The traditional account comes from Sima Qian, writing around 100 BCE. Laozi was an archivist of the Zhou court whom Confucius consulted on ritual; disgusted with the decline of the dynasty, he rode west, and at the frontier pass a gatekeeper asked him to set down his teaching before he vanished. Sima Qian himself hedges this, offering rival identifications and conceding that nobody knows what became of the man. Most modern scholars therefore treat Laozi as legendary or composite, and the book as an accretion of sayings from several hands, assembled in the fourth or third century BCE.
Two archaeological finds transformed the argument. In 1973, tomb three at Mawangdui near Changsha, sealed in 168 BCE, produced two nearly complete copies on silk. Both put the De section before the Dao section, reversing the received order, and both are close enough to the received wording to show that the text was substantially fixed by the early Han.
In 1993, tomb one at Guodian in Jingmen, Hubei, dated to around 300 BCE, produced three separate bundles of bamboo slips carrying Laozi material: some 2,000 characters corresponding to about thirty-one chapters of the received eighty-one, in a quite different order, with nothing at all from the closing chapters.
The two finds pull in different directions and both are useful. Guodian proves the material existed well before 300 BCE, killing the late datings that had placed the whole book in the third century. It also shows that at that point it was not yet a fixed eighty-one-chapter work with a settled sequence. Mawangdui shows how quickly it became one.