Adrian Hester

You say that like it's a bad thing.

Damn, sorry, I allowed myself to get rushed and searched on the wrong guy. The post I should have linked to is this one.

She also supported legalized abortion, which you oppose, and unlike her pronouncements on homosexuality she considered abortion a central issue in Objectivism, going so far as to urge her readers to vote against Reagan in 1976 because he opposed abortion: "I urge you, as emphatically as I can, not to support the candidacy of Ronald Reagan. I urge you not to work for or advocate his nomination, and not to vote for him...This description ["a conservative in the worst sense of that word"] applies in various degrees to most Republican politicians, but most of them preserve some respect for the rights of the individual. Mr. Reagan does not: he opposes the right to abortion." Neither are your views on abortion. You disagree with Ayn Rand on abortion yet consider yourself a defender of her work and beliefs. You should extend the same courtesy to those who disagree with you about homosexuality. [Edited to add link.]

The cost of the time, labor, and thought put into acquiring aand processing the things found in nature to where they are usable by others. Again the equation of things as found in nature and those same things rendered usable by human time, labor, and thought, both indifferently described as "resources" to be divided among the group with no regard for the ones who made them usable: In short, again the attempt to obliterate the human mind as our distinctive means of survival.

Liberal contradicts this bizarre claim below: "Cooperative sharing of resources is only the sharing of extra resources among a group by a member who does not require those resources to maintain an average degree of comfort." So in fact cooperative sharing requires that someone else have what is shared. Note the sleight of hand in this phrase: "felt as a 'dependence'". There's no effort on liberal's part to explain why this is actually dependence, or even to define "dependence" at all; instead, liberal assumes somehow that there is no need even to justify his claim that the animal feels a certain way, never mind equating this feeling with reality. Again liberal gleefully ignores productive activity and the use of the human mind and labor, which are precisely what underlie the distinction between the earned and the unearned. Even gathering resources available and ready for use in nature requires the labor of gathering it, and most resources only become ready for use after human processing--labor, time, and the use of the human mind to discover how to do so in the first place. So contrary to liberal's last sentence, the dependency in cooperative sharing is precisely dependence on the labor and minds of those others who gathered or processed the things of nature in the first place, not on the Earth. It's a revealing argument though--by ruling labor and mind out of consideration from the get-go, this allows him to pretend that the cooperative group is itself a part of nature on the same level as the Earth, which liberal seems to consider a desirable state of affairs. Ayn Rand pointed out that leftists seek freedom on the intellectual level, conservatives on the material level, since both sides want to control the realm it considers metaphysically important. Our poster liberal exemplifies this perfectly.

I'm not a professional historian, I'm a linguist. You don't need historical training to treat such issues, just determination, curiosity, and time. So starting with the Qin you start getting positive trends even with your method: Hardly ringing evidence for a universal trend towards decay! It's a question of how reliable you yourself to know your data to be. Otherwise you're just picking cherries some boss has laid out a path for you to pick along. But the question is how to prove this to someone like me skeptical of your claim about entropy. (And technically, I fitted a power law to your data, though by the quickest method I could, so the details doubtless differ.) The very first step of using cumulative data introduces a bias towards the earlier data that I pointed out renders your method nugatory since it gaves negative trends for reversed time series as well in certain circumstances. If used for comparison of alternate methods, no, they're useful.

I wrote a lengthier reply earlier that I tried to post just as the power went out, so it's lost. While not as fleshed out with examples, at least I hope this version gains in clarity from concision. No, institutions of headship or chieftainship involving lineages over several generations are well-nigh universal whenever we look at non-literate societies. The Samoans, for example, had well-developed kingships and aristocratic societies with well-defied lineages despite not being able to write, as did the Hawaiians, the Bedouins of Arabia, the Salishan peoples of the Pacific Northwest, and on and on and on. It is clear that such societies kept track of their history through oral transmission, and while the details were unclear and fluid, the very fact that their rules had lineages was apparently universal. And when the great powers have had dealings with any part of the world peopled by other societies, we know from their records that they encountered states with ruling lineages, if only in the fact that those rulers paid them tribute in submission recorded in their monuments; these states could be quite large and powerful, as with the various steppe empires neighboring China. And remember: The Iliad and Odyssey themselves were preserved for centuries by oral transmission; oral transmission is not as reliable as writing, but it’s hardly just to conclude, as you do, that without writing these people had no sense of the past or of their neighbors and thus their states cannot be counted as dynasties. In any case, we again run into a basic methodological problem with your entire approach: What do you mean by “dynasty” anyway? We know these dynasties you do recognize were surrounded by states with lineages of rulers; we can also be fairly sure that these latter were shorter-lived than your dynasties on average. You exclude them for reasons you haven’t fleshed out, but until you support such a distinction, all you have done is sidestepped the fact that your methodology quite clearly introduces systematic biases towards longer average durations in the earliest periods (before 500 BC), biases that moreover are very great because so few dynasties of medium or short length would be expected to be recorded. First of all, you’ve made a basic assumption, bolded above, that needs a great deal of argument: That what gives a society stability is one and the same thing for all societies at all times and places with any social structure, political institutions, technical achievements, or economic systems, as if exactly the same factors make a skyscraper and a tent stable in the same degree. However, feudal monarchies and bureaucratic monarchies are very different, as are sedentary and nomadic societies, agricultural and herding societies, and industrial and pre-industrial societies, and it is a dubious enterprise to lump them all together in periods when one or another form of social, political, or economic organization is spreading at the expense of others. So instead of throwing out data from fundamentally dissimilar cases you practice the equally egregious error in the other direction of lumping them all in together. Ah, but we know that other states did exist, and some were quite large and powerful without leaving records of their own. We also know of others by name that did have contemporary records, but these have been lost--the Greek kingdom of Bactria, for example, from which only coins, archeological remains, and brief, ever so brief mentions by later Greek historians survive. Similarly, the Toba Wei (who established the Jin states in north China) might have possessed writing, but no trace of it has survived. And there are many other examples, such as the Hephthalites of Central Asia, the southern Arabian kingdoms before Islam (much of what we know about them is due to their importance to the struggle between Rome and Persia). So what? That entirely misses the point. Spain and France competed bitterly in that period against each other and against Britain, Austria, and their neighbors, and in so doing they gobbled up many shorter-lived neighboring states and put ends to their dynasties. You lump this process of consolidation in with the co-existence with little mutual influence of Spain and China on opposite ends of the same landmass, which is dump oranges in the apple bin and sell them all as Granny Smiths. Ah, but "merely being able to come up with an alternative hypothesis" is an uncharitable description of what I've done. I'm pointing out alternative factors we know existed and pretty clearly were in play but that you ignore, and whose exclusion introduces systematic extensive bias towards your conclusion. It doesn't deal your case a fatal blow, but it does constitute a major body blow against the reliability and completeness of your data. You have to ask, good enough data for what? For the dynasties for which they survive, yes, but as representative of the experience of the rest of human societies at the time, hardly. They introduce systematic biases that you rely on as model cases of the processes you claim to detect. Which entails something akin to unconscious cherry-picking. If a historical schema is over-inclusive, then to support it in argument you have to disregard lots of contradictory data, which Toynbee and Spengler most certainly did--and what would their criteria have been? That the contradictory data was unimpressive. Unimpressive why? Because it didn’t fit the program. You can call it something besides cherry-picking if you like; in any case, it’s the sort of error that historians are trained from their earliest years of study to recognize and avoid, and thus strikes historian critics of Toynbee and Spengler as basic errors of the crudest sort. “John has more moral failings than Bill” is hardly a shining encomium, and “John has more moral failings than Bill, so Bill deserves respect” is fallacious. In any case, I think Ayn Rand’s emphasis on basic philosophical positions much more convincing than Toynbee’s Christian apologetics or Spengler’s etatist pessimism.

First of all, you include on the one hand the Liao and the Jin dynasties and on the other the two Song dynasties, even though they coexisted. (The Liao was a state ruling the north China plain and vast stretches of the steppes founded by the Khitans, an early Mongolic people; the Jin was founded by the ancestors of the Manchus after they revolted against the Liao. The two Song dynasties were their native rivals in the south.) Second, there's no cause to include the Shun or the supposed "empire of China," and in any case the period given for the latter is dubious. The first was founded after the death of the last Ming emperor by a rebel peasant leader who captured Beijing (amid other powerful figures in the rest of China) in April 1644; his forces were defeated a month later by the Manchus, and his supposed dynasty was just a claim to the throne and a name. The duration of a year given in that list refers to his lifespan after April 1644, not to any period of effective rule. The second was founded by the first president of the Republic of China, Yuan Shikai, who served an elected term of 4 years, 1912-1916. In December 1915 he decided to reestablish the monarchy and declared himself emperor; the empire was abolished three months later and far from ruling any part of China effectively, it greatly sped the collapse of the republic into a congeries of domains of competing warlords (a period that lasted until Chiang Kai-Shek's Northern Campaign of 1926). So that figure you have of 4 years confounds his republican term of office and his three months of failed empire-building. In both cases they are historical curiosities that for whatever reason (they do have some slight historical importance, but not of the same degree as the Tang and Song) Wikipedia included out of a whole constellation of earlier futile and abortive bids for power that they ignored. Third, you include several periods of disorder and disunity as single units: The Three Kingdoms period after the end of the Later Han (the terms Former Han and Later Han are probably more mainstream than the geographical designations, but both are common enough currency, both in Chinese and Western scholarship), the Jin dynasty (which ruled only the northern part of China; the rest of China was divided among the so-called "16 Kingdoms," a period lasting from 304-439, and thus overlapping with the following period), the period of the Southern and Northern dynasties, when again China was divided into more than one state, and the period of the Five Dynasties and Ten Kingdoms after the collapse of the Tang. By lumping these all together, you follow a native tradition common in Confucian historiography (which means pretty much any Chinese history written in the imperial period, but identifies the intellectual source of the tendency) to disregard from consideration all dynasties that were not decided on whatever grounds (usually sheer success) to have received the Mandate of Heaven. Those periods were ones in which no agreement was reached on whether the Mandate had been passed down at all or simply withheld by Heaven, and thus much complexity is written off by using a simple canonical list. (Though even here the list is not canonical: The Shun certainly was not canonical, and by no means was Yuan Shikai's empire.) On the other hand, when the Mongols established the Yuan dynasty, their political program, such as it was, put them forward as successors to both the Chinese native dynasties and the empires of the steppes, and as a result the Confucians who wrote history for the Yuan wrote dynastic histories of both Song dynasties as well as the Liao and the Jin; had the Mongols not conquered the south of China, it is doubtful the successor of the Southern Song would have commissioned histories of the Liao and Jin. More generally, the choice of whether a dynasty was canonical or not was determined partly by the political circumstances of its successor, for the successor would in imperial Chinese political tradition have the obligation to compile a history (actually more of an encyclopedia of governance and administrators with lots of added almanacs) of its predecessor; after a period of disunion, the state that unified China (or at least part of China) could pick and choose favorites in the past with some freedom. Similarly, your figures are based on the view of the Former Han dynasty starting in 206 BC, but in fact the Qin dynasty was overthrown by a number of local rulers, and the one who declared himself supreme, Xiang Yu, ruled from 206-202 BC, before he was finally killed by the ruler of the Former Han, his former vassal. Since this was before the full-fledged intellectual dominance of Confucianism (which pretty much established its dominance in governance and political thought with the eclectic syncretism of Dong Zhongshu around the middle of the Former Han), it was the historical writings and schema of Sima Qian that set the date of 206 BC as the founding of the Former Han and the end of the Qin. (His biography of Xiang Yu is well worth reading, by the way, well-written, well-drawn, and wonderfully dramatic.) Fourth, while you include the Xin dynasty that overthrew the Former Han, your list leaves out many other cases of effective usurpation that simply weren't treated as canonical by later scholars. For example, Wu Zetian (Empress Wu) established what she called the Zhou dynasty (same character as the Zhou dynasty beloved of the Confucians) from 690 to 705; it ruled China effectively and was not ended by strife--Empress Wu abdicated due to illness and handed the govrnment over to a son, who reestablished the Tang dynasty. In essentials her dynasty was almost the same as the Xin dynasty, except that the founder of that dynasty (a son of a Former Han empress) was killed by an army of peasant rebels and the Han reestablished by the descendants of the emperors of the Former Han; whether you view the Hans and the Tang in either case as one long dynasty or two small dynasties separated by a short usurping dynasty is a matter of debate and in the list you use is something that was decided purely and simply by the political needs of the victor over the usurper. Fifth, the actual dates of each dynasty used by Wikipedia are somewhat arbitrary. Their article on Chinese dynasties includes one example: The Qing unified much of Manchuria by 1599 and took all the measures necessary to make a successful bid for traditional dynastic power by 1616; they established the Qing dynasty as such in 1636 when they gained rule over the Inner Mongolian tribes and received the imperial seal of Chinggis Khan; they captured Beijing in 1644; they defeated the last Ming pretender in 1662; but they also had to fend off a massive revolt in the period 1674-81 (the Revolt of the Three Feudatories). Which date do you choose? Similarly, the Mongols defeated the Jin and declared war on the Southern Song in 1234; they effectively (well, effectively by their lights, but they had quite different views of governance than the Chinese, so a Chinese view would be "effectively and cruelly") ruled much of north China from 1215, but it was only because Khubilai Khan had been trained by Confucian scholars as a young man that he saw the need to declare himself emperor in the Chinese tradition in 1271, and only in 1279 defeated the Southern Song and took control of all of historical China. So why 1271 in his case (the year he declared the dynasty) but 1644 for the Qing (the year they captured the rival capital)? Sixth, this list of dynasties covers a long period in which there was massive systematic change in the very character of Chinese society and political thought, as well as extensive geographical expansion; it's not geography or social structure or political structure but ethnic continuity that determines what you have chosen to compare as like to like. (For example, historically China was situated on the North China Plain around the Yellow River, and the Chinese still revere this as the cradle of Chinese civilization; when that area was ruled by non-Chinese dynasties after about 300 AD, Chinese literati in the south mourned themselves as exiles in their own homeland--they lived under native dynasties but what they saw as their true homeland was under alien rule. A similar situation held in the Southern Song.) Similarly with the social structure: The Shang and Zhou were feudal states; the emperor enfeoffed each local ruler in a ceremony in which (at least under the Zhou) he handed his vassal a handful of earth from the territory in which he established him. As a result, the emperor had an important political role as the figurehead who legitimized the power of the local ruler, regardless of his degree of political power in the imperial realm--this is why the end of real imperial power by the end of the Spring and Autumn period didn't lead to the end of the dynasty. Had it been expedient to do so, the Zhou would have been deposed. In terms of real political power rather than polite fictions, the Warring States period was a period of competition and disorder like the Three Kingdoms and other such periods, and note that while the Zhou dynasty was recognized, actual political power was vested in the title of Hegemon (ba in Chinese), which was the title that all rulers until towards the end of the Warring States period strove for--it's quite analogous to the Japanese case, in which effective imperial power ended in 1092 with the establishment of the Kamakura Shogunate, but because of Japanese political ideals the emperors could not be deposed even if they were purely ornamental throughout much of Japanese history. The Qin on the other hand represented the culmination of the bureaucratizing trends in all of the individual states of the Warring States period, during which the aristocracy exterminated itself en masse as a group or class and power and administration turned increasingly from the personal to the bureaucratic (whenever a state fell, its victor enslaved much of the leading nobility and made the rest commoners; at the same time, Confucianism was spreading as the ideology of the commoners who staffed the administration, and their political views were for a fully bureaucratic, rigidly moral administration in the service of a legitimate ruler); however, the mixed feudal system of the later Warring States period did not die out immediately; significantly, Xiang Yu took the title of "Hegemon-King" (bawang) and the other rulers as kings (wang), such as the founder of the Han dynasty as King of Han (a state around the Han River, hence the name of his dynasty), precisely the political system of the Warring States and thus a repudiation of the Qin system; and while the Han kept the less objectionable features of Qin bureaucratism, they still found it necessary to enfeoff a number of powerful local figures as princes who revolted throughout the Former Han dynasty. (It was that process that finally killed off the remnants of the Warring States political tendencies.) Of course, bureaucrtization continued to develop throughout Chinese history (selection of officials by examination rather than personal recommendation by a patron and birth into an aristocratic family was not established for much of the administration until the Tang dynasty, and selection by examination as the sole source of officials until some time in the Northern Song, I believe, and it was only after that time that an ideology of personal advancement purely by merit and scholarship regardless of the circumstances of birth spread throughout society and showed up strongly in literature), but there was a sea change between the Zhou and the Tang that makes treating them as part of one long history but the dynasties of France or Italy as distinct from the Roman Empire essentially arbitrary. There is a great difference between a feudal monarchy and an absolute monarchy heading a bureaucratic administration--at least as great as the difference between the latter and a modern republic. To focus on monarchy to the exclusion of all other political factors is a drastic oversimplification that by its very nature lumps together many dissimilar regimes that should be treated separately. Finally, very little is known about the actual history of the Shang dynasty; the Confucians revered the Shang and in lieu of actual detailed information about it treated it with a simple schema of dynastic establishment and decline. It is unclear that it was actually one unified dynasty with effective rule throughout that period, nor do we know the details of whatever political fictions might have been erected to give a gloss to the machinations of political actors behind the scenes or to give legitimacy to fundamentally new institutions, actors, or families. If we restricted ourselves to the same level of detail for the rest of Chinese history, the Eastern and Western Zhou would have to be treated as one single dynasty, as would the Former and Later Han and the Northern and Southern Song; the Jin and Liao would have to be excluded entirely, and the Three Kingdoms period and the following period of disorder probably lumped in with the Han and the Tang, respectively; the Sui and Qin would probably disappear (they were usurpers whose innovations, which were pretty extensive, were picked and chosen among by their successors), and the Yuan would as well, most likely. So no, I don't think it valid to include the Shang; it's simply incomparable in the amount of actual hard facts to the rest of Chinese history. (And I've already stated my objections to including the first two that you ilist above--tell me, do you include Adam and Methuselah in your king lists somewhaere? If not, then you should exclude those two periods as well without the slightest qualms.)

As with the mythical state of Utopia. If we reverse the data, then for the raw data we find an exponent of -0.0084 and for the cumulative mean an exponent of -0.0081. So no, your method does not show what you want it to show; it's simply unsound.

Even though I know fairly little about Portugal, I still know enough to question the wisdom of treating the period of Spanish rule between 1580 and 1640 on the same level as native dynasties. "Cumulative mean"? So in other words you're basing vast claims about human behavior on the general phenomenon of regression towards the mean! Let's do this differently and take your same data with duration versus dynasty number. In that case you have a decline that is much less precipitate; if you fit a power law to it you get an exponent of -0.03149, with a correlation coefficient of 0.34. (Note that the number of digits I've kept is far greater than appropriate, but then you do the same.) It would be an interesting mathematical exercise to start with time series fitted to a power law with a low correlation coefficient and a low absolute power exponent, represent the expected cumulative mean as a time series, fit that to a power series, and find the expected power exponent for that; I suspect due to regression towards the mean you'd find the expected exponent to be negative and rather larger in absolute value for a wide range of initial parameters. So yes, another reason your methodology is entirely suspect--ironically, through the simple operation of randomness and the law of large numbers! And in fact we can test this very easily with the Chinese dynasty data. Start with the data set that includes the Shang dynasty with a duration of 510 years (an oops above--that, not 540, was the duration I used throughout), breaking up the Eastern Zhou into the Spring and Autumn and Warring States periods (which is actually beneficial to your argument, since that tips the exponent from slightly positive to just very very barely negative), and including the Liao and Jin (non-Chinese dynasties ruling nortern China) rather than the two Song dynasties (native dynasties ruling south China in the same period), though the differences resulting from that last choice are entirely negligible. This set of choices, which still has many points to quibble about, gives an exponent of -0.0006763, based on my figures, with a correlation coefficient of 0.009. If you then calculate the cumulative means and fit a power law to that, you (or a least I) find an exponent of -0.022 with a correlation coefficient of 0.88. But you'll find that with a wide range of data sets; using the cumulative mean wipes out vast amounts of variation and greatly exaggerates the effects of earlier dynasties on the results. This we can see if we do a simple thought experiment: The geographical region of Utopia has been ruled by 20 dynasties; the first one lasted 500 years, the second 200 years, the third 30 years, and each of the subsequent dynasties lasted 10 years longer than the predecessor of its predecessor. (This gives a sharp initial drop followed by a slightly rising heavily jagging rise similar in very broad outline to the Chinese dynastic data.) The general trend is rising, and if fit to a power law has an exponent of 0.0084; however, if you fit the cumulative means to a power law regression towards the mean predominates and you have an exponent of -0.014. (Moreover, the correlation coefficients are 0.145 and 0.696, respectively, for the cumulative means wash out much of the variation in the actual figures, so that the power law fits the cumulative-mean data very much better.) What is fatal to the trustiness of your method is that if we reverse the data we sometimes get much the same result! For the same Chinese data reversed, the cumulative mean when fit to a power law has, yes, a negative exponent, -0.012. So whether going forwards or backwards in time, you get a negative exponent either way! For the Portuguese data reversed, on the other hand, the shortness of the last two dynasties suffices to give positive exponents, 0.031 for the raw data and 0.0168 for the cumulative mean. Well yeah, if you fit a time series to a power law, in lieu of other information you can expect the power law to be declining in half the cases. If you then add in regress towards the mean by relying on cumulative means, especially with small absolute values of the exponent, of course you're going to have a predominance of power law declines! In any case, your French data are a mass of incompatible regimes--not dynasties since you're lumping in republics with empires and monarchies. Here is the second half of your data with the names of the regimes in question labeled and classified: 12 / 137.3 First Republic (usually divided into different regimes as follows: Constitutional monarchy, republic, the Directorate, and the Dictatorship of Napoleon) 10 / 121.4 First Empire (Bonapartist) 1 / 108 Bourbon Restoration following Napoleon's abdication (pre-revolutionary monarchical line) 0.5 / 97.2 Return of Napoleon from exile at Elba 15 / 89.7 Second Bourbon Restoration (pre-revolutionary line) 18 / 83.8 July Monarchy (Orleanist, an offshoot of the Bourbon line) 4 / 77.6 Second Republic 18 / 73.4 Second Empire (Bonapartist) Not only do you lump in republics with dynastic regimes, you put the period from Napoleon's abortive return from Elba until his defeat at Waterloo on the same level as the Carolingians! Suffice it to say that's a sign of utter historical ignorance. Before the 1750s there was very little antimonarchical sentiment in Europe, but once it spread, then yes, there was a great deal of political instability in Europe, and particularly in France. This passel of regimes is a major reason for your results, but it's signally unclear it's fair to include them. If we exclude the First Republic and all later regimes, we find that the decline is much less marked (and in this case is obscured by taking the cumulative mean, thanks to the sagging middle): -0.064 for the raw data, -0.016 for the cumulative means (correlation coefficients of 0.26 and 0.31, respectively), versus by my figures -0.13 for the overall raw figures, -0.036 for the overall cumulative means (correlation coefficients of 0.65 and 0.93, resp.). For the reversed data excluding post-Revolutionary regimes, we find +0.064 for the raw data (as we should) and +0.014 for the cumulative means. It would be interesting to take all of your data, reverse them in time, and fit the resulting cumulative means to a power law and see how many of them remain negative--given the low correlation coefficients for the raw data you've already presented, it's a toss-up, but that's because your method is fundamentally unsound for showing what you want to show.

First of all, note that before about 500 BC, very few dynasties are recorded--the historical record is incredibly sparse for that period, and in many cases the king lists that you are relying on probably include many legendary and semi-legendary figures who were given reign periods that can't be checked and were probably fabricated out of whole cloth for cultural reasons. (Like the sage emperors in China.) Besides that, the only dynasties from that time we'd have records of are the ones so powerful and so long-lived as to make such an impact on later societies, for very few societies had writing then! The average durations you rely on for the period before 500 BC, in other words, are simply not at all comparable in completeness, accuracy, or reliability with later periods. Moreover, the period after 1800 saw republican revolutions throughout Europe and Latin America; that too renders their inclusion suspect since an entirely new ideological factor had come on the scene. So, if we focus only on the years between 500 BC and 1800 AD, we find a general downward trend by number of dynasties: The average duration of all dynasties in years is 273.5-1.549*N (N=number of dynasties), with a correlation coefficient of 0.73 and error about the mean of 49.6. But even on your own terms, that makes sense: Thy're not all independent at all, but rather competing within larger geographic areas, so of course some win out at the expense of others, a process that is more intense the greater the number of dynasties. But you can't blithely disregard the serious problems of reporting either--the historical record is very spotty, so you have a general trend throughout the period of 1000-1800 AD of consolidation against a much greater mass of documentation from 1500 on; a similar pair of competing factors is at play for the period before the collapse of Rome. What I see is that you have no idea of the factors at play on a true "detail level." The data simply are not comparable throughout the period; taking the cumulative average is a meaningless exercise without such detailed treatment. I see you actually do have a glimmering of the difficulties you've not addressed, but the fact that you treat them as open questions is significant. Ah, but I'm not "postulating a few missing ones," I'm pointing out systematic gross errors with strong time dependencies that knock your entire methodology into a cocked hat. A result based in great part by including semi-historical and pre-historical (i.e., legendary) dynasties and ignoring basic questions of the reliabilityand completeness of the sources of your data. No, we don't even know THAT this rule operates at all. Makes me think of Toynbee--a vast historical schema to which he fit cherry-picked data with no regard for the historiographical difficulties at the base of his entire enterprise.

More generally, if succession is from the father to the son, the if the father is long-lived the son will be fairly mature upon accession to the throne and simply in the course of things will not be expected to have as long a lifespan. Wrong? Or do you mean incomplete? Those are very different claims.

Post your data for China--the names of the dynasties and their durations upon which you based your result. I'm fairly familiar with Chinese history and unlike you don't have to rely on the vagaries of Wikipedia for what constitutes a "dynasty." I'll add up front that I ran the same test and found results at great variance with yours--positive power-law exponents in all cases, regardless of how you define "dynasty" (for it's a vague term that meant different things at different times), provided you use only the dynasties for which the dates are actually solidly supported in the historical or archeological records. (Which excludes the Xia and Shang dynasties; it's ridiculous on the face of it if you know anything about Chinese history and culture to even consider including the Three Sovereigns and Five Emperors, who are purely legendary figures of the same sort as Adam and Methuselah and with much the same overinflated lifespans.) The existence of the Shang is at least established, though the dates are sketchy; if we include a duration of 540 years (the various figures are around that), then and only then do I get negative exponents--very small negative exponents in some cases, very small positive exponents in others, with correlation coefficients between 0.02 and 0.06. (The Xia is charitably described in many sources as "possibly legendary.") In other words, the data are such that you have to include one semi-historical dynasty about which virtually nothing of its history or its reign periods are known in order to support your thesis, and even then it's statistically nugatory. More generally, in fact, a discussion of the Chinese data would clarify many of the problems with your blithely relying on others to do your basic ground-clearing. Two initial questions: 1. Do you fit your power-law curve to the number versus duration of each dynasty, or to the beginning year versus duration of each dynasty? (For me this doesn't matter; I get a positive power in all cases without the Shang and positive or negative but with very small absolute values with the Shang included. It's an important question though to pin down your methodology.) 2. Why haven't you reported your correlation coefficients? In all cases mine were less than 0.18, which is hardly impressive evidence. Once I see your data, I'm sure I'll have a large number of questions about your principles of selection--though you've already admitted you most likely fobbed off the responsibility for answering them onto Wikipedia.

Chinggis Khan died in 1227, before war with the Southern Song dynasty even started (in 1234); moreover, there were only two Chinese states (in the sense of states based in historically Chinese lands with predominantly Chinese populations), and in fact under the Southern Song Chinese culture, science, technology, and printing flourished, and indeed the Chinese economy not only expanded extensively (population growth without a decrease in per capita production) but intensively (an increase in per capita production). In fact, the Song saw one of the few pre-modern industrial revolutions (fueled by advances in metallurgy, paper production, and so on, with literati reinvesting their earnings in trade and industry rather than rent-seeking). See Eric Jones' Growth Recurring: Economic Change in World History. More than that, the Song did not neglect their military; they established the first standing Chinese navy in the 1130s, for example, and by the end of the dynasty (1279) had a million men in arms. The problem is that the other "Chinese" state in northern China, the Jin (founded by the Jurchens, ancestors of the Manchus who founded the Qing dynasty in 1644), fell to the Mongols in 1234, after which the Mongols were able to use their armies and military technology. Even so, it still took 45 years after war started (when the Jin finally fell, the Southern Song seized some of the historic Chinese capitals in the North China Plain against their agreement with the Mongols, which was taken as an act of war) for the Mongols to conquer the Southern Song; the Southern Song was hardly a rotten shell just waiting for a sharp kick to knock it over. And it was a shame the Southern Song did fall, because it was their fall that doomed the Song policies of reduced government regulation, military prowess, and their vibrant cultural accompaniments in the eyes of later Neo-Confucian scholars and rulers.

A positive action of self-defense, in which case the consequent death of the fetus is justified. What you are saying then is that unlike every other sphere of social life, a pregnant woman does not have the fundamental right to deal with others or not as she wishes. Why? Precisely because in this case the fetus cannot support itself. You are saying that it is precisely because a fetus is not rational, not independent, not able to live on its own--and that its needs somehow give it rights that no other human being has. But again, if the fetus does not have the right to live at its mother's expense, then it doesn't matter how healthy it may be, it can be aborted--killed as an act of self-defense. Rather, parents are responsible for their children because they consented to support them by not giving them up for adoption--or if they are adoptive parents, by the fact that they adopted them. In this respect it doesn't matter whether they created them or not. An arbitrarily asserted responsibility, yet again, begging the question. Any sufficiently crippled human being at any age cannot support himself--and by your own argument, he still possesses a right to existence just as a fetus does (the fetus's right to existence, you say, is what remains when you pare away all the other rights recognized for different stages in life, therefore their right to existence is the same). But he does not have the right to be supported at another's expense against that person's consent. The fetus has that same right of existence--therefore by your own admission it has no more right to support than a crippled genius does. Saying that the cases are different because of the mother's responsibility to her fetus for her actions smuggles in the conclusion you claim to prove: She only has such a responsibility to it if it has a right to such support, and the only basis you have for claiming it has such a right is this arbitrarily asserted "responsibility" a mother has to her fetus, based unlike any other human rights on that fetus's needs. "Requirements." In other words, its need. Need gives no one a right to anything. The need for a fetus to be carried to term in order eventually to grow into a fully rational human being does not give it the right to the support necessary for it to do so--this right can be assumed legally if consent to support the child has been given, first by giving birth, second by keeping the child or adopting it as the case may be. (This is why birth is the most convenient dividing line for when an immature human being is given legal protection.) But the very fact that a woman aborts a fetus shows that she does not consent to giving it birth. "Metaphysical status"? At best a fetus is a human being like any other, and as such has no more right to support because of its needs than any other human being does, and if a woman does not consent to support it, it's self-defense to remove it from her support, even though it must kill the fetus to do so. In fact, as far as "metaphysical status" goes, you've erected the fetus into a super-human entity whose simple fact of existence gives it a right to support that no other human beings possess, by the simple fact of its needs, precisely because it is not rational or independent (and, I might add, a right that it loses at birth). Consenting to sex does not imply consent to any of its unintended consequences. Yes, you are responsible within certain standards of reasonable expectations for the unintended consequences of your actions, but that responsibility is to other persons who have legal standing--the fact that a fetus results unintentionally from a consensual act does not by that fact make it into a person with legal standing to whom one owes responsibility. No, you're concerned with the needs of fetuses (which you have insinuated through their "context"), which you see as negating the rights of women.