Kansei Engineering as a Tool for the Design of Traditional Pattern

2.1 Screening of lotus pattern in Mogao Cave of the Tang dynasty

The Tang dynasty is the golden age of the development of Dunhuang mural art; plants as the main decorative patterns in the Tang dynasty also began to perfect [20,21,22,23]. A large number of algae lotus patterns were collected through literature review and field research methods, and five Dunhuang cultural research and design-related experts were invited to screen the patterns, remove similar lotus patterns, and retain the different patterns. Finally, 14 lotus patterns were selected as shown in Figure 1.

Figure 1

Representative lotus pattern sample in the caisson of the Mogao Grottoes in the Tang dynasty.

2.2 Deconstruction of lotus pattern design elements

Through expert interviews and focus group discussions and referring to relevant research methods, a comprehensive analysis of 14 caisson lotus samples was carried out. The algae lotus pattern was disassembled into five design elements, such as the lotus center structure, the petal shape, the combination of the flower heart and the outer petals, the combination of the outer petals, and the position of other decorations. To further refine the design and obtain more accurate conclusions, these five design elements are subdivided into various form items. For example, the structure of the lotus center is divided into cross type, 米 type, and mission type. Petal morphology is divided into flat petal, curl petal, moiré petal + flat petal, moiré petal + peach-shaped petal, moiré petal + hook petal + peach-shaped petal, flat petal + hook petal + moiré petal, curl petal + hook petal + moiré petal, and flat petal + peach-shaped petal + moiré petal + hook petal. The combination of the flower heart and the outer petals is divided into three forms, such as the flower heart connected to the outer flower petals, the flower heart separated from the outer flower petals, and the single layer petal. The combination of petals is divided into no outer layer, multiple layers closely connected, and multiple layers sparsely connected. The position of the decoration is divided into the flower heart, around the lotus, no decoration, and the flower heart + around the lotus. The deconstruction of the lotus pattern design elements is shown in Table 1.

2.3 Perceptual evaluation experiment

Through extensive collection of pattern perceptual adjectives in magazines, books, research literature, and other materials related to patterns and then inquiry and survey in a small range, invite five design-related professionals to select adjectives that appropriately describe the pattern features from the collected adjectives, and to remove overlapping and inappropriate adjectives, 7 pairs of perceptual adjectives were selected from the 40 pairs, which are lively–dull, simple–complex, neat–disordered, tight–loose, cabinet–atmospheric, move–static, and unique–ordinary.

This study used the semantic difference method to conduct evaluation experiments. The selected lotus pattern samples and perceptual adjectives are combined using a Likert 5 scale. If we take the perceptual adjective “living–dull” as an example, 1 point is very lively, 2 points are a bit lively, 3 points are neither lively nor dull, 4 points are a bit dull, and 5 points are very dull.

2.4 Data processing and analysis

The results were processed and analyzed using mean value processing and correlation analysis, multiple linear regression analysis, and other mathematical methods in SPSS 24.0 software.

3.1 Data processing of perceptual evaluation of caisson lotus pattern in Mogao Caves of the Tang dynasty

In this survey, online questionnaires were distributed to young consumers aged between 20 and 35 years. A total of 60 questionnaires were sent out. A total of 57 valid questionnaires were returned, and the recovery rate was 95%. The results of 57 surveys were processed by Excel software, and the mean value of each sample under each perceptual vocabulary was obtained as shown in Table 2.

To further analyze the main feelings of each lotus pattern and the differences between them, the highest and lowest scores of each lotus pattern are listed in Table 3.

As can be seen from the Table 3, samples 1, 5, 6, 7, and 12 have a neat feeling; samples 1, 7, and 10 have a static feeling; samples 2 and 4 have a lively feeling; samples 2, 3, 4, 5, 9, 11, 12, and 13 have a complex feeling; samples 3 and 8 have a feeling of movement; samples 5, 9, and 11 have a tight feeling; samples 6 and 12 have an atmospheric feeling; sample 8 has a loose feeling; sample 10 has a cabinet and ordinary feeling; sample 13 has a unique feel; and sample 14 has a simple and dull feel. At the same time, all the samples in Table 3 did not show the feeling of disordered. In Table 2, the samples with disordered feelings were 8 and 10, the scores were 3.14 and 3.07, and the scores were lower. It can be said that all the samples have almost not disordered.

3.2 Correlation analysis between perceptual vocabulary and design elements

Then using SPSS 24.0 software, the design elements are independent variables, and the perceptual vocabulary evaluation value is the dependent variable. First, the correlation analysis is carried out to find out the design elements that are significantly related to the sensible vocabulary. As can be seen from Table 4, the design element that is significantly associated with the perceptual adjectives lively–dull is the petal shape; the Pearson correlation coefficient is −0.568, with a significant value of 0.035, followed by the combination of the flower heart and the outer petals; the Pearson correlation coefficient is 0.381. The design elements related to the simple complicated are petal form, the position of decorations, and the combination of outer petals; the Pearson correlation coefficients are 0.711, 0.380, and −0.304, respectively. Petal shape is the most relevant with the significant value of 0.004. The design elements related to neat–disordered are the combination of the flower heart and the outer petals, the lotus center structure, the combination of the outer petals, and the petal shape. The Pearson correlation coefficients are −0.590, 0.450, 0.391, and 0.367, respectively. The petal shape is closely related to the perceptual vocabulary tight–loose significant, with a significant value of 0.006 < 0.01, the combination of the flower heart and outer petals, and the combination of the outer petals are also related to it; the Pearson correlation coefficients are 0.469 and −0.342, respectively. The petal form of the design element is also significantly correlated with the cabinet–atmospheric perceptual vocabulary, with a significant value of 0.013 < 0.05, followed by the combination of the flower heart and the outer petals; the Pearson correlation coefficient is 0.339. The design elements related to move–static are the position of decorations and the petal shape; the Pearson correlation coefficients are 0.565 and −0.357, respectively. The petal shape and the combination of the outer petals are related to unique–ordinary, and the petal shape is significantly correlated with it; the Pearson correlation coefficient is −0.596; and the significant value is 0.022.

3.3 Multiple linear regression analysis of perceptual evaluation data

Then multivariate linear regression analysis is carried out to finally find out the correspondence between the design elements of lotus pattern and the sensible vocabulary and establish the regression model (as shown in Table 5).

As can be seen from Table 5, all models have a good fitting degree, and the values of R² are all above 0.78. Among the seven pairs of perceptual vocabulary, the design element of the petal shape has influence on the other six pairs of perceptual vocabulary, except that it has no effect on the neat–disordered perceptual vocabulary. The combination of the flower heart and the outer petals has influence on the neat–disordered, tight–loose, and unique–ordinary. The position of decorations has an impact on the simple–complex and move–static two sets of perceptual vocabulary. The structure of the lotus center only affects neat–disordered.

3.4 Formulation of design element combination coding table

With regard to the design elements in the model of Table 4 and the samples in Table 3, to find out which design elements will affect the perceptual vocabulary, according to the design element coding in Table 1, a design element coding combination table corresponding to the perceptual vocabulary is formulated. Taking the perceptual vocabulary lively–dull as an example, in the model, the petal shape has an influence on it; the samples with lively features in Table 3 are 2 and 4; the samples with dull features are 14; by observing the petal shape of samples 2, 4, and 14, it can be concluded that the petal shape encoded as 27 and 26 gives a lively feeling, and the petal shape encoded as 21 gives a dull feeling. According to this method, the design element codes corresponding to each pair of perceptual vocabulary are specifically obtained, as shown in Table 6.

By comparing the design element coding in Table 6 with the sample picture 1, it can be concluded that the petal shape of codes 26 and 27 has a lively sensation, and the petal shape of code 21 has a dull feeling. The design element combination codes 254353 and 284253 present a complex feeling, while the samples 9, 11, and 12 corresponding to code 284253 happen to be the common pattern of the Mission flower and the Baoxiang flower in the Tang dynasty caisson. This shows that modern young consumers think that the traditional lotus pattern, such as the typical Mission flower and Baoxiang flower, gives people a complex feeling. The samples 2, 3, 4, and 13 corresponding to the codes 274251, 264252, 224154, and 244151 have no complex pattern characteristics, such as the Mission flower and the Baoxiang flower, but the lotus pattern has a decorative object, which also gives a complex feeling. Among all the samples, only the 14th sample has a simple character. By analyzing the shape of the petals, the combination of the outer petals, and the position of the ornament, it can be seen that the two-layered different-sized lotus petals combined with the hear decorate with a simple ornament; it gives a simple feeling. For neat–disordered sensuous vocabulary, all samples are presented in a neat sense; by analyzing the center structure of the sample and the combination of the flower heart and the outer petals, it can be found that the center structure of the lotus pattern exhibits a cross shape, a 米 shape, and a mission shape, which itself has a symmetric and balanced characteristic; the outer petals then radiate outward from small to large layers, but it is only when the flower heart is combined with the outer petals, it will give a little disordered feeling. The design element combination code 283242 has a tight feeling, corresponding to the physical samples 9 and 11, and it can be seen that the complex multilayered Mission flower and the Baoxiang flower's petal shape give a tight feeling. The difference between the design elements of codes 253143 and 243143 is the shape of the petals. By analyzing samples 5 and 8, it can be seen that the complex multilayered petal shape of sample 5 gives a tight feeling, while the less number of petal layers and shape of sample 8 give a feeling of looseness. In the design element coded 2331, the lotus flower heart is connected to the simple petal shape giving a cabinet feeling. The complex petal shape, the flower heart does not connect with the outer petals, such as the samples corresponding to the codes 2532 and 2832, gives an atmospheric feel. By comparing the coding of the move–static vocabulary design elements with the sample, it can be seen that the curl petal produces a rotational illusion due to the inward curling of the petals; at the same time, the movement of animals or characters in the center of the flower or the surrounding area will also give people a sense of movement. Analysis of the design elements of codes 2441 and 2342 shows that the single-layered combination of Moiré petal and peach-shaped petals gives a unique feeling; Moiré petal with similar shapes and sizes that are closely arranged in multiple layers gives a common feeling.

3.5 Verify the regression model of perceptual vocabulary and design elements

To verify the reliability and validity of regression model of perceptual vocabulary and design elements, five lotus patterns were randomly selected from the Mogao Caves in the Tang dynasty caisson to verify the accuracy of the correspondence between sensual vocabulary and design elements (as shown in Figure 2).

Figure 2

Verification samples.

First, five lotus patterns are used as verification samples to analyze the design elements, bring the analysis results into the model corresponding to each group of perceptual vocabulary, and calculate the perceptual predictive value of the verification sample. Then, a perceptual questionnaire was created on the network for this five verification samples, and it was distributed to 20 target consumers to conduct a perceptual evaluation about them. Based on the results of the survey, the average of the evaluation results of five verification samples was calculated. To verify the accuracy of the model, the relative error method is used to normalize the perceptual prediction value and the measured value of the questionnaire (as shown in Table 7).

From Table 7, it can be seen that in addition to the two groups of perceptual words “Simple” and “Tight,” the relative error of the results obtained by the verification sample 4 is 5.3%; the “Tight” group of perceptual words—the relative error of the results obtained in sample 3 is 5%; other perceptual words’ relative errors are below 5%; and the average relative error of each group of perceptual words is below 4%. That is to say, the model corresponding to each group of perceptual vocabulary obtained in this study has certain credibility and can predict consumers’ perception about the lotus pattern of the caisson in the Tang dynasty.