Tokyo: New Science Co., Ltd.; 2020. [Google Scholar]
3.Uehara M. [Study of “Kawara” Roof Tiles] Kawara wo Yomu (in Japanese). Tokyo: Kodansha Ltd.; 1997. [Google Scholar]
4.Okawa K. [The Beauty of “Kawara” Roof Tiles: Unearthed Japanese History] Kawara no Bi: Umoreta Nihon Kodaishi (in Japanese). Tokyo: Shakai Sisousha Co., Ltd.; 1966. [Google Scholar]
5.Okawa K. [Ancient “Kawara” Roof Tiles of Japan] Kodai no Kawara (in Japanese). Yougyou-Shi Hakubutsukan (Museum of Ceramic History); 1996.
6.Uehara S. [Study of Ancient Roof Tiles in Ryukyu Islands] Ryukyu Koga no Kenkyu (in Japanese). Ginowan: Youju Shorin Ltd.; 2013. [Google Scholar]
7.Ikeda Y. [Relationship between Jeju Island in Koreal and Ryukyu Islands based on Material Culture in the Goryeo Period] Bussitsu Bunka Kenkyu kara Mita Kankoku Cheju-tou to Ryukyu-rettou: Kourai Jidai wo Chushin to shite (in Japanese). Ryudai Asian Stud. 1998;2: 1–10. [Google Scholar]
8.Sekiguchi K. [Transition and Historical Background of Ceramic Roof tile Manufacturing Based on Ancient Roof Tiles unearthed from “Honmaru remain” of Katsuren Castle (Katsuren-jo)] Okinawa ni okeru Zouga-gijutsu no Hensen to Sono-aida no Jijou: Katsuren-jou Honmaru-ato Shutsudo Koga wo Chushin ni shite (in Japanese). Kōkogaku zasshi (Journal of the Archaeological Society of Nippon). 1976;62: 222–239. [Google Scholar]
9.Yamazaki S. [Study of Roof tiles in the Middle Age of Japan] Chu-Sei Gawara no Kenkyu (in Japanese). Tokyo: Yuzankaku, Inc.; 2000. [Google Scholar]
10.Uehara S. [Transition of Mincho Style Roof Tiles Unearthed from “Nishi no Azana” Region in the Shuri Castle (Shuri-jo)] Shuri-jo Nishi no Azana chiku Syutsudo no Mincho-kei kawara to sono Suii (in Japanese). Nantō kōko (The journal of the Okinawa Archaeological Society) 1994;14: 153–186. [Google Scholar]
11.Ishii R. Corpus and typological classification of eave-tile endpiece designs of Ryukyu modern roof tiles (in Japanese). Bulletin of the Department of Archaeology, The University of Tokyo. 2006; 20: 109–148. doi: 10.15083/00027626 [DOI] [Google Scholar]
12.Ishii R. Transition of the roof on Shurijo castle Seiden. Josai Management Review (in Japanese). 2020;16: 157–181. doi: 10.20566/18801536_16_157 [DOI] [Google Scholar]
13.Tsukada S. [Study of Historical Epitaph of Ryukyu Kingdom] Ryukyu-koku Hibunn-ki (in Japanese). Tokyo: Keigaku Shuppan Co., Ltd; 1970. [Google Scholar]
14.Okinawa Prefectural Archaeological Center, editor. [Report of Archaeological Survey in the Main Palace of Shuri Castle (Shuri-jo)] Syuri jyou-shi -Seiden chiku hakkutsu chousa houkokusyo (in Japanese). Nishihara: Okinawa Prefectural Archaeological Center; 2016. [Google Scholar]
15.Tei H. Kyūyō [edited by Kyūyō kenkyū-kai] (in Japanese). Tokyo: Kadokawa Shoten Publishing Co., Ltd.; 1974. [Google Scholar]
16.Asato S. [Transition of the Main Place Platform of Shuri Castle (Shuri-jo)] Shurijyo seiden kidan no hennsen (in Japanese). Syurijyo kenkyu (Journal of Shuri-jo Study) 1996;2: 5–15. [Google Scholar]
17.Chen K. Shi ryukyu roku [translated with notes by Harada N]. Ginowan: Youju Shorin Ltd.; 1995. (in Japanese). [Google Scholar]
18.Takara K. [Academic Resource 1: Historical Resources about Official History] Shiryou-hen 1: Seishi kankei (in Japanese). In: Takara K, Yamazato J, Tomiyama K, Maehira F, Akamine M, Karimata S, editors. [Study of Disaster Based on the Historical Resources in Okinawa] Okinawa no Saigai Joho ni Kansuru Rekishi Bunnkenn wo Syutai toshita Sogo Kenkyu; 2008. pp. 69–97. http://ir.lib.u-ryukyu.ac.jp/handle/20.500.12000/8987. [Google Scholar]
19.Ishii R. [Analysis of “Kawara” Roof Tiles Unearthed from “Watanji-mura” remain: Discussion about the Early modern Kawara in Ryukyu] Watanji-mura Ato Shutsudo Kawara no bunseki: Ryukyu kinsei kawara no teigi ni Kanshite (in Japanese). Bulletin of the Tsuboya Ceramics Museum 2015;15: 23–43. [Google Scholar]
20.Uehara S. [Historical Relationship Between Ryukyu Islands and Surrounding Regions Based on the Study of Ancient Roof Tiles] Koga kara mita Ryukyu rettou no Kouryu-shi (in Japanese). Cultura antiqua 2000;52: 167–171. [Google Scholar]
21.CIE. Improvement to Industrial Colour-Difference Evaluation (Vienna: CIE Publication No. 142–2001, Central Bureau of the CIE, 2001). [Google Scholar]
22.Taylor, G. Python Colormath Module. Available online: https://pypi.python.org/pypi/colormath/.
23.Aoyama H, Yamagiwa K, Fujimoto S, Izumi J, Ishikawa R, Kameshima S, et al. A new nondestructive approach to chemical analysis of potsherds using an X‐ray fluorescence microscope: Case study about the past pottery manufacture in the Yaeyama Islands. X-Ray Spectrom. 2018;47: 265–272. doi: 10.1002/xrs.2837 [DOI] [Google Scholar]
24.Yamagiwa K, Fujimoto S, Aoyama H, Izumi J, Kameshima S. A possible new oldest pottery group in the Southern Ryukyu Islands, Japan: comparative analysis of elemental components of potsherds from the Shiraho-Saonetabaru Cave Site. J. Archaeol. Sci. Rep. 2019;26: 101879. doi: 10.1016/j.jasrep.2019.101879 [DOI] [Google Scholar]
25.Segata N, Izard J, Waldron L, Gevers D, Miropolsky L, Garrett WS, et al. Metagenomic biomarker discovery and explanation. Genome biol. 2011;12: 1–18. doi: 10.1186/gb-2011-12-6-r60 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Cheung S, Nakamoto M, Hamuro Y. NYSOL: a user-centric framework for knowledge discovery in big data. Int. J. Knowl. Eng. 2015;1: 214–218. doi: 10.18178/ijke.2015.1.3.037 [DOI] [Google Scholar]
27.Urasoe City Board of Education, editor. [Survey Report of “Urasoe Youdore” vol. 2: About the Kawara Concentration Remain] Urasoe Youdore 2: Kawara-tamari ikou hen (in Japanese). Urasoe: Urasoe City Board of Education; 2005. [Google Scholar]
28.Ishii R. Relationship with roof tile and Ryukyu history—kingship, legislation, interpretation, regional interaction (in Japanese). In: Nishimura M, Shinohara H, Okamoto H, editors. [Vietnam, Korea, and Ryukyu in East Asia Based on Archeological and Historical Study] Shuen to chushin no gainen de yomitoku higashi ajia no Etsu Kan Ryuu: Rekishigaku kokogaku kenkyu kara no shiza. Suita: Institute for Cultural Interaction Studies, Kansai University; 2012. pp. 161–177. [Google Scholar]
29.Teruya Y, Hanashiro K, Yoza N, Ginoza T. [Scientific Study of Ancient Roof Tile in Shuri Castle (Shuri-jo)] Shurijo furu gawara nado no rikagakuteki kenkyu (in Japanese). Okinawa Industrial Technology Center Report. 1988;16: 51–75. [Google Scholar]
30.Jordan MM, Montero MA, Meseguer S, Sanfeliu T. Influence of firing temperature and mineralogical composition on bending strength and porosity of ceramic tile bodies. Appl. Clay Sci. 2008;42: 266–271. doi: 10.1016/j.clay.2008.01.005 [DOI] [Google Scholar]
31.Teruya Y, Ginoza T, Yoza N, Hanashiro K. [Study About Reproduction of Ancient Roof Tiles in Shuri Castle (Shuri-jo)] Shurijo seiden fukugen gawara nado no kaihatsu ni kansuru kenkyu (in Japanese). Okinawa Industrial Technology Center Report. 1988;16: 89–105. [Google Scholar]
32.Morgan WR. Oxidation and loss of weight of clay bodies during firing: a report of an investigation conducted by the engineering experiment station. University of Illinois in cooperation with the Clay Products Association. 1936. Available from: http://hdl.handle.net/2142/4389. [Google Scholar]
33.Sakurai K. Sulfur content in wood. Bull. For. & Fo. Prod. Res. Inst. 1986;335: 141–145. (in Japanese). [Google Scholar]
34.Nakama Y. [History of Forest Cultivation System in Okinawa] Okinawa Rinya Seido Riyou-shi Kenkyu (in Japanese). Naha: Media Express Co., Ltd; 2011. [Google Scholar]
35.Gou A, Wang J. The development of roof color in ancient China. Color Res. Appl. 2010;35: 246–266. doi: 10.1002/col.20564 [DOI] [Google Scholar]
36.Okamoto H. Structural transformation of Ryukyu Kingdom in the 17th and rarly 18th centuries: as an intersection of cultural interaction. In: Masanari N, Sato M, Kimura M, Okamoto H, editors. Cultural reproduction on its interface: from the perspectives of text, diplomacy, otherness, and tea in East Asia. Suita: Institute for Cultural Interaction Studies, Kansai University; 2010. pp. 3–17. [Google Scholar]
37.Dunnell RC. Style and function: a fundamental dichotomy. Am Antiq. 1978;43: 192–202. doi: 10.2307/279244 [DOI] [Google Scholar]
38.Boyd R, Richerson PJ. Culture and the evolutionary process. Chicago. University of Chicago Press; 1988. [Google Scholar]
39.Gilbert SF, Epel D. Ecological developmental biology: the environmental regulation of development, health, and evolution. 2^nd ed. Sunderland: Sinauer Associates, Inc. Publishers; 2015. [Google Scholar]

PLoS One. doi: 10.1371/journal.pone.0277560.r001

Decision Letter 0

Abbas Farmany

5 Aug 2022

PONE-D-22-12204A new perspective on the evolution of “Kawara” roof tiles in Ryukyu: a multidisciplinary non-destructive analysis of roof tile transition at Shuri Castle, Ryukyu Islands, Japan.

PLOS ONE

Dear Dr. Aoyama,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Sep 19 2022 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Abbas Farmany

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that Figure 1 in your submission contain copyrighted images. All PLOS content is published under the Creative Commons Attribution License (CC BY 4.0), which means that the manuscript, images, and Supporting Information files will be freely available online, and any third party is permitted to access, download, copy, distribute, and use these materials in any way, even commercially, with proper attribution. For more information, see our copyright guidelines: http://journals.plos.org/plosone/s/licenses-and-copyright.

We require you to either (1) present written permission from the copyright holder to publish these figures specifically under the CC BY 4.0 license, or (2) remove the figures from your submission:

1. You may seek permission from the original copyright holder of Figure 1 to publish the content specifically under the CC BY 4.0 license.

We recommend that you contact the original copyright holder with the Content Permission Form (http://journals.plos.org/plosone/s/file?id=7c09/content-permission-form.pdf) and the following text:

“I request permission for the open-access journal PLOS ONE to publish XXX under the Creative Commons Attribution License (CCAL) CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). Please be aware that this license allows unrestricted use and distribution, even commercially, by third parties. Please reply and provide explicit written permission to publish XXX under a CC BY license and complete the attached form.”

Please upload the completed Content Permission Form or other proof of granted permissions as an "Other" file with your submission.

In the figure caption of the copyrighted figure, please include the following text: “Reprinted from [ref] under a CC BY license, with permission from [name of publisher], original copyright [original copyright year].”

2. If you are unable to obtain permission from the original copyright holder to publish these figures under the CC BY 4.0 license or if the copyright holder’s requirements are incompatible with the CC BY 4.0 license, please either i) remove the figure or ii) supply a replacement figure that complies with the CC BY 4.0 license. Please check copyright information on all replacement figures and update the figure caption with source information. If applicable, please specify in the figure caption text when a figure is similar but not identical to the original image and is therefore for illustrative purposes only.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: N/A

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: See uploaded PDF document for many details and feel free to share it with the authors (I wrote it assuming they would see it). In general, the analysis of the tiles themselves seems innovative, but the authors struggle, without success, to place their work in a broader social and historical context. I recommend cutting out most of this social and historical material and making what remains more precise and accurate.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Attachment

Submitted filename: Reader report Shuri Castle roof tiles.pdf

Click here for additional data file.^{(664.5KB, pdf)}

PLoS One. 2022 Nov 16;17(11):e0277560. doi: 10.1371/journal.pone.0277560.r002

Author response to Decision Letter 0

25 Sep 2022

＜Comments to Reviewer＞

Thank you for reviewing our manuscript. We think your peer review made our paper better by reaffirming the content of our discussion. On the other hand, we have only partially agreed with your suggestions. The reasons and our considerations are detailed below.

Report on “A new perspective on the evolution of “Kawara” roof tiles in Ryukyu”

This article reports the results of a sophisticated technical analysis of Shuri Castle roof tiles from different time periods and attempts to hypothesize the broader significance of the authors’ laboratory analysis. The authors also attempt to set their analysis of roof tiles in a historical context.

I am not qualified to comment on the technical analysis and methodology itself. Because that realm appears to be the authors’ area of expertise, I assume in this report that their analysis is sound. My comments deal with roof tile geography, historical context, the authors’ attempt to link their technical findings with social change/social history, and some aspects of writing. Let me say at the outset that there are problems in each of these areas. I think they can be resolved or partially resolved, but doing so will require significant effort. I have underlined some words and phrases for emphasis.

Roof Tile Geography

As the authors know, Shuri Castle was not the only large stone-walled gusuku with roof-tiled structures dating from the fourteenth century. Of course, there are pragmatic constraints with respect to the size of the sample set in a study like this one. So I can well understand why the authors analyze only tiles from Shuri Castle. However, there it least one major tile-related issue about which the authors are silent. Scholars have long speculated about the significance of roof tiles made in almost exact imitation of late Goryeo (Korean) tiles. (The authors often refer to Goryeo as Kōrai [or Kourai], the Japanese pronunciation of the Goryeo dynasty in Korean history, 918-1392). These tiles are found at Shuri Castle, Urasoe Castle, and Katsuren Castle. One issue regarding the tiles is their date. In all three cases, the tiles display the date癸酉. Some scholars argue the sexagenary characters indicate 1273 and others argue for 1333. Ikeda Yoshifumi, for example, favors 1333 partly on the grounds that large stone-walled gusuku themselves came into existence only in the fourteenth century. I find Ikeda’s argument convincing. Other scholars, especially those who tend to regard Ryukyu’s official histories as accurate sources, favor 1273 because of the claim that Eiso resided at Urasoe castle (although some scholars dispute this claim about Eiso).

In Ancient Ryukyu (University of Hawaii Press, 2013), Richard Pearson favors 1273. He also notes that analysis of the tiles indicates that the clay all came from the same quarry, located near Nago. I am away from my office as I write this report, and I do not have access to my library. So unfortunately, I cannot provide page numbers. However, Person’s discussion of roof tiles is extensive and can easily be found via the book’s index.

I am aware of over twenty different works that discuss these Korean-style roof tiles and their significance, and I am sure there are many other works I have not seen. So what is the point with respect to the current manuscript? Obviously the authors have analyzed only tiles from Shuri Castle at this point. Instead of vague speculation about social matters (more on that below) in an attempt to justify the article, perhaps a better (or additional) approach might be to discuss the possible significance of their analysis with respect to the Korean tile academic debate. Perhaps their analysis, if extended to the other two locations, can she useful light on this matter.

[One other background point, which might be worth mentioning, is that stone-walled gusuku in the Ryukyu islands [as opposed to trench-and-earthworks gusuku like Sashiki ui gusuku or Nago gusuku] predate Japanese castles by at least a century. So from where did the knowhow to build them come? Chinese castles are entirely different in terms of their construction. But Korean mountain fortresses (山城), of which there are over 2,000 extant examples, resemble Ryukyuan stone-walled gusuku almost exactly (including the three major types of wall construction). One excellent source for the Korean fortresses is Guglib munhwajae yeonguso 國立文化財硏究所 (National Research Institute of Cultural Heritage, Korea). 2012. Hangug gogohag jeonmun sajeon, seonggwag bongsu pyeon 韓國考古學專門事典: 城郭, 烽燧篇. Daejeon, South Korea: Guglib munhwajae yeonguso.]

＜Comments to Reviewer＞

First, we focused on Shuri Castle, because it is one of the cultural centers and tiles have been used for a relatively long period of time in the same place. We approach the evolution of the tiles by scientifically analyzing the changes in the tiles limited to Shuri Castle. Therefore, as the reviewer pointed out, the analysis was conducted only for the Shuri Castle tiles.

We are also aware that there is some debate about the date of manufacture of the Kourai tiles, as has been pointed out, but we understand that no definite decision has yet been reached at this time.

And we inferred social issues based on the data analyzed in this study, not merely vaguely imagined. In those analyses we used three types of tiles (Yamato, Kourai, and Mincho), not only Kourai. Through comparisons of these three types, we discussed the relationship between technological changes and social context.

Historical Context

Of course, none of us can be experts on everything. The literature on Ryukyuan history and archaeology is vast, and over the past fifteen years or so, our understanding of Ryukyu’s deep past has changed significantly. Obviously, certain points

are debated and many questions remain unsettled. My comments below are based on a general consensus of recent (past 15 years) work in ancient Ryukyu. Nearly all of this research is in Japanese, but one important, relevant book in English is Richard Pearson’s work mentioned above. In the paragraphs below I will paste in statements from the manuscript and comment on them (manuscript line numbers included). Overall, the problem is a lack of specificity with respect to time periods and concepts.

In the Ryukyu Islands, Kawara was introduced 72 after about the 13th or 14th century AD, when multiple types of roof tiles appeared with several origins 73 assumed to be mainland Japan, the Korean peninsula, or mainland China. The ancient roof tiles, as a 74 typical artifact, were evidence of the prosperous international relations between the Ryukyu kingdom 75 and its surrounding countries.

The authors cite an early article by Ikeda Yoshifumi. I recommend they read some of his work from 2012 onward—he often repeats himself, so any of several book chapters or articles will do. As Ikeda points out, although some gusuku sites existed in the 13th century, the stone walls and tile-roofed structures all date from the 14th century. As for origins of gusuku era kawara: mainland Japan and Korea, yes; China, no.

＜Comments to Reviewer＞

Since Ikeda noted that matter in his 2012 reference (Ikeda 2012), we found no need to change the citation. In addition, reviewer has noted that Ikeda has made it clear in his recent writings that tiled buildings appeared after the 14th century, but this is not accurate. Ikeda (2019), for example, states that "the appearance of buildings with stone walls and platforms is in the fourteenth-century period" (Ikeda 2019:29). However, it is not clear whether the only buildings with platforms were tile-wiping buildings, and Ikeda does not state that tile-roofed buildings appeared after the 14th century. There are several possible dates for the appearance of Kourai roof tiles including the late 13th century, and the issue has not been settled. It is not appropriate to define the appearance of the roof tiles as being after the 14th century on the basis of the appearance of the platform building alone, and we have used a wide range of dates in our manuscript. Furthermore, we do not discuss in this section the date of appearance of buildings shingled with Kourai style roof tiles. Therefore, we have not adopted the reviewer's recommended citation.

[Reference]

Ikeda, Y. 2019. Ryukyu Rettou-shi wo Horiokosu: Juuichi-Juuyonseiki no Ijuu, Koueki to Syakai-teki Henyou (The Unearthed History of Ryukyu Islands: Maritime Trade, Migration, and Social Transition). Chuusei gaku kenkyu-kai (ed.). Ryukyu no Chuusei (Middle Age of Ryukyu). Koushi Shoin inc, Tokyo: 13-37.

Now, for something more subtle but very important. It is common for many writers to throw around the term “Ryukyu kingdom” casually and loosely. This term is very tricky because it means two different things across time. The first meaning (1370s-ca. 1500): some people (or fictitious people) on the island of Okinawa conducted official trade with China starting in the 1370s. These people held the title “wang” which is always translated as “king.” However, the title wang was simply a license to conduct official trade with China. It says nothing about the domestic circumstances of these “kings,” of which there were as many as four at a time in Okinawa, ca. the late 1300s. The second meaning (ca. 1500-1879). When did a Ryukyu kingdom in the usual sense of the term (a centralized government headed by a monarch) come into existence? The traditional date is 1429, but that is clearly a creation of later centuries. There is no exact date, but one possibility is Shō Shin’s successful (for him) invasion of Yaeyama in 1500. Nevertheless, additional warfare was needed to complete his conquests of the Ryukyu islands. Moreover, the earliest extant written documents used for domestic governance (jiresisho) are from the late 1520s. There is much more to say, but I would put the start date of a genuine Ryukyu kingdom at about 1520 or 1530—roughly 100 years later than the usual claim based on the official histories.

So “the prosperous international relations between the Ryukyu kingdom and its surrounding countries” doesn’t really make sense with respect to the 13th or 14th centuries. There was no Ryukyu kingdom then, and as I mentioned, the first people holding the title “king” (meaning trade license) appeared in the 1370s.

＜Comments to Reviewer＞

Pearson (2013) described the beginning of the Ryukyu Kingdom period to AD 1429 (p. 234). The Okinawa Archaeological Society (2019) also described the beginning of the Ryukyu Kingdom period in the early 15th century, including AD 1429 (some argue the late 14th century); the theory of the late 16th century is not a consensus within Okinawan archaeologists, and we do not adopt it.

[References]

Pearson, R. 2013. Ancient Ryukyu: An Archaeological Study of Island Communities. University of Hawai'I Press, Honolulu.

Okinawan Archaeological Society (ed.). 2018. Nantou Kouko Nyuumon: Horidasareta Okinawa no Rekishi, Bunka (Japanese Southern Islands Archaeology: Unearthed History and Culture of Okinawa). Borderink, Okinawa.

Here is a more accurate statement: Kawara first appeared in Okinawa in the early fourteenth century in connection with the construction of large fortresses (gusuku). The Korean peninsula was one source of kawara or kawara-making knowhow, and it is also likely that mainland Japanese kawara or kawara-making knowhow also came into the Ryukyu islands at about this time. The arrival of kawara corresponds to a time when local powers based in harbors in Okinawa pursued vigorous regional trade throughout the East China Sea.

79 Ryukyuan roof tiles appeared at least as early as the 13th to 14th centuries AD (This period corresponds 80 to the Gusuk period in 10th to 14th century of Ryukyu Islands) and are presumed to have been used in 81 Gusuku, the large stone-walled castles and fortresses that are representative of this period. The oldest 82 tiles are classified into two styles: Kourai and Yamato.

Again, instead of “as early as the 13th to 14th centuries” I recommend “circa the early 14th century.” Roof tiles were not “presumed to have been used” in gusuku. Instead, they were used in gusuku, specifically for the roofs of certain structures inside these gusuku. This second point is mainly an issue of diction or style, and it is subtle. For academic writing in English, I recommend clear declarations when the evidence supports them (more on this topic below).

＜Comments to Reviewer＞

As we have written before, there are various theories on the date of appearance of tiles in Ryukyu, and we cannot make a definite determination at this time. Therefore, we have given a wide range of date, "as early as the 13th to 14th centuries"

The Kourai style is associated with Korea’s 83 Goryeo Kingdom (AD 918–1392) due to external characteristics, such as design, shape, paddling traces, 84 imprinting, and inscriptions on their surface [4–6]; specifically, they are assumed to be connected to 85 refugees from the failure of the “Sambyeolcho” Rebellion in Goryeo (AD 1273)

This is a good point. Indeed, as Kurima Yasuo and others have argued, all major changes in early Ryukyuan history came from influxes of people from outside the islands. For example, during the early gusuku era (11th-12th centuries; I would date whole era as the 11th-15th centuries), migrants from Japan (mainly) and Korea swamped out the sparse indigenous population of Jōmon people (confirmed abundantly by recent ancient DNA studies). Moreover not only did the Sambyeolcho Rebellion have an impact on Okinawa, the gradual collapse of the Goryeo dynasty from ca. 1350 onward did as well. For example, the occupants of Urasoe Castle, operating under the name Satto, almost certainly came from Korea.

In the 14th and 15th centuries AD, the Ryukyu archipelago was unified into the Ryukyu Kingdom 92 and became part of the Chinese dynasty’s tributary system.

Because there is so much confusion about this era, please be more specific and accurate. Here is one possibility: From the 1370s, local powers in Okinawa engaged in official trade with China.

That is all I would say. The “Ryukyu archipelago” did not become unified under Shuri’s control until the 16th century. I realize the authors are not historians, but brief consideration of well-documented events such as Shuri’s invasion of Yaeyama in 1500, Shuri’s conquest of Kumejima in 1506, and Shuri’s conquest of Yonaguni (ether the 15-teens or 1520s) indicate the approximate time when all of the Ryukyu islands became, at least nominally, under Shuri’s control: the early 16th century.

＜Comments to Reviewer＞

We changed to “the Okinawa island” as suggested (p5, line 92).

Although the 125 Castle’s origins remain unclear, some historical and archaeological research has suggested that the main 126 palace was constructed in the mid-15th century (“Ankoku-zan Juka-boku no Kihi”: inscription on stone 127 monument built in 1427). It is presumed to have had a tile roof at this time because of the large number 128 of tiles excavated [12–14]. This main palace was destroyed by fire at least three times (AD 1453, 1660, 129 1709) (“Kyuyou”, a historical record written in AD 1743–1745).

The first sentence is odd. Although indeed we do not know the exact month or year that Shuri Came into being, 1405 or 1406 is very close. It is also important to note that the current, Chinese-looking version of the castle dates from about the 1490s or ca. 1500-1510 (part of the larger process of the kingdom coming into being). We know this in part because of the detailed descriptions of Shuri Castle by Koreans temporarily residing in Okinawa in the mid 1400s. They describe a castle that looked much different from its later iterations. These Korean descriptions of Shuri castle are widely discussed in the Japanese literature. Takara Kurayoshi, for example, includes a summary in many of his recent books and book chapters.

＜Comments to Reviewer＞

We do not agree with the reviewer's consideration because there are no citeable references. Therefore, we left it unchanged.

Because no clear traces of firing or production facilities for 13th- to 15th-century roof tiles have 391 been found in the Ryukyu Islands, it has not been possible to conclude whether they were produced in 392 this region or were imported. However, our results suggest that, at the very least, the technical 393 background of the production of these two roof tile styles at Shuri Castle (perhaps also at Urasoe Castle) 394 had similarities that can be detected in their color, paste characteristics, and internal structure. This might 395 mean that both the Kourai and Yamato roof tiles were made in the Ryukyus.

I agree with the final sentence, and indeed I would state the conclusion with greater strength. The first statement might be correct, although furnaces and kilns using Korean technology existed Tokunoshima from the late 11th century (to produce Kamuiyaki) and temple bells were cast in Shuri castle (and probably other locations) during the 15th century. So the basic knowhow probably existed during the 14th and 15th centuries. In any case, I think that the origin of the clay for the Korean-style roof tiles in Shuri castle, Urasoe Castle, and Katsuren has been known definitely for some time—an area near Nago, whose precise name escapes my mind. Again, my apologies for not providing specific page numbers, but see Pearson’s discussion for the details and relevant references in Japanese. So I suggest re-writing this paragraph to make the point more forcefully. The overall situation was technical knowhow (and almost certainly actual human technicians), which came from Korea and mainland Japan, but production taking place in Okinawa.

＜Comments to Reviewer＞

Our data did not reveal the origin if the technologies, nor did we identify its provenance. Therefore, we have taken the current expression and cannot say much more than that.

Although Ryukyu had the opportunity to import 461 techniques from Japan and China, it did not use these techniques but developed its own unglazed reddish 462 roof tiles. The reasons for this remain unclear, but as Ryukyu was in the complicated situation of being 463 under the control of Satsuma while having a feudal relationship with China [36], it is possible that they 464 did not prioritize the incorporation of a single country’s technology. The reddish color may also have 465 been a result of the feng-shui values that affected the policies of the Ryukyu Kingdom from the Qing 466 dynasty at that time [37]. This suggests that the background to the appearance of colored roof tiles in 467 Ryukyu encompasses not only the production technology and fuel resources, but also the socio-political 468 background. 469

The first sentence above is odd insofar as the original techniques and technology were clearly Korean (and possibly mainland Japanese) imports. But the authors transformed these outside technologies into something “uniquely” Ryukyuan in preceding paragraphs (a move probably best avoided).

More important: This whole paragraph is nothing but idle speculation by writers who clearly have only a superficial understanding of Okinawan history and society at this time. Again, I acknowledge that we cannot be experts in all fields, so I hate to sound critical. Nevertheless, a statement like “as Ryukyu was in the complicated situation of being 463 under the control of Satsuma while having a feudal relationship with China [36], it is possible that they 464 did not prioritize the incorporation of a single country’s technology” does not even make superficial sense—in addition to being nothing but speculation. Even general survey histories of Okinawa would note that during the early-modern era (ca. 1609-1879), the Ryukyu kingdom adopted extensive technologies from both China and Japan in fields like medicine, manufacturing, agriculture, and astronomy to name a few. Why, therefore, would it be reluctant to do so regarding roof tiles? And as for “The reddish color may also have 465 been a result of the feng-shui values that affected the policies of the Ryukyu Kingdom from the Qing 466 dynasty at that time,” please explain what you mean, specifically, with respect to fengshui. In any case, here and throughout all of the concluding sections, there is too much idle, unconvincing speculation.

＜Comments to Reviewer＞

We have removed the description of Feng Shui, as indicated (p 24, line 464).

However, sine this section discussed the color change in “Mincho style tiles”, the reviewer's suggestion “the original techniques and technology were clearly Korean imports” is not appropriate. We also discussed the relationship between the technology of neighboring countries and color change in Ryukyus, and only mentioned social context as one of possible explanation.

Important: the material from lines 493 to 520 is especially confusing. It all needs to be re-written. First, clarify exactly who has proposed which hypothesis. Second, explain each hypothesis better so that it is clear what the difference between them—and yours is. I have read these sentences three times and I cannot tell what exactly the authors mean.

＜Comments to Reviewer＞

In this paragraph, we discussed the transition (evolution) of tiles from a material culture perspective, based on the results obtained from the analysis of this study. In those discussions, we have also crearly presented past studies (by Dunnell, Boyd R, Richerson PJ) and then have proposed our new hypothesis. In other words, we discussed the history of Ryukyu (historical context), but rather presented a hypothesis regarding the evolution of material culture.

Also important: The authors periodically speculate about fuel shortages. For example:

In the case of transition pattern 2, it is considered 515 that the entire roof tile culture was subjected to the social stress caused by fuel shortages, but some of 516 the firing methods showed an especially strong reaction to it, which led to a transition in the color and 517 pore size of the tiles. In other words, the roof tile artisans responded to this stress by changing their 518 firing methods.

However, nowhere in the paper do the authors clearly present evidence of fuel shortages at certain times. Because this alleged fuel shortage is the only specific social factor that the authors claim (as opposed to vague speculation), it should be well explained. In other words, is there scholarly evidence for a significant fuel shortage during the early-modern era? If so present it (I cannot help here, because I am not aware of any such a claim in the literature). Or, if the physical analysis of the tiles either proves or suggests a fuel shortage, the authors need to explain the causal chain with clarity. Different firing methods? OK, that seems more reasonable and fairly obvious based on the evidence, but why does a different firing method mean that there was a fuel shortage? Perhaps it does, but if so, the authors need to explain it clearly.

＜Comments to Reviewer＞

The evidence on fuel shortages and their social context has already been clearly presented (line 452-457).

We 528 therefore provided a new perspective on the evolution of the Ryukyuan roof tile by proposing that it was 529 caused by external stresses and various reactions of traits to such stressors. As a result of our 530 multifaceted analysis, we conclude that the uniqueness of Ryukyuan roof tiles is the result of the 531 accumulation of external stresses in Ryukyu society and the various responses and evolutions of each 532 trait to such stressors.

The authors repeat some variation of this conclusion several times (indeed, too many times). It makes no sense. What exactly are “external stresses and various reactions of traits to such stressors”? Other than the alleged fuel shortage, the authors are at pains even to speculate about specific possibilities. Herein lies the paper’s major problem. The authors have conducted what appears to be an innovative analysis of roof tiles. Excellent. But they cast about, without success, trying to find some kind of larger significance for their work. Granted, coming up with something is difficult. I can think of only one thing.

＜Comments to Reviewer＞

Again, we have discussed the evolution of Ryukyuan tile from material culture perspective. The purpose of this paper is not to complement the history or origin of roof tiles in Ryukyu.

I suggest removing most or all of this material. A better approach, albeit a more modest one, might be to study the literature on the Korean roof tiles I mentioned at the start of this report and suggest how the authors’ method of tile analysis might shed useful light on this ongoing academic debate.

If indeed there are any relevant social “stressors” (a term I suggest the authors abandon) for which there is good evidence, then the authors should state their case with clarity and specificity, backed up with suitable citations to relevant scholarship.

＜Comments to Reviewer＞

" Stress" is a "cultural evolutionary” term already presented in Boyd and Richerson. This is also made clear in the manuscript (p. 26, line 498)

Writing

The quality of the writing needs improvement. Many sentences contain vague wording, and the authors use the passive voice excessively (e.g., “is presumed” “can be assumed” etc.). Organization at the level of paragraphs tends to break down in the latter third of the paper. Small spelling errors are common. And there is more. I am not going to go on at length about writing, leaving that matter to the editor. However there is one important issue, the spelling or Romanizing of Japanese words.

The authors spell Japanese words inconsistently. I provide some typical examples:

<(“Kyuyou”, a historical record written in AD 1743–1745).> The typical manner of spelling this word (球陽) would be Kyūyō. Note that both the “u” vowel sound and the “o” vowel sound are long. The conventional way to indicate long vowels is with macrons, ū and ō in this case. Writing out the vowels, while not conventional, does make sense and indeed it corresponds to the Japanese kana spelling きゅうよう. However, in this case, the Roman letter spelling should be Kyuuyou, not Kyuyou.

Looking at the bibliography, the Romanization of Japanese titles is inconsistent in numerous ways. The sound fu ふ is often spelled fu, but sometimes it appears as hu (as in huusui 風水). These sounds should all appear spelled as “fu” or “fuu.” Moreover, long vowels are indicated in some cases but not others, and capitalization is irregular. To take but one example, here is a book title:

Shuen to chushin no gainen de yomitoku higashi ajia no Etsu Kan Ryuu: Rekishigaku kokogaku kenkyu kara no shiza

Using the authors’ system it should appear as (note also that Higashi Ajia [East Asia], is a proper noun):

Shuuen to chuushin no gainen de yomitoku Higashi Ajia no Etsu, Kan, Ryuu: Rekishigaku, koukogaku kenkyuu kara no shiza

Or more conventionally:

Shūen to chūshin no gainen de yomitoku Higashi Ajia no Etsu, Kan, Ryū: Rekishigaku, kōkogaku kenkyū kara no shiza

Almost every entry in the bibliography contains Romanization errors, as do some words in the main text.

＜Comments to Reviewer＞

This paper has been reviewed in English by a professional agency. Variants in romanization were the result of following the original notation of the journals cited.

Attachment

Submitted filename: Response to Reviewers_20220918.docx

Click here for additional data file.^{(29.5KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0277560.r003

Decision Letter 1

Abbas Farmany

31 Oct 2022

A new perspective on the evolution of “Kawara” roof tiles in Ryukyu: a multidisciplinary non-destructive analysis of roof tile transition at Shuri Castle, Ryukyu Islands, Japan.

PONE-D-22-12204R1

Dear Dr. Aoyama,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Abbas Farmany

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: N/A

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #1: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #1: Yes

**********

6. Review Comments to the Author

Reviewer #1: In a few cases the authors correctly note that my comments missed or misconstrued something (e.g., regarding the term "stress"). Overall, however, the authors were dismissive regarding many of my substantive comments. I have no intention of spending any more of it on this matter except for one item:

It is also important to note that the current, Chinese-looking

version of the castle dates from about the 1490s or ca. 1500-1510 (part of the larger process of the

kingdom coming into being). We know this in part because of the detailed descriptions of Shuri Castle

by Koreans temporarily residing in Okinawa in the mid 1400s. They describe a castle that looked much

different from its later iterations. These Korean descriptions of Shuri castle are widely discussed in

the Japanese literature. Takara Kurayoshi, for example, includes a summary in many of his recent

books and book chapters.

＜Comments to Reviewer＞

We do not agree with the reviewer's consideration because there are no citeable references. Therefore,

we left it unchanged

Here are your references:

Higashionna Kanjun _Higashionna Kanjun zenshuu_, vol 3, pp. 81-85

Joseon Royal Records (朝鮮王朝実録), Entry nos. 125 & 126 (1462) in the edition edited by Ikeya Machiko, Uchida Akiko, and Takase Kyouko, _Chousen ouchou jitsuroku Ryuukyuu shiryou shuusei_. (The account is by Ryang Seong).

Takara Kurayoshi, _Ryuukyuu oukokushi no tankyuu_ (Ginowan, Youju shorin 2011), p. 32.

Makishi Youko, "Ryuukyuu oukokujuugoseiki chuuki ikou no kinaisei-tekina tokuchou to oujou girei no hensen," In _Okinawa bunka kenkyuu_, No. 38 (2012), p. 162.

--end--

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

**********

PLoS One. doi: 10.1371/journal.pone.0277560.r004

Acceptance letter

Abbas Farmany

7 Nov 2022

PONE-D-22-12204R1

A new perspective on the evolution of “Kawara” roof tiles in Ryukyu: a multidisciplinary non-destructive analysis of roof tile transition at Shuri Castle, Ryukyu Islands, Japan.

Dear Dr. Aoyama:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Abbas Farmany

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Appendix. Supplementary tables S1-S11.

(XLSX)

Click here for additional data file.^{(72.7KB, xlsx)}

S2 Appendix. Supplementary methods and figures S1-S10.

(PDF)

Click here for additional data file.^{(1.4MB, pdf)}

Attachment

Submitted filename: Reader report Shuri Castle roof tiles.pdf

Click here for additional data file.^{(664.5KB, pdf)}

Attachment

Submitted filename: Response to Reviewers_20220918.docx

Click here for additional data file.^{(29.5KB, docx)}

Data Availability Statement

All relevant data are within the paper and its Supporting Information files.

[pone.0277560.ref001] 1.ICOMOS. Gusuku Sites and Related Properties of the Kingdom of Ryukyu. 2000. [Cited 2021 June 27]. In: World Heritage List [Internet]. Paris: World Heritage Centre 1992. Available from: https://whc.unesco.org/en/list/972/.

[pone.0277560.ref002] 2.Esaka T, Serizawa C, Sakadume S. [New Japanese Archaeological Dictionarly] Shin, Nihon Kouko-gaku Jiten (in Japanese). Tokyo: New Science Co., Ltd.; 2020. [Google Scholar]

[pone.0277560.ref003] 3.Uehara M. [Study of “Kawara” Roof Tiles] Kawara wo Yomu (in Japanese). Tokyo: Kodansha Ltd.; 1997. [Google Scholar]

[pone.0277560.ref004] 4.Okawa K. [The Beauty of “Kawara” Roof Tiles: Unearthed Japanese History] Kawara no Bi: Umoreta Nihon Kodaishi (in Japanese). Tokyo: Shakai Sisousha Co., Ltd.; 1966. [Google Scholar]

[pone.0277560.ref005] 5.Okawa K. [Ancient “Kawara” Roof Tiles of Japan] Kodai no Kawara (in Japanese). Yougyou-Shi Hakubutsukan (Museum of Ceramic History); 1996.

[pone.0277560.ref006] 6.Uehara S. [Study of Ancient Roof Tiles in Ryukyu Islands] Ryukyu Koga no Kenkyu (in Japanese). Ginowan: Youju Shorin Ltd.; 2013. [Google Scholar]

[pone.0277560.ref007] 7.Ikeda Y. [Relationship between Jeju Island in Koreal and Ryukyu Islands based on Material Culture in the Goryeo Period] Bussitsu Bunka Kenkyu kara Mita Kankoku Cheju-tou to Ryukyu-rettou: Kourai Jidai wo Chushin to shite (in Japanese). Ryudai Asian Stud. 1998;2: 1–10. [Google Scholar]

[pone.0277560.ref008] 8.Sekiguchi K. [Transition and Historical Background of Ceramic Roof tile Manufacturing Based on Ancient Roof Tiles unearthed from “Honmaru remain” of Katsuren Castle (Katsuren-jo)] Okinawa ni okeru Zouga-gijutsu no Hensen to Sono-aida no Jijou: Katsuren-jou Honmaru-ato Shutsudo Koga wo Chushin ni shite (in Japanese). Kōkogaku zasshi (Journal of the Archaeological Society of Nippon). 1976;62: 222–239. [Google Scholar]

[pone.0277560.ref009] 9.Yamazaki S. [Study of Roof tiles in the Middle Age of Japan] Chu-Sei Gawara no Kenkyu (in Japanese). Tokyo: Yuzankaku, Inc.; 2000. [Google Scholar]

[pone.0277560.ref010] 10.Uehara S. [Transition of Mincho Style Roof Tiles Unearthed from “Nishi no Azana” Region in the Shuri Castle (Shuri-jo)] Shuri-jo Nishi no Azana chiku Syutsudo no Mincho-kei kawara to sono Suii (in Japanese). Nantō kōko (The journal of the Okinawa Archaeological Society) 1994;14: 153–186. [Google Scholar]

[pone.0277560.ref011] 11.Ishii R. Corpus and typological classification of eave-tile endpiece designs of Ryukyu modern roof tiles (in Japanese). Bulletin of the Department of Archaeology, The University of Tokyo. 2006; 20: 109–148. doi: 10.15083/00027626 [DOI] [Google Scholar]

[pone.0277560.ref012] 12.Ishii R. Transition of the roof on Shurijo castle Seiden. Josai Management Review (in Japanese). 2020;16: 157–181. doi: 10.20566/18801536_16_157 [DOI] [Google Scholar]

[pone.0277560.ref013] 13.Tsukada S. [Study of Historical Epitaph of Ryukyu Kingdom] Ryukyu-koku Hibunn-ki (in Japanese). Tokyo: Keigaku Shuppan Co., Ltd; 1970. [Google Scholar]

[pone.0277560.ref014] 14.Okinawa Prefectural Archaeological Center, editor. [Report of Archaeological Survey in the Main Palace of Shuri Castle (Shuri-jo)] Syuri jyou-shi -Seiden chiku hakkutsu chousa houkokusyo (in Japanese). Nishihara: Okinawa Prefectural Archaeological Center; 2016. [Google Scholar]

[pone.0277560.ref015] 15.Tei H. Kyūyō [edited by Kyūyō kenkyū-kai] (in Japanese). Tokyo: Kadokawa Shoten Publishing Co., Ltd.; 1974. [Google Scholar]

[pone.0277560.ref016] 16.Asato S. [Transition of the Main Place Platform of Shuri Castle (Shuri-jo)] Shurijyo seiden kidan no hennsen (in Japanese). Syurijyo kenkyu (Journal of Shuri-jo Study) 1996;2: 5–15. [Google Scholar]

[pone.0277560.ref017] 17.Chen K. Shi ryukyu roku [translated with notes by Harada N]. Ginowan: Youju Shorin Ltd.; 1995. (in Japanese). [Google Scholar]

[pone.0277560.ref018] 18.Takara K. [Academic Resource 1: Historical Resources about Official History] Shiryou-hen 1: Seishi kankei (in Japanese). In: Takara K, Yamazato J, Tomiyama K, Maehira F, Akamine M, Karimata S, editors. [Study of Disaster Based on the Historical Resources in Okinawa] Okinawa no Saigai Joho ni Kansuru Rekishi Bunnkenn wo Syutai toshita Sogo Kenkyu; 2008. pp. 69–97. http://ir.lib.u-ryukyu.ac.jp/handle/20.500.12000/8987. [Google Scholar]

[pone.0277560.ref019] 19.Ishii R. [Analysis of “Kawara” Roof Tiles Unearthed from “Watanji-mura” remain: Discussion about the Early modern Kawara in Ryukyu] Watanji-mura Ato Shutsudo Kawara no bunseki: Ryukyu kinsei kawara no teigi ni Kanshite (in Japanese). Bulletin of the Tsuboya Ceramics Museum 2015;15: 23–43. [Google Scholar]

[pone.0277560.ref020] 20.Uehara S. [Historical Relationship Between Ryukyu Islands and Surrounding Regions Based on the Study of Ancient Roof Tiles] Koga kara mita Ryukyu rettou no Kouryu-shi (in Japanese). Cultura antiqua 2000;52: 167–171. [Google Scholar]

[pone.0277560.ref021] 21.CIE. Improvement to Industrial Colour-Difference Evaluation (Vienna: CIE Publication No. 142–2001, Central Bureau of the CIE, 2001). [Google Scholar]

[pone.0277560.ref022] 22.Taylor, G. Python Colormath Module. Available online: https://pypi.python.org/pypi/colormath/.

[pone.0277560.ref023] 23.Aoyama H, Yamagiwa K, Fujimoto S, Izumi J, Ishikawa R, Kameshima S, et al. A new nondestructive approach to chemical analysis of potsherds using an X‐ray fluorescence microscope: Case study about the past pottery manufacture in the Yaeyama Islands. X-Ray Spectrom. 2018;47: 265–272. doi: 10.1002/xrs.2837 [DOI] [Google Scholar]

[pone.0277560.ref024] 24.Yamagiwa K, Fujimoto S, Aoyama H, Izumi J, Kameshima S. A possible new oldest pottery group in the Southern Ryukyu Islands, Japan: comparative analysis of elemental components of potsherds from the Shiraho-Saonetabaru Cave Site. J. Archaeol. Sci. Rep. 2019;26: 101879. doi: 10.1016/j.jasrep.2019.101879 [DOI] [Google Scholar]

[pone.0277560.ref025] 25.Segata N, Izard J, Waldron L, Gevers D, Miropolsky L, Garrett WS, et al. Metagenomic biomarker discovery and explanation. Genome biol. 2011;12: 1–18. doi: 10.1186/gb-2011-12-6-r60 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0277560.ref026] 26.Cheung S, Nakamoto M, Hamuro Y. NYSOL: a user-centric framework for knowledge discovery in big data. Int. J. Knowl. Eng. 2015;1: 214–218. doi: 10.18178/ijke.2015.1.3.037 [DOI] [Google Scholar]

[pone.0277560.ref027] 27.Urasoe City Board of Education, editor. [Survey Report of “Urasoe Youdore” vol. 2: About the Kawara Concentration Remain] Urasoe Youdore 2: Kawara-tamari ikou hen (in Japanese). Urasoe: Urasoe City Board of Education; 2005. [Google Scholar]

[pone.0277560.ref028] 28.Ishii R. Relationship with roof tile and Ryukyu history—kingship, legislation, interpretation, regional interaction (in Japanese). In: Nishimura M, Shinohara H, Okamoto H, editors. [Vietnam, Korea, and Ryukyu in East Asia Based on Archeological and Historical Study] Shuen to chushin no gainen de yomitoku higashi ajia no Etsu Kan Ryuu: Rekishigaku kokogaku kenkyu kara no shiza. Suita: Institute for Cultural Interaction Studies, Kansai University; 2012. pp. 161–177. [Google Scholar]

[pone.0277560.ref029] 29.Teruya Y, Hanashiro K, Yoza N, Ginoza T. [Scientific Study of Ancient Roof Tile in Shuri Castle (Shuri-jo)] Shurijo furu gawara nado no rikagakuteki kenkyu (in Japanese). Okinawa Industrial Technology Center Report. 1988;16: 51–75. [Google Scholar]

[pone.0277560.ref030] 30.Jordan MM, Montero MA, Meseguer S, Sanfeliu T. Influence of firing temperature and mineralogical composition on bending strength and porosity of ceramic tile bodies. Appl. Clay Sci. 2008;42: 266–271. doi: 10.1016/j.clay.2008.01.005 [DOI] [Google Scholar]

[pone.0277560.ref031] 31.Teruya Y, Ginoza T, Yoza N, Hanashiro K. [Study About Reproduction of Ancient Roof Tiles in Shuri Castle (Shuri-jo)] Shurijo seiden fukugen gawara nado no kaihatsu ni kansuru kenkyu (in Japanese). Okinawa Industrial Technology Center Report. 1988;16: 89–105. [Google Scholar]

[pone.0277560.ref032] 32.Morgan WR. Oxidation and loss of weight of clay bodies during firing: a report of an investigation conducted by the engineering experiment station. University of Illinois in cooperation with the Clay Products Association. 1936. Available from: http://hdl.handle.net/2142/4389. [Google Scholar]

[pone.0277560.ref033] 33.Sakurai K. Sulfur content in wood. Bull. For. & Fo. Prod. Res. Inst. 1986;335: 141–145. (in Japanese). [Google Scholar]

[pone.0277560.ref034] 34.Nakama Y. [History of Forest Cultivation System in Okinawa] Okinawa Rinya Seido Riyou-shi Kenkyu (in Japanese). Naha: Media Express Co., Ltd; 2011. [Google Scholar]

[pone.0277560.ref035] 35.Gou A, Wang J. The development of roof color in ancient China. Color Res. Appl. 2010;35: 246–266. doi: 10.1002/col.20564 [DOI] [Google Scholar]

[pone.0277560.ref036] 36.Okamoto H. Structural transformation of Ryukyu Kingdom in the 17th and rarly 18th centuries: as an intersection of cultural interaction. In: Masanari N, Sato M, Kimura M, Okamoto H, editors. Cultural reproduction on its interface: from the perspectives of text, diplomacy, otherness, and tea in East Asia. Suita: Institute for Cultural Interaction Studies, Kansai University; 2010. pp. 3–17. [Google Scholar]

[pone.0277560.ref037] 37.Dunnell RC. Style and function: a fundamental dichotomy. Am Antiq. 1978;43: 192–202. doi: 10.2307/279244 [DOI] [Google Scholar]

[pone.0277560.ref038] 38.Boyd R, Richerson PJ. Culture and the evolutionary process. Chicago. University of Chicago Press; 1988. [Google Scholar]

[pone.0277560.ref039] 39.Gilbert SF, Epel D. Ecological developmental biology: the environmental regulation of development, health, and evolution. 2^nd ed. Sunderland: Sinauer Associates, Inc. Publishers; 2015. [Google Scholar]

PERMALINK

A new perspective on the evolution of “Kawara” roof tiles in Ryukyu: A multidisciplinary non-destructive analysis of roof tile transition at Shuri Castle, Ryukyu Islands, Japan

Hiroaki Aoyama

Kaishi Yamagiwa

Wataru Taira

Takeshi Kon

Roles

Abstract

Introduction

Fig 1. Location of Shuri Castle and exterior view of Main Hall.

Materials and methods

Site information

Materials

Image scanning and color distance analysis

X-ray fluorescence microscopic analysis

X-ray computed tomography (CT)

Pores identification, counting, and measurement

Eight values for internal microstructure features

Consideration of data reliability

Data analysis

Results

Quantitative color differences analysis

Fig 2. Color distance analysis of roof tiles.

Comparative analysis of chemical components

Fig 3. Chemical component analysis of roof tiles.

Physical feature analysis for the internal microstructure of the roof tiles

Fig 4. Analysis of microstructural features.

Correlation analysis of the factors contributing to the color change of the ancient roof tiles

Discussion

1. Transition from Kourai and Yamato to Early Minchō style

2. Unique color transition after Minchō style and its background

3. New perspective on the evolution of Kawara in Ryukyu

Fig 5. Summary of roof tile evolutionary process.

Fig 6. Two hypotheses to explain the different transition patterns.

Conclusion

Supporting information

Acknowledgments

Data Availability

Funding Statement

References

Decision Letter 0

Abbas Farmany

Roles

Author response to Decision Letter 0

Decision Letter 1

Abbas Farmany

Roles

Acceptance letter

Abbas Farmany

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases