Enough, already (you may be saying), with this modern stuff. What about a nice ancient question? The nice ancent question will tend to show that sometimes simple methods are more revealing than new and sophisticated ones.
For eleven of the twelve reigns of Lu princes for which data are provided in the Spring and Autumn chronicle (the other reign only lasted two years, and may be atypical), and for all states represented in the record, we want to know the relationship, if any, between the number of diplomatic missions to other states and the number of acts of war. We convert the numbers of diplomatic and military events to percentages, in order to equalize the unequal lengths of the reigns. We then fill in our table, as on the previous page, calculate the mean for the War and Diplomacy events, subtract it from each value as before (to get its difference above or below the mean), and then calculate Columbs 6, 7, and 8. The positive events are again shown in red:
1 2 3 4 5 6 7 8 Reign (1st Year) x (Dip) y (War) (x-mx) (y-my) (x-mx)(y-my) (x-mx)² (y-my)² 01 Yin 0722 1.82 1.64 +0.13 -0.45 -0.0585 0.0169 0.2025 02 Hwan 0711 2.33 0.78 +0.63 -1.31 -0.8253 0.3969 1.7161 03 Jwang 0693 0.75 1.34 -0.94 -0.75 +0.7050 0.8836 0.5625 05 Syi 0659 1.39 2.55 -0.30 +0.46 -0.1380 0.0900 0.2116 06 Wvn 0626 2.50 2.11 +0.81 +0.02 +-0.0162 0.6561 0.0004 07 Sywaen 0608 1.50 2.61 -0.19 +0.52 -0.0998 0.0361 0.2704 08 Chvng 0590 2.39 2.78 +0.70 +0.69 +0.4830 0.4900 0.4761 09 Syang0572 2.35 2.48 +0.66 +0.39 +0.2574 0.4356 0.1521 10 Jau 0541 1.38 1.25 -0.31 -0.84 +0.2602 0.0961 0.7056 11 Ding 0509 1.47 2.60 -0.22 +0.51 -0.1122 0.0484 0.2601 12 Ai 0494 0.69 2.88 -1.00 +0.79 -0.7900 1.0000 0.6241 SUM (S) 18.57 23.02 a = -0.3010 b = 4.1497 c = 5.1815 mean (S/11) 1.69 2.09
after which, by formula, Coefficient of Determination (D) is:
D = a² / bc = -(0.3010)² / (4.1497)(5.1815) = -0.090601 / 21.5017 = 0.0042
The minus sign means that, over all, diplomacy and war are inversely related: the more of one you have, the less you have of the other. But the absolute value of 0.0042, or zero to within two decimal places, tells us that neither that nor any other linear relationship is very strong. The test has thus failed to point to anything of interest for the Spring and Autumn period as a whole.
Does that mean there is nothing of interest in the data? Not necessarily. The correlation coefficient throws away a lot of data, and the interest may lie in some of the thrown away data. Let us now revert to the primitive method of making a bar graph of War and Diplomacy for each of our eleven reigns. In the following graph, the height of the blue bar is proportional to the incidence of diplomacy events, and the height of the red bar is proportional to war events, in each reign:
1 2 3 5 6 7 8 9 10 11 12
And what does this tell us? First, for the middle and late parts of the graph (reigns 5-12), the rate of war is high and more or less constant (with the conspicuous exception of Reign 10), whereas the rate of diplomacy fluctuates more or less independently of the war rate, though there is a certain decline toward the end. Second, in the early part of the graph (Reigns 1-3), the war rate, though variable, is generally lower, and for part of that period, a rise in the diplomacy rate coincides with a fall in the war rate, and vice versa. That is, there may be not one situation here, but two: an early period, where war and diplomacy relate inversely (Reigns 1-3) and a late one, where war is more or less constant at a high level, while diplomacy tends to diminish (Reigns 5-12).
That's at least suggestive. Now, if we take the ultimate step of reading the text, do we learn anything that matches these observations, or the noted exceptions to them?
Reading The Text
It might seem that a suggestion offers itself for the atypical Reign 10 (Jau-gung, 0541-0510), namely, that it was atypical because for the last 7 of its 31 years, Jau-gung was in exile after a failed coup against the Three Great Families of Lu. During his exile, Lu undertook no military verntures, perhaps because by custom only the ruler had the right to call out the army, and though the Three Families had exiled Jau-gung, they had not set up a replacement. There was thus no recognized ruler in the capital. But this accounts for only the last quarter of Jau-gung's long reign, whereas a notably lower incidence of military events applies also to the pre-exile three-quarters. It seems that something is indeed going on here, and that it is something larger than the exile of Jau-gung. There was a general system-wide lull in the fighting, in all the states.
The decline in diplomacy relative to war in this late period can also be explained, if we patiently notice what the diplomatic events, when they occurred, were meant to accomplish. It had been the custom in the middle years for allies to meet prior to a campaign, evidently to agree on details, and separately assemble for the campaign itself. In the late years, this preliminary conference tended to get telescoped into the assembly event, so that the number of campaigns is no longer matched by an equal number of meetings. The decline is diplomacy, then, seems to be in part no loss of faith in diplomacy, but rather an improved war planning process.
We have now defined two periods within Spring and Autumn: a middle one of high war level and preliminary diplomacy, and a late one of continued high war level but more efficient staff planning.
What about the first period, reigns 1-3? It was during the first three reigns that Lu pursued a more or less independent policy within its region, seeking to become top state south of the Tai-shan, and intimidating its small neighbors, both Sinitic and non-Sinitic. This entailed many visits to nearby places, and some use of force on those nearby places. The roughly reciprocal relationship between diplomacy and war in this period seems reasonable as a reflection of this two-pronged Lu foreign policy.
With Syi-gung (0659-0627), things changed. The central state Jin became prominent and powerful, and it took charge of coordinating military resistance to nonSinitic southern Chu. The smaller states, including Lu, were more or less compelled to fall in with Jin policy. From 0632 (Syi-gung 28th year), Jin took an openly dominating role, compelling the powerless Jou King to acknowledge his hegemony, and also exerting policy pressure on Lu among other states. It cannot be said that from this time Lu really had a foreign policy of its own. It became a laggard and resentful member, but still a member, of the northern coalition which divided its time between defending against external threats and warring for internal supremacy.
Such are the things we can learn, or more exactly, be prompted to learn, by careful use of the numerical data which are implicit in some ancient text. A procedure that analyzes situations for us, without our having do think about the data ourselves, would be nice, but few situations are amenable to that result. A procedure that tells us where to exert our own analytical skills would also be valuable, and this, it seems, is what we have in this instance. The better to keep our own analytical skills in good shape.
- James Legge. The Chinese Classics. v5 "The Ch'un Ts'ew with the Tso Chuen." It is the former you want; the actual chronicle. Ignore the commentary.
17 Dec 2006 / Contact The Project / Exit to Outline Index Page