The Death of Confucius
The traditional death date of Confucius is 0479. It is given by the interpolated entry in the version of the Chun/Chyou text which is associated with the Dzwo Jwan. Maspero proposed that the actual date might have been a quarter of a century later, a proposal noted with interest by Waley and recently endorsed by Riegel. We have reviewed the evidence for the the dates and ancestry of Confucius, and we find that the traditional death date turns out to be generally sound.
The Confucius date problem cannot be solved in isolation; there are several related problems which must be considered at the same time. We here note some of the difficulties which a valid solution of this problem must resolve, and some of the implications which our solution suggests.
1. Rulers. No source, anecdote, or tradition down through Han, including the anti-Confucian anecdotes of pre-Han and Han times, describes Confucius as living under any Lu ruler later than Ai-gung (reigned 0494-0468). If Confucius had spent his last and thus most influential years under Dau-gung (r 0468-0432), we would expect that fact to have left some trace in the record. There is no such trace.
2. Career. Analects tradition shows Confucius as (a) delayed in his career by early poverty, (b) seeming to have direct knowledge of Jau-gung (r 0541-0510), (c) having interviews with Ding-gung (r 0509-0495) and also with Ai-gung (r 0494-0468), and (d) living at least to the age of seventy. None of these stories need be true; no story in any ancient text need be true. But these stories together imply a consistent background assumption by the Analects writers. That assumption can be summarized thus: Confucius's early career years, that is, until roughly the age of thirty, were spent under the reign of Jau-gung; he reached career maturity, perhaps about the age of fifty, under Ding-gung but not earlier, and he died at the age of seventy or a little later during the reign of Ai-gung. Let us designate the unknown year of his death as X. A point twenty years before his death (X-20, or his 50th year) should thus fall within the reign of Ding-gung, and a point forty years before his death (X-40, or his 30th year) should fall within the reign of Jau-gung. The former condition is the more restrictive. Arithmetic will show that the range of death dates which satisfy it is 0489-0475 if Confucius's age at death was seventy, and 0489-0472 if (as some sources hold) it was seventy-three.
This is the general trend of the earliest anecdotal and inferential evidence about Confucius. Roughly speaking, it identifies a range of 0489-0472 for the probable death date of Confucius. This range includes the traditional death date 0479. It however excludes all years from 0468 onward, which are the ones that would agree with the Maspero proposal. This alone is sufficient to discredit the Maspero proposal. From here on, we will consider it as discredited, and consider the consequences of that decision, which turn out to be compatible with a wide range of additional evidence.
3. Death. Again by explicit Analects tradition, which is agreed to be on the whole the best available tradition, Confucius's son Bwo-yw predeceased him. Later family tradition (Kungdz Jya-yw) is that Bwo-yw died in his own fiftieth year.
4. Confirmation. The claim that Bwo-yw predeceased his father is borne out by references in ritual texts to Confucius's death and funeral. In all of these, without exception, it is not family members, but disciples, who are in charge of the funeral arrangements, and serve as the chief mourners. If Confucius's son had survived him, and was of adult years at the time of his death, then the son would presumably have taken the leading role on the occasion of Confucius's funeral. If the traditional date of Confucius's death (0479) is right, and the points noted above suggest that it is, then Bwo-yw must have died not later than 0480. Accepting instead the range of Confucius death dates produced by argument #2 above, the extreme possibility is that Bwo-yw died not later than 0473. This point is vital for the traditional view of the transmission of Confucian doctrine, for which see the following note on Dz-sz.
5. Birth. In Shr Ji 47, Confucius's kin in succeeding generations are given as Bwo-yw and then Dz-sz. The implication is that Dz-sz was the son of Bwo-yw. Even on the extreme assumption that Dz-sz was conceived in the last year of Bwo-yw's life, and born after Bwo-yw's death, Dz-sz's birth (by #4) can be no later than 0472. The view of SJ 47 is evidently that Confucius's teachings descended in the family, with no interlude of disciple transmission. The Kung Tsungdz (Latter Han) puts Dz-sz's birth in 0479, early enough for a few months of overlap with his grandfather, but not allowing for any very complicated contact between them. That text actually shows Dz-sz and Confucius in conversation at a high philosophical level. This seems so strained as to be self-refuting. Chyen Mu, for one, dismisses at as evidently wrong.
6. Maturity. Several stories in the Mencius portray Dz-sz in high office. In all those stories, the Lu ruler whom Dz-sz advises is Mu-gung (r 0410-0378). The stories themselves date from no later than the first half of the 03c, and are thus not inordinately far removed from the presumed facts. A brief and previously unknown text recovered from the Gwodyen 1 tomb reinforces these bits of evidence, from a source outside the Mencius and (as far as we know) independent of the Mencius. That text also shows Dz-sz advising Mu-gung, and indeed admonishing him. The probable date of the Gwodyen tomb is within a decade of c0288. This story is thus an independent voice direct from the probable period of the Mencius (most of which text consists of writings of the posthumous Mencian school or schools, and thus also dates from the first half of the 03c). Sywndz in criticizing Mencius brackets him with Dz-sz, thus seeming to imply that Mencius was part of a lineal tradition going back to Dz-sz. If so, Mencius would have known the school traditions concerning Dz-sz. The earliest testimony about Dz-sz is therefore that of the Mencius together with the compatible testimony of the Gwodyen document, and on the hostile evidence of the Sywndz, the Mencius testimony may be well informed. The Mencius and the Gwodyen document together consistently suggest that Dz-sz served no other ruler than Lu Mu-gung. On the most natural assumptions as to career patterns, Dz-sz should thus have reached his fiftieth year within the reign of Mu-gung (0410-0378). This requires that Dz-sz's birth year be in the span 0459-0427. This is not within the range of the possibilities available from #5 preceding. To reconcile the two, it is necessary to make less natural assumptions, such as a very delayed public career, and/or an extremely long life for Dz-sz. The latter is the Kung Tsungdz solution, and as noted above, even that solution does not provide a humanly plausible scenario for Dz-sz. It is very likely that the Kung Tsungdz is motivated by a wish to provide for a continuous family transmission of Confucius's doctrines. That is, the Kung Tsungdz scenario is a piece of Kung family apologetics.
If we dismiss it as such, we then need to consider what other transmission possibilities there are.
7. Stewardship Gap. We have seen that the Dz-sz story presents difficulties at both its ends. Orthodoxy requires that he have reached years of discretion within Confucius's lifetime (implying a birthdate not long after the beginning of the 05th century), and have been able to hold office as late as Mu-gung (implying personal vigor near the end of the 05th century or even later). If either of these claims is abandoned, the orthodox account fails, and an intervening stage in the lineage of the Confucian school must be assumed between Confucius and Dz-sz. Given the early testimony about Dz-sz serving Mu-gung (the 03c portions of the Mencius), and given the lack of early testimony that Dz-sz was born within Confucius's lifetime and received teaching from him (the Analects; the accounts of Confucius's funeral in the ritual texts; the fact that Dz-sz is not listed in the inventories of disciples, though Bwo-yw is), it seems that the point which must be abandoned is point #5, the claim that Dz-sz was born early enough to have served as a transmitter of Confucius's teachings. Once we are rid of that complicating claim, the better evidenced tradition that Dz-sz served Mu-gung (point #6), may be allowed to stand.
8. Disciple Phase. If the previous argument is accepted, it unavoidably requires that we posit a transmission stage, and thus a Confucian school leadership stage, between Confucius and Dz-sz. If there were kinsmen available for this leadership, the Kung Tsungdz would most probably have made use of them. It follows that the interim leaders must have been other than kinsmen. In this context, it is proper to ask if any early tradition provides information about leadership in matters Confucius in the years after the death of Confucius. One datum, already noticed, is that the ritual texts portray only disciples as involved with Confucius's funeral and mourning period. Another is that the early Analects (in LY 8:3-7) seemingly portrays the disciple Dzvngdz, at the time of his death, in the role of head of the Confucian school. All this suggests a disciple-dominated milieu in the years directly following the death of Confucius. We are led by the tendency of all this evidence to reject the Kung family claim of unbroken Kung family stewardship, and to posit a stage of disciple stewardship and transmission between Confucius (died not later than 0472) and Dz-sz (born not earlier than 0459, and reaching career maturity about the end of the 05c).
9. Choices. The decision about Confucius's death date thus comes down to the question of continuity or discontinuity between Confucius and Dz-sz in the Confucian school lineage. Maspero, following Shr Ji 47 and the Kung Tsungdz tradition or some scenario compatible with it, chooses to reject point #1 (the traditional death date of Confucius). This results in a continuity scenario, but creates problems with absolutely all of the other evidence. It thus appears to be a pious but untenable solution. The best evidence rather indicates that point #5 (the direct successor role of Dz-sz) is the unsound position. Rejecting point #5 results in a discontinuity scenario, but a discontinuity scenario which is compatible with all the early evidence. We complete that scenario by accepting point #8: there was a period of disciple headship in Confucius's successor school.
10. Verification. If this is true, then at some point the Confucian school must have undergone a rectification of its own historical memory, in which its actual disciple heritage would have been replaced or discredited. As it happens, we can see just such a rectification process at work in the disciple pantheon of LY 11:3, whose valuations of disciples are wildly at variance with those explicit in earlier chapters of the Analects (LY 5-6). This restructuring of the disciple tradition in turn seems to be parallel with the drastically different values expressed in the respective portions of the Analects, with rvn ("otherness") being often mentioned in LY 4-9 (despite the unsuccessful denial of this fact in the interpolated LY 9:1, on which see Brooks Word), but entirely absent from chapters in the vicinity of the LY 11 disciple pantheon (LY 10-11), being replaced there by a focus on li ("behavioral propriety").
The final position which we reach by the above argument suggests a consistent historical picture which has been disturbed by one factor: the desire of the Kung family to assert an unbroken Kung stewardship for Confucius's ideas. That desire is intelligible as a position profitable to the Kung family, but it seems to have no claim on general acceptance. We therefore reject it.
For a more detailed argument on these and other points, see the references given above plus our Mentorship monograph, an abridged version of which appears as Appendix 4 in The Original Analects.
Postscript 2008: Nothing is more consistent in the anecdotal literature, including the anecdotes now contained in various documents in the Li Ji ritual collection, than the picture of Confucius's disciples presiding over his funeral. For this to have taken place, it would seem to be ritually necessary, not only that Confucius's son Bwo-yw, who would normally have been the chief mourner, should have predeceased Confucius (and this detail both family tradition and the Analects text do supply), but for Bwo-yw in turn to have left no male heir who could have taken on that function. We have seen that Dz-sz cannot have been Bwo-yw's son. If he was instead Bwo-yw's grandson, then Dz-sz's father (Bwo-yw's son) would most likely have been in his middle twenties at the time of Confucius's death, and perfectly capable of taking part in the funeral rites. Then either (a) he too had died before 0479, or (b) Dz-sz was not Bwo-yw's grandson either, but some other Kung of that generation. It is the latter alternative that better accounts for the evidence, including the sometimes venomous tone of Analects 11 when it is dealing with the individuals who made up, or were influenced by, the early disciple transmission.
13 May 2000 / Contact The Project / Exit to Results Page