The debate over the authenticity of the Liezi goes back centuries. Many proofs have been offered for and against the claim that it is an authentic pre-Qin work, but by far the most convincing form of proof has been the evidence provided by the language of the Liezi itself. In the past, however, the strength of this evidence has not been fully appreciated; as a result, this linguistic evidence has not received the attention it deserves.
There are, of course, substantial studies of linguistic issues in the Liezi, where previous scholars have cited a number of linguistic forms in Liezi which do not appear in pre-Qin texts. (“Linguistic forms” here include words, phrases, grammatical particles, and syntactic forms). However, many late forms in the Liezi have remained unnoted; in fact, the total number of these is many times greater than previous scholars have observed.

This study thus undertakes a complete re-examination of the language of the Liezi to determine exactly how much of the Liezi text can be classified as postdating the pre-Qin era. Based on this new examination, it then offers a comprehensive re-assessment of the actual date of the composition of the Liezi. In this reassessment, it is important to note that there are many passages in the Liezi that duplicate authentic Old Chinese texts. However, these are of little help in determining the features which are characteristic of the Liezi text and the linguistic era which these features reflect. Thus this study focuses on the portions of the Liezi which have no parallels in Old Chinese works. Note that here Old Chinese is used as a general cover term for both pre-Qin and Western Han texts.

Summarizing the results of this study, the Liezi contains over 260 linguistic forms unseen in any Old Chinese texts, approximately 90 forms which first appear in Western Han works, and approximately 40 forms which were rarely used in pre-Qin texts. In terms of frequency, forms in the first group appear 403 times, forms in the second 138 times, and forms in the third 181 times, a total of 722 instances. Based on this evidence, we infer that much of the non-parallel portion of Liezi (the passages which have no parallels from authentic Old Chinese works) was written between the Eastern Han dynasty and the Wei and Jin dynasties.
In these non-parallel passages, linguistic forms unseen or scarce in pre-Qin works appear in large numbers throughout; they essentially appear in all passages of any length. This is strong evidence that, excluding the pre-Qin parallels, most of the Liezi is not a pre-Qin work. In addition, the distribution of linguistic forms which are unseen in Old Chinese texts throughout the portion of the non-parallel Liezi text also offers strong evidence that most of this portion postdates the Old Chinese era; some of these passages are no earlier than the Western Jin era, and at least one chapter is most likely from the Eastern Jin.

This study concludes that of the eight chapters of the Liezi, the majority were independent works, compiled or written by more than one person. The study also concludes that the texts in the Liezi we have today were compiled by Zhang Yi (張嶷), the grandfather of Zhang Zhan (張湛). The material he used to compile the Liezi came mainly from the Wang and Zhang families and their friends and relatives. Based on the evidence, the so-called Liu Xiang “Liezi Xinshu Mulu” is also most likely a forgery.