Abstract
In this study, we compare statistical properties of ancient and modern Chinese within the framework of weighted complex networks. We examine two language networks based on different Chinese versions of the Records of the Grand Historian. The comparative results show that Zipf’s law holds and that both networks are scale-free and disassortative. The interactivity and connectivity of the two networks lead us to expect that the modern Chinese text would have more phrases than the ancient Chinese one. Furthermore, by considering some of the topological and weighted quantities, we find that expressions in ancient Chinese are briefer than in modern Chinese. These observations indicate that the two languages might have different linguistic mechanisms and combinatorial natures, which we attribute to the stylistic differences and evolution of written Chinese.
Highlights
With the publication of seminal works [1,2], complex networks have become a popular research topic in statistics, sociology, biology, and other fields over the past 20 years [3,4,5,6]
Investigation of real-world networks of various kinds has led to the study of many complex systems from the viewpoint of complex networks including ecology [7], computation [8], coding [9], cell and molecular biology [10], protein [11], neuroscience [12,13,14], human brain [15,16,17], and communication networks [18]
We explore in detail the architectures of the two language networks through weighted network representations of the ancient and modern Chinese versions of the Records of the Grand Historian and find that they exhibit different behaviors
Summary
With the publication of seminal works [1,2], complex networks have become a popular research topic in statistics, sociology, biology, and other fields over the past 20 years [3,4,5,6]. Some scholars have explored language networks in those terms, with many important linguistic properties being discovered by Li [20] and Liang [25] They discussed only undirected and unweighted networks, and no comparisons have been made regarding the characteristics of different dialects or stages of Chinese. In our research, the main purpose is to investigate the similarities and differences between the language networks for ancient (ALN) and modern (MLN) Chinese. Both networks are constructed in the same manner and treated as weighted directed graphs.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
