David Henige
To kill an error is as good a service as, and sometimes even better than, the establishing a new truth or fact. (Darwin 1903, 2:422)
Recently a devotee of J.R.R. Tolkien’s Middle-Earth cycle bethought himself to estimate the size of the elf population there (Losack 1987). Unfortunately, it appears that no formal count of these creatures survives, forcing the author to extrapolate from certain figures cited in his sources, particularly the reported size of various elf military detachments. In addition he calculated average generations and “typical” family size, adding and deducting various elements of elf society arbitrarily. In the end he determined that at its height the elf population reached one million and then began to decline precipitately.
Sound far-fetched? Not to anyone familiar with recent studies of the contact population of the New World. In this essay I will show that much of this work has chosen to base itself on sources that are not very different from the Tolkien canon and has adopted methods uncannily similar to those used to estimate the elf population of Middle-Earth.
[The Credulous Man] finds most delight in believing strange things, and the stranger they are, the easier they pass with him; but he never regards those that are plain and feasible, for every man can believe such [Butler 1970:265].
The size of the aboriginal population of the Americas was the object of guessing games almost from the Discovery itself, and enjoyed a considerable vogue in the latter part of the eighteenth century (Gerbi 1973: 50-120; Commager and Giordanetti 1967). Still, no earlier interest was as sustained or as extensive as that shown during the past forty years or so. This modern movement was initiated in the 1940s by the so-called Berkeley School—Lesley B. Simpson, Sherburne F. Cook, and Woodrow W. Borah (with Carl O. Sauer as godfather)—and has continued ever since. This work has been extensive, undoubtedly impresses at first sight, and has been extremely influential both within the field of historical demography and in the larger scholarly world.1
The systematic distortion of the historical record gained momentum with the Berkeley School and it too continues unabated. At first, the conclusions of the High Counters (as I shall term them) were fairly modest and relied largely on the evidence in the early writings on the conquest of Mexico (and later of Peru, Hispaniola, and Central America). But as their estimates began to spiral dizzily upwards from two million in central Mexico to twenty-five million for the same area, other materials were conscripted into the study, particularly tribute records (for want of a better term) from the pre-Conquest period. Today it is routine for the high counters to advance figures for the entire New World that range upwards of one hundred million, with North America’s portion of the whole growing ever greater, and with no end in sight for the spiraling increases.2
The High Counters were able to raise their estimate twelvefold in large part because they abandoned sound historical method, which ordains that all relevant sources be sought out and studied critically, both textually and contextually; that accepting and rejecting sources be done explicitly; that dissenting points of view be recognized and also dealt with explicity; that sources be cited fully and honestly; and that conclusions be framed as cautiously as the evidence requires. Under these rules of evidence, such statements as “in the absence of any contrary evidence” (Cook and Borah 1971-79, 3:152) become inadmissible when no such evidence can possibly exist. As we will have occasion to notice, though, the High Counters have not conspicuously followed any of these rules, perhaps realizing that to do so would severely hamstring their flights of fancy.
The High Counters have written thousands of pages, and here I can deal only with limited aspects of that production. On the assumption that extraordinary claims require extraordinary evidence, I will concentrate on the High Counters’ use of written post-conquest sources and will generally confine my discussion to the works of Simpson, Cook, and Borah, as well as to Henry Dobyns, for this seems to be the direct line of descent of the baton. The presuppositions and techniques behind the High Counters’ use of pre-conquest materials have been criticized briefly by Henige (1978b) and more extensively by Zambardino (1980).
Along the way, the Berkeley School, like most movements, came to resemble an echo chamber—self-congratulating, self-validating, and exclusionary.3 Its members have no doubt that they are on the side of the angels and do not expect their views to be nourished by disagreement. Quite the contrary; dissent is not entertained, and critics necessarily speak only from ignorance and want of skill. After all,
[a] complex and most bewildering series of calculations and adjustments is necessary to make allowance for such matters [how to multiply, how to add]; the calculations and adjustments in turn must be based upon close study of the society, the administrative system, and the circumstances of the making of the count or assessment. It is no wonder that people unacquainted with such techniques or with the methods of historical verification [sic] are tempted to dismiss all such studies as mere legerdemain. (Borah 1976:26)
Borah expatiated at some length in this vein, speaking of “the application in full of very elaborate and exacting techniques of verification” and “painstaking reading” of the sources. Finally, in an outburst of irrepressible modesty, he concluded by confessing that the whole enterprise of the High Counters has been firmly anchored in the lofty standards of textual analysis established by the Bollandists and Maurists (Borah 1976:24-25, 33).
All these statements have an unedifying whiff of divine-right-to-be-right pronouncements by a prophet declaiming arcane knowledge vouchsafed him by a preternatural being, and the last amounts to a piece of preposterous, if unintended, self-parody. For the High Counters, textual criticism consists in seeking out and then justifying belief in, and only in, the very highest numbers in the sources.4 But at least, as I will argue here, Borah in fact well christened the techniques of the High Counters: “mere legerdemain.”
Know for the truth that in the army there were 300 giants, 163,000 infantry all armed with the skins of hobgoblins, 3,400 men at arms, 3,600 double cannons and arquebusiers without number; 94,000 pioneers, 450,000 prostitutes. . . . (Rabelais 1965:137)
The High Counters often turn for comfort to warrior counts—early estimates of Indian armed forces. The reason is simple enough—such estimates very frequently make their appearance in the historical record. Interpreting warrior counts involves two issues: what proportion of the general population do they represent; and how accurate are they? Comment on the first I defer for the moment. About the second a great deal might be said, far more than is possible here. The technique of the High Counters is occasionally to notice briefly criticisms of warrior counts, peremptorily dismiss them, and hasten to assure readers that they find the counts persuasive. High Counters tend to find them so reasonable in fact that they have no compunction in accepting any estimate of Indian armies as reasonably accurate (Henige 1986a:703-8). But they prefer the larger estimates. In the process the High Counters repeatedly grant eyewitness estimates privileged status, though on what grounds can hardly be imagined. Virtually all the several hundred examples of inflated warrior counts from other times and places of which I am aware entered the historical record by way of participants, and there are no reasons to accord eyewitness testimony any more credibility than we might give any other military estimates (e.g., Hollingsworth 1969:227-32).
The Spanish began by having trouble estimating even the small numbers of Guanches in the Canaries—estimates of the population of Gran Canaria ranged from 6,000 to 60,000 (Fernández-Armesto 1982:10-11). For Hispaniola the difficulties only increased (Henige 1978a). In turn, with regard to Mexico, the problems were fewer only than the estimates themselves. Unlike most other students of the subject, the High Counters lend eager ears to these numbers, which are usually in the hundreds of thousands. Those who do not actually regard them as underestimates (e.g., Dobyns 1966:407-8) claim to find them “a faithful and exact portrayal” and “eminently reasonable.” (Cook and Simpson 1948:23; Cook and Borah, 1971-1979, 1:9). In doing so they never wonder how the estimates were made, either physically or psychologically, which is remarkable in light of both the wide range of discrepancy and particularly in view of the fact that the Spanish chroniclers believed they were recounting the springing to life of the marvelous deeds in the chivalric romances of the time (Leonard 1949; Weckmann 1984:117-223; Fernández-Armesto (1982:208-211).
A single example must do here to suggest the dimensions of the problem, but it is an extraordinarily apt one. In one of his numerous estimates of the formidable forces that he must (and did) overcome, Cortés asserted that the enemy numbered “more than 149,000” men. This curious statement forces us to believe that the number was also under 150,000; in other words, Cortés was effectively claiming to be able to estimate within 0.6 percent the size of an army he had never seen before! How was he able to accomplish this miraculous feat? Had the enemy conveniently arrayed itself in formations of exactly 1000, or maybe 149, or even 74.5 each? Or perhaps Cortés merely counted arms and legs and divided by four? Or did the enemy commander carelessly leave behind a roster of men on active duty at that very moment? All possible alternatives? Of course. Any of them attractive? Of course not. Have the High Counters suggested any other possibilities? Not at all. Whence this peculiar number then? Like all the numbers in his dispatches, Cortés offered it on a take-it-or-leave-it basis. But then the High Counters require only the number, not its explanation.
Such numbers of course are no more than metaphors for “a lot,” but then “a lot” is not multipliable and affords no basis for extrapolative flights of fancy into the millions (Blázquez Garbajosa 1985). Cortés’ estimates were nothing more than one brief tableau in a tapestry reaching back to the time of Ramses II of Egypt and forward to the Iran-Iraq conflict or United States Football League crowd estimates.
To return to the first question posed above—the relationship of army size to population size—we need to know more about the circumstances than we do in this case. Elsewhere Cook (1973:13) had the answer for the New England Indians. Their armies were about one-quarter of the entire population—certainly a reasonable enough figure under most circumstances. But here, it seems, we have the stirrings of a problem. The estimates of army sizes in Mexico, however inflated, are still far too low to serve the High Counters’ cause very well. Even if we accept every one of these at face value and we further accept that no Indians were counted twice (or more often), the numbers of Indian soldiers would be less than one twenty-fifth of the population eventually posited for the area by the High Counters.
This embarrassing state of affairs arises because, while the High Counters were increasing their estimates for central Mexico from some two million in 1946 (Cook 1946:98) to over twenty-five million scarcely fifteen years later (Borah and Cook 1963), the warrior counts in the record stubbornly remained the same. To what extent this has given the High Counters pause is anyone’s guess since they never discuss their sources critically. But, while they have never publicly disowned reliance on warrior counts, it is not without interest that Cook and Borah (1966:231) eventually decided that maybe, after all, “such materials must be considered of low order of credibility” or were even . . . “unreliable” (Borah 1976:26; cf. Dobyns 1966:407-408).
Our own judgment is, that in his eagerness to make discoveries, our author has done what nearly everybody is sure to do under similar circumstances. He has opened both his eyes and his mouth a little too wide, and swallowed a great deal more than he has sufficiently tasted. (Anonymous 1969:485)
The High Counters (Cook and Simpson 1948:18-22; Cook and Borah 1971-1979, 1:12-13) also rely heavily on baptismal figures to give their estimates respectability—they are the only sources actually to include the magic term “millions.” These figures (not records, I emphasize) are embedded in the histories of the Christianization of Mexico written by Toribio de Benavente (better known as Motolinía), Jeronimo de Men-dieta, and Juan de Torquemada. As always, the High Counters try hard (e.g., Cook and Borah 1971-1979, 1:12) to convince us that each of these sources constitutes independent corroboration of the others. They do not. Mendieta relied heavily on Motolinía, and Torquemada on both Motolinía and Mendieta (Alcina Franch 1973:267-270; León-Portilla 1979:309-12).
Motolinía began the process by offering several conflicting figures. At one point (Motolinía 1941:118) he estimated that “more than four million” Indians had been baptized between 1521 and 1536. Then, only a few passages later (1941:120-21), he looked again at his information and broke it down “in two ways.” First he estimated (1941:120-21) that “about five million” Indians had been baptized “by [Franciscan] friars” and named several, together with the numbers (all well rounded) each allegedly baptized. He then looked at the distribution “by towns and provinces.” Dividing New Spain into four regions, he found that, oddly enough, in each of them “more than one million” Indians had been baptized. For good measure, he added that, since the figures for this regional breakdown had been gathered, another half million baptisms had taken place, bringing the two calculations into closer alignment. Then—and this is distinctly odd, to be sure—Motolinía seemed to forget that his last two sets of figures actually duplicated each other, and promptly concluded (1941:121) that “more than nine million” Indians had been baptized.
To explain this the High Counters assure us that Motolinía “thought the [regional] figures too conservative” and so promptly doubled them (Cook and Simpson 1948:20-21; Cook and Borah 1971-1979, 1:12). Again, this is not the case. Motolinía prefaced his statement of nine million with “therefore” (“por manera que”) indicating that in his mind—erroneously—that figure emanated directly from his earlier calculations and not from any doubts.
When his turn came, Mendieta (1870:275) unlike the High Counters was not taken in by Motolinía’s shaky arithmetic. He noted that Motolinía had stated that “about five million” Indians had been baptized by 1536, and added that by 1540 another million had been added to the total, but provided no details.
Writing much later, and basing himself explicity on Motolinía and Mendieta, Torquemada (1977, 5:237) had some strange things to say. He cited Motolinía’s regional figures almost verbatim, but claimed that they covered the period from 1524 to 1540. In other words, by reporting the same numbers but for a different span of years, Torquemada in effect claimed either that there had been no baptisms between 1521 and 1524 or none between 1536 and 1540! Then he added another touch of his own (Torquemada 1977, 5:237): the Franciscans had baptized “another two million ... in other [but unnamed] provinces and towns.” In this way he “corroborated” Mendieta’s total of six million.
All in all, this comedy of errors is quite an unconvincing modus op-erandi. Clearly enough, despite his claims of “actual counts,” Motolinía had no such thing at his disposal. In his turn Torquemada’s tortuous efforts to bring various figures into harmony led him to offer self-con-tradictory claims. Interestingly, though, however slavishly he followed Motolinía’s procedures, Torquemada was no more inclined than Mendieta to accept Motolinía’s figure of nine million. On the other hand, the High Counters (Cook and Borah 1971-1979, 1:12-13) profess to see “nothing improbable” in even that high figure.
The arithmetical skills exhibited by Motolinía and his followers are not likely to impress skeptics, but Henry Dobyns is no skeptic. As an indication of the vast numbers of Indians available for baptism, Dobyns (1966:405) approvingly cited several cases where Motolinía justified his high figures (even Motolinía seems to have doubted them!) by providing examples of hard working friars baptizing everyone in sight. Motolinía (1941:126) related one case where two priests baptized 14,200 Indians in five days, or one Indian per priest per minute for the entire five days. For Dobyns this “rings true.” Dobyns is even willing (Dobyns 1966:405) to credit that a single priest could have baptized 14,000 Indians in a single day—one baptism every six seconds non-stop! Motolinía himself made no greater claim than that a single hardy priest had baptized 10,000 Indians one busy day, or a more leisurely pace of one baptism every 8.5 seconds. Putting his predecessors to shame, Dobyns concluded (1966:405) in a whirlwind of extrapolative energy that “sixty priests would [sic] have baptized 2,184,000 Indians annually.” The mind bog-gles.
But enough of this. We could ask how many times some Indians were baptized or why the missionaries’ estimates were always so vaguely rounded off to the nearest million or why (or whether) the missionaries believed what they wrote, but there seems little point to that. In the context of the times, of themselves, and of the procedures they describe, the baptismal figures do not abide scrutiny. They are, more than anything else, paeans to the power of the Holy Spirit and, as such, useful indicators only of the mental world in which the missionaries operated.5
But if it is difficult to understand the High Counters’ credulity, it is even harder to explain their constant massaging of the baptism figures. As they stand in the historical record, these figures are, quite like the warrior counts, an embarrassment to the High Counters’ own figures, which actually serve to indict the missionaries for inefficiency. But this is only the last of several puzzling features that beset recent use of baptismal figures by the High Counters.6
It [Oz] has nine thousand, and fifty-four buildings, in which lived fifty-seven thousand three hundred and eighteen people. (Baum 1910:29)
In addition to estimating numbers of living persons the Spanish were fond of estimating numbers of dwellings and even the number of skulls they encountered. Since neither houses nor skulls move about while being counted, we might expect that discrepancies in estimating them would be few and minor. Surprisingly (not surprisingly?) this proves not to be the case. In a passage that mixes misleading statements with an unconscious candor more characteristic of the earlier than the later work of the High Counters, Cook and Simpson (1948:31) addressed the problem with respect to Tenochtotlán:
Firmly embedded in the historical literature is the statement that the city contained 60,000 houses, which would imply a population of 300,000. This estimate first appears in Gómara . . . No explicit statement of the sort, however, is contained in the letters of Cortés, the True History of Bernal Díaz, the narrative of the Anonymous Conqueror, or in Andrés de Tapia.
Cook and Simpson then went on to cite two seventeenth-century estimates, one that Tenochtotlán contained 80,000 houses in 1519, the other 120,000.
What are we to make of this? Surely a statement that appears in no primary source can hardly be described as “firmly” embedded in the historical literature.” Nor are we told just what constitutes “the historical literature” in question, only what does not. Cook and Simpson fail to cite a single source besides Gómara in which the figure 60,000 appears.
Cook and Simpson then proceed to discuss various (and varying) estimates of the population of towns contiguous to Tenochtotlán, reducing or increasing these figures at will in order to produce a figure they deem acceptable, all without a single demonstrably accurate datum. Their effort in this respect highlights one difficulty in interpreting dwelling counts: the identity of that which is being counted. As Cook and Simpson note (but only to further their own calculations), the cities around the lake of Texcoco form a conurbation in which it is extremely difficult to distinguish city ‘boundaries’ when estimating, so we cannot assume that each chronicler was defining “Tenochtotlan” or “Texcoco” or any other place name in consistent fashion. The same applies for other parts of Mexico, where it is often impossible to know whether a reference is to a town or its region. When in doubt, the High Counters have been wont to accept the highest count and then apply it to the smallest possible unit, thereby ensuring the extrapolation of the highest possible number.7
The High Counters have not attempted a comprehensive canvass of dwelling estimates; presumably such a study would only turn up further unwelcome discrepancies.8 What does seem clear is that when estimates agree with each other, one of them is textually dependent on the other, rendering the agreement meaningless.
In turn, in his study of Timucuan population, Dobyns is not averse to squeezing all he can—and then some—some from the less than a handful of estimates of Timucuan town sizes that are available (Henige 1986a:707, 715). In fact Dobyns (1983:209) goes so far as to take a single estimate of a single village to extrapolate the size of the entire Timucuan population. In general the practice is becoming more widespread, even when it involves accepting a single observation or one (or the average) of several discrepant estimates and then conjecturing about the average number of persons per dwelling, average dwellings per town, average towns per polity, etc.
Something must be said about such averaging, a technique which of course is commonplace in historical demography as a means of “reconciling” ranges of figures. To the extent that the figures involved are known to be accurate, averaging has a place—must have a place. But for the New World, averaging merely replaces several unknowns with another unknown and at the same time suppresses the reality of discrepant estimates. The notion that averages are really a species of Golden Mean (e.g., Cook and Simpson 1948:30-34) is nonsensical. No figure in a range may be correct, or the entire range may be too high or too low. In these circumstances averaging is just an expedient to mask the notionality of the numbers being averaged.
Finally, we must bear in mind that the Spanish habit of estimating dwellings in urban settings was in a long-established tradition, a tradition that also had a long-standing history of being inaccurate. Whether we turn to European estimates of their own cities or their estimates of the cities of alien societies or other peoples’ estimates of European cities, we find consistent patterns of over estimation and hyperbole; the estimates are victims of culture contact.9
Let us estimate the number of cannibals at a minimum, say 4,000. Say that each one ate one person a month; and we arrive at a total of 48,000 persons eaten during one year . . . If to this we add those killed in battle, drowned, starved to death, devoured by wild beasts . . . we arrive at a population of [two million] [Ellenberger 1912:225].
The theme of human sacrifice also figured prominently in the accounts of the chroniclers, whether secular or clerical. How could it not, when it provided the strongest possible rationale for the conquest of Mexico? Oddly, though, there are only two estimates (Cook 1946:88-89) of the numbers of skulls in the pyramids that purportedly dotted the landscape of Mexico in 1519. Bernal Díaz provided the first. Cook (1946:88) quoted him as writing of one town: “there were piles of human skulls so regularly arranged that one could count them, and I estimated them at more than one hundred thousand.” Actually, Bernal (Díaz del Castillo 1982:116) was slightly more cautious than this, saying only that “it appeared to me that there were more than one hundred thousand skulls.” If they were as easily calculable as Bernal said they were, why did he not do so? And was it not thoughtful of the Indians to have accumulated such a nice round number of skulls just in time for the Spanish to count them?
The other, and slightly more notorious, estimate is that Andrés de Tapia (Tapia 1866:583) made of the rack of skulls in the main temple in Tenochtotlán. Tapia arrived at a figure of 136,000 skulls and the High Counters (Cook 1946:88-89; Cook and Borah 1971-1979, 1:11-12; Borah:231-232) have had no trouble at all in accepting this. In doing this they overlook at least one early demonstration (Garcia 1902:xiii) that Tapia’s estimate was impossibly high. Still, the beguiling character of large numbers is nowhere more graphically illustrated, and the Tapia count has also led to no end of silliness by cultural materialists (Harner 1977; Harris 1977:107-110; compare Price 1978 and Sahlins 1978) hoping to show that the Aztecs were so driven by a protein deficiency that they dined on their fellows in exceedingly large helpings.
In criticizing the cultural materialists Ortiz de Montellano (1983) points out (in contradiction to Cook, who could “see no reason for not accepting their figures at face value”) that in the circumstances it is only surprising that no more than two exaggerated accounts have come down to us.10 Cook and Borah (1966:232), though, think of these estimates as “sober and serious.” If the High Counters had familiarized themselves with other reactions of Europeans to real and alleged practices of human sacrifice in the non-European world, perhaps they might have reconsidered.11
When you write again, cite fairly the Records and Authorities you make use of, and endeavour to find out the genuine and just sense of them; and abuse not, nor trouble the World by currupting them with false, partial, factious, and opinionative Glosses [Brady 1684:325].
A trait shared widely among the High Counters is the unscholarly practice of heaping praise on each other while ignoring or dismissing uncongenial sources and unwelcome criticism. Thus the detailed on-site inspections of Alonso de Zorita are stigmatized (Cook and Borah 1960:13) as “partisan” because Zorita observed that the encomenderos tended to overstate the number of Indians under their control, whereas the High Counters much prefer to see these numbers as undercounts.
But perhaps the most interesting example of the ostrich at work is the High Counters’ treatment of Bernal Díaz del Castillo. As the author of one of the most extended accounts of the conquest of Mexico and adjacent areas, Bernal must at all costs be shown to be embraced by those intent on ascribing high populations to the area. No problem. Cook and Simpson (1948:23; cf. Cook and Borah 1971-1979, 1:23) are more than happy to credit Bernal’s account as being “full of firsthand observation which supplements the statements of Cortés to an extraordinary degree” and go on to claim that “the testimony of two such experienced witnesses must be accepted . . . as probably coming fairly close to the truth.” To drive the point all the way home, they go on (Cook and Simpson 1948:23) to assert that “the size of Indian armies . . . as stated throughout [sic] the accounts of Cortés and Bernal Díaz, corresponds with the values given for the preceding decades by native historians.”
This sounds promising; but wait, why do we then find that Bernal’s testimony is used rather sparingly by the High Counters? After being cited for one set of figures (Cook and Simpson 1948:24)—which were only one-third as large as those Cortés cited for the same incident—Bernal virtually disappears from the writings of the High Counters, to reappear (Cook and Borah 1971-1979, 1:11) only as a kind of ancillary witness to the din created by the large crowds in the marketplace of Tenochtotlán. The reason for this erratic treatment becomes apparent as soon as we look at what Bernal really said. We then find that he is not the comfortable bedfellow of Cortés or Gómara that the High Counters imply. Nor would he likely have welcomed their brief if warm embrace.
In fact Bernal was scathingly critical of Gómara’s account of the conquest and so, by inference, of Cortés’ version, and his numbers are always very much smaller.12 More to the point, Bernal was repeatedly critical of one particular aspect of Gómara’s account, his propensity for . . . high numbers. Early in his narrative Bernal (Díaz del Castillo 1982:34) charged that Gómara’s version of things was “very much contrary” to what really happened, especially when he wrote of “great cities and such large numbers of their inhabitants” and that Gómara would just as soon write “eight thousand” as “eight.” Eventually, tiring of criticizing Gómara at every juncture, Bernal offered a final portmanteau critique. There he was intent on pointing out that Gómara was wrong to attribute to Cortés alone things that his whole force had accomplished. Once again, he castigated Gómara (1982:296) for his habit of exaggerating numbers:
This chronicler says that so many thousands of Indians encountered us on the expedition and such high numbers have no rhyme or reason. And he also speaks of cities and towns and villages in which were so many thousands of houses, when there was not a fifth part of them ... He would as soon write one [or eight] thousand as eighty thousand. By this boasting he believes that his history will be pleasing to those who hear it, by not saying what really happened.13
These are strange words indeed from someone whom the High Counters represent as “supplementing to an extraordinary degree” the very historian whom he criticized so severely.
Bernal Díaz’s fate at the hands of the High Counters exemplifies, but scarcely exhausts, their penchant for tendentious selection and characterization of the materials they choose to use and abuse. There can be no doubt that the High Counters were quite aware of Bernal’s true feelings about the size of the population of central Mexico. Since they could not afford to ignore him altogether, they distorted his views beyond recognition with heavy-handed artfulness. Despite their efforts, Bernal remains a Banquo at their feast, a hovering presence that can be ignored but not exorcised.
The original is unfaithful to the translation. (Borges 1974:732)
No one would deny that translating is an art that can abide no timeless or straightforward rules, but there are general principles under translators ordinarily operate. The most important of these, but one of which the High Counters seem not to be aware, enjoins that any translation reflect the general linguistic usage of the time, as well as that of the author. Another axiom of translating, which the High Counters seem to understand more thoroughly, is that even the slightest twist in the interpretation of a word or phrase can have significant ripple effects. With this in mind I would like to examine several instances of translations that do not bear scrutiny well.
As I have noted elsewhere (Henige 1978a:220n), Cook and Borah (1971-1979, 1:378) saw fit to translate Columbus’ phrase “mayor que,” which he used to compare the size of Hispaniola with that of England, as “more populous.” Since they were far more interested in numbers of people than in numbers of square miles, they found this a handy way to suggest that Columbus had discerned a population in Hispaniola exceeding that in England. Columbus was wrong in thinking that Hispaniola was larger than England, and Cook and Borah were just as wrong in suggesting that Columbus was referring to people. The degree of calculation in each error is best left to the imagination. In his litany of baptism statistics Motolinía (1971:108) referred to one of his regions as “pueblos de la Mar del Sur.” This is rather ambiguous, but not so ambiguous as to justify translating it as “South Sea coast,” which was exactly what Cook and Simpson (1948:20) decided was most appropriate. Certainly it was useful for their own argument because it allowed them to claim that Motolinía excluded “most of Oaxaca, parts of Guerrero, and much of Michoacán,” forcing them to add 3.9 million to the totals they had devised from Motolinía’s figures (Cook and Simpson 1948:22). In their haste to aggregate, they neglected to mention that Torquemada, who followed Motolinía very closely, rendered (Torquemada 1977, 5:237) Motolinía’s phrase as “Michoacán and Matltzinco, which is the valley of Toluca, with their provinces,” thereby including at least Michoacán in Motolinía’s totals.
Recognizing a useful tactic, Henry Dobyns has proved to be the equal of the Berkeley School in the art of Mystical Translation. In fact, in his Their Number Become Thinned, he has managed to surpass them in the sustained application of the rule that a little mistranslation can go a long way.14 But Dobyns’ record is longer than this. In his attempt to calculate the ratio of depopulation for the entire New World Dobyns treated the reader to several examples of this tactic. For instance, he asserted (Dobyns 1966:405) that in 1609 a Spanish soldier “saw a total of sixteen million baptisms . . . recorded in a Franciscan convent at Xochimilco.” This conjures up an image of huge registers filling room after room with substantiable details of Indian baptisms—certainly a source to be reckoned with by those who would doubt the wisdom of believing in the calculating ability of Motolinía and others. But this was not quite what Dobyns’ source said. Quiros stated (Fernández de Quiros 1609 and 1866:507) that he “saw written in a convent . . . that the friars alone baptized sixteen million Indians.”
“Recorded” and “written” are not synonyms. Recording is a specialized form of writing and implies (as Dobyns must have known) a certain procedure of registering data in detail and/or on the spot.15 Quiros may have seen register after register, but all he said was that he had seen something written to the effect that sixteen million Indians had been baptized. It would be possible but pointless to chronicle the further adventures of the High Counters in the land of translating, but a systematic campaign to deploy mistranslations as a means to persuade doubters is evident. It betokens a contempt for both their sources and their readers.
I have shown in many instances, how partial you have been in your Citations, taking some fragments, or parcels of Sentences, or sometimes a short Sentence you thought might serve your turn, and always leaving what you could not but know would have destroyed your Notion or Argument [Brady 1684:325].
The High Counters’ practice of the genteel art of decontextualization brings to mind Ronald Reagan’s question: “Where is the rest of me?” Reading a citation or quotation in the writings of the High Counters is sometimes reminiscent of perusing a movie ad or a blurb for a bestseller—it really isn’t all there. Even the ellipses are missing . . . As far as I can tell, Henry Dobyns is in a class by himself as far as wrenching materials from any context but his own preconceptions. Already in 1966 Dobyns (1966:405) cited Motolinía’s nine million baptisms without indicating that Motolinía had offered other, lower, figures as well. By eliminating Motolinía’s self-contradictions Dobyns presumably hoped to make this excessively high figure more appealing by masquerading it as Motolinía’s only guess.
In his study of the Timucuan Indians and beyond, Dobyns suffers from what might be called hambre-phobia; that is, he goes to great lengths to expunge any mention of hunger in his allusions to the historical record. This aversion seem to be in aid of thrusting onto center stage the notion of epidemics (discussed below) as the chief cause of depopulation. At any rate, whenever the sources specify hunger as the sole or as a contributing factor, Dobyns eliminates it (Henige 1986b), often adding a phrase that makes it appear as though disease was actually cited when it was not.
No example of the High Counters’ studied indifference to textual integrity is more outrageous than their handling of Las Casas’ estimates of the contact population of Hispaniola. In all, Las Casas made no fewer than twelve of these (Henige 1978a:222-25). Five times he estimated that there had been about one million Indians, another six times that there had been about three million, and once that there had been “many” millions. But if Las Casas was confused, the High Counters seem not to have noticed it. Not one of them (Borah 1964:379; Cook and Borah 1971-1979, 1:376, 386, 393; Borah 1976:14) mentions the figure of one million, including Cook and Borah’s extended discussion of the sources for estimating that population.16
How can this be? It seems that the High Counters simply stop looking when they find the figures they want. If they discover these before they canvass an entire body of literature, so much the better. If they discover other, lower, figures along the way, no matter either; these are simply ignored. Which of these decidedly non-Bollandist alternatives this case represents is unclear. It might be sheer laziness since the lower figures generally appear in the earlier and less accessible writings of Las Casas. On the other hand, the figure of one million also appears in his Historia de las Indias, which Cook and Borah duly cite—for the three million figure.
More than any other discursive strategy, the High Counters’ penchant for treating texts as a chaos of broken atoms constantly allows them to appear to be startled by their own proofs. In this strategy they invoke the conventions of a ‘scientific’ style, according to which the disinterested scholar finds himself confronted with findings that impose themselves by their sheer inner cogency, quite independently of, if not in outright contradiction to, the scholar’s own working hypothesis. Nothing could be further from the truth.
In Larynx and Pharynx, two cities as large as Rouen or Nantes . . . there died of plague more than twenty two hundred and sixty thousand and sixteen people within the last eight days [Rabelais 1965:172].
The High Counters acknowledge Las Casas as their ideological ancestor, but events of late might have caused them some ambivalence in that regard. They must still admire the abandon with which he threw impossibly high numbers around. On the other hand, as their own estimates continue to escalate, they might well be dissatisfied with the causes to which Las Casas attributed Indian depopulation. Virtually without exception he ascribed it to hunger, to Spanish abuses of the encomienda system, and to their astonishing feats of slaughter in battle.
These reasons seemed to satisfy Las Casas but the High Counters have recently discovered their philosophers’ stone which enables them to reconcile both the low numbers described in the actual counts and the extraordinarily high numbers they posit as having existed just a few years earlier. This deus ex machina is disease—disease introduced by the Europeans, presumably from the moment Columbus set foot on Guanahani. Yet somehow Las Casas was oblivious to all this, even while claiming eyewitness status to the death of millions of Indians in Hispaniola and elsewhere (Henige 1978a:222-225). The High Counters never tell us how this could be, beyond disingenuously claiming that Las Casas “was merely the earliest and best-known commentator upon the biological [sic] plight of the Indian” (Dobyns 1966:412; cf. Dobyns 1983:24).
Of course the High Counters were not the first to notice the effects of the newly introduced diseases on the Indians, but they are among the first to claim (e.g., Dobyns 1983:34-35) both that the pre-conquest New World was virtually free of any infectious diseases at all and (e.g., Dobyns 1973:295-296) that most of the post-conquest mortality occurred very early and well beyond the purview of any recorded observation. Thus, according to Dobyns (1983:13, 217) the smallpox epidemic that struck the West Indies and Mexico (and possibly Central America) between 1518 and 1524 also “swept through all [sic] of the most densely populated portions of the Americas” and probably had a mortality rate of 50 percent or more. Voilá, before any Europeans could start counting Indians, at least half of them had already died! Las Casas, Cortés, and others not only were not exaggerating, they could only have been seriously underestimating the extent and speed of the depopulation.
There is no denying that from the standpoint of the High Counters this is truly a marvelous argument. Certainly no one holding this point of view can ever be refuted. One does not even need to produce any evidence—the argument presumes the absence of it as part of its own logic! Doubters might doubt, but they can do no more, and doubt is, we would all concede, rather like sticks and stones. As a result, this convenient doctrine appears on its way to becoming a new orthodoxy, at least with respect to North America. Under its terms, there is no figure that is too high; infer a sufficient number and variety of epidemics within a sufficient period of time (as Dobyns 1983:254-270 does for Florida) and any finite number can easily be reduced to any small number.
Unfortunately for the cause of the High Counters, Dobyns has not done an effective job in dealing with the evidence he adduces for such disease “episodes” (Henige 1986b) and he would best have left it entirely aside, leaving his case firmly entrenched in the domain of the Possible without attempting to whisk it to the land of the Real. Other attempts to interpret the historical record as suggesting earlier and wider-ranging epidemics have not been particularly successful either. In most cases the sources simply do not allow this—and willing meanings to resisting texts or drawing conclusions from their silence (“in the absence of any evidence to the contrary . . .”) are hardly acceptable substitutes.
It is all too easy to confuse the possible with the probable, or even with the real, particularly when in hot pursuit of a desired conclusion. Doing so as part of testing hypotheses is legitimate, in fact inescapable, historical method. But adopting it as standard operating procedure without the slightest effort at devil’s advocacy is unacceptable. For the moment—and apparently for all time—the notion that newly introduced infectious diseases quick assumed epidemic (or “pandemic”) proportions throughout the New World is, by its very nature, detached from any evidence. Accepting it is solely a matter of belief, an odd posture for the High Counters, who relish (as we see from Borah’s quote cited above) in claiming a certain “statistical” luster status for their work. In the meantime the study of the effects of well-documented disease episodes in specific contexts, while contributing to our knowledge of historical epidemiology, can have little effect on Native American historical demography beyond suggesting slightly higher population levels, a matter of saying something like X +, which is hardly ever enlightening.17
Whether it is the truth of history or fiction matters not, because the example is not supplied for its own sake, but for its significance [John Bromyard, quoted in Owst 1933:155].
What accounts for the phenomena discussed in this paper? Why is there at this moment an increased and sustained interest in the contact population of the New World supported by a slapdash and intuitive method of operation that ensures that only unsuspected—and indemonstrable—high population levels are “found” (a word the High Counters often use to describe their ‘success’)? Two attitudes predominate in the writings of the High Counters: a palpable sense of self-evident truth value and a belief that all questions have answers rather than, as would be more circumspect in the circumstances, that all answers have questions.
In referring to criticisms of Las Casas’ free use of large numbers, one commentator recently had this to say: “Such an approach has something repulsive about it ... It would like to prove Las Casas a liar but let the murderers go scot-free because they killed only 8, 5, or 3 million Indians instead of 20 million. That is the way the National Zeitung protects the German facists, claiming that not 6 million Jews were killed but at most 5 (Enzensberger 1974:13).” This is probably an extreme expression of this point of view, but not necessarily an extreme sentiment in the field of contact demography about the unquestioned accuracy of congenial texts. Many of the High Counters, especially those writing today, seem to see their work as having many of the trappings of a morality play. We see this in much of Dobyns’ work (e.g., Dobyns 1966:395-396; Dobyns 1973:292-294; Dobyns 1983), which fairly exudes with moral indignation. Not content with referring to DeSoto’s men as “Spaniards” or “soldiers” or even “invaders,” he prefers the harsher term “marauders.” And of course these “marauders” were armed, wittingly or unwittingly, with an entire arsenal of biological weapons. By this angle of vision, ‘enlightened’ scholars are to foster this image by inflating the number of victims that resulted—and keep inflating it until the specter of genocide appears. In effect, in “counting” millions of Indians they deliver a moral judgment, minus the evidence to support it (Borah 1976:19-20).
Rather paradoxically, the other main tenet of the High Counters is the belief that any number can serve as a proxy and render the unknowable knowable, the uncertain certain. This is to be accomplished by pressing into service techniques that will demonstrate that the size of at least part of the contact population can be measured and the measurement tested. Such procedures can then be encoded and extrapolated from one part of this universe to all others. The uncritical acceptance of Dobyns’ crypto-scientific 20:1 depopulation ratio is simply the most obvious manifestation of the view that, given time and technique, modern man can turn the silences of history into the facts of history.18 In this mode of thinking the very lack of information becomes a force in its own right, giving unbridled rein to ingenuity cast adrift.
Does disagreeing with this belief necessarily mean that there were not about (or even exactly) 25.2 million Indians in central Mexico in 1519? Or even one hundred million in all of the New World? Not at all. Is it likely that there were? Not at all. This latter view, prompted by common sense and enjoined by the available evidence, is only reinforced by the persistent but misguided efforts of the High Counters and particularly by the irredeemably damaging effects of their nonchalant attitude towards that evidence.
Close analysis of the materials used by the High Counters awaits the extended effort it will require and this obliges us to withhold final judgment of the value and the validity of the uses to which the High Counters have put them.19 In the meantime, to judge from the skill with which they have deployed the post-conquest data they have used, it is difficult to imagine that they have managed to treat pre-conquest materials any more responsibly (Zambarino 1980). Consequently, any interim judgment can only be that the very methods of the High Counters require us to reject their conclusions.
A final point remains to be made. Some readers may well wonder whether it is necessary to devote so much attention to appraising critically the methods of the High Counters. Does this particular dead horse really need to be flogged? Surely they condemn themselves by their own irresponsibility? Sadly, this is not true, or at least far from entirely true. As I suggested in opening this paper, the sedulously cultivated illusion of the High Counters has been, and continues to be, influential, to the point where the techniques they have developed—and their attitude towards sources—have sprouted clones in Africa (Henige 1986c) and Australia (Butlin 1983), in both cases spawned by the same fatal combination of the lack of evidence and the apparent need for expiation.20 Why this should be so is perhaps not for historians to surmise, but certainly historians have an obligation to test any set of arguments, if only to lend them credibility. In this case at least the High Counters, with their unerring propensity for the improbable, badly fail that test.
1 Without going into detail here, let me mention works by W.H. McNeill, Ladurie, and others, which attempt to provide a unifying framework for world history and in doing so feature disease prominently. See as well Crosby 1986; Hopkins 1983.
2 A recent survey of these estimates is Thornton (1988:22-32). Thornton inclines toward believing in larger contact populations but offers no estimate of his own and is not uncritical of the work of the high counters. Stuart (1987 45-46) mentions the contact population only in passing, also offers no estimate, and credits Dobyns, apparently without irony, with a “reassessment of the sources.”
3 For instance, Dobyns 1976 did not manage to cite a single work critical of the High Counters, thereby belying its sub-title, ceasing to be a work of reference, and becoming a propaganda tool. It is an uncommon man who can glimpse the possibility of conducting a vendetta in a bibliography.
4 It is revealing that in seeking to demonstrate what he meant by Bollan dist-like textual analysis, Borah argued against accepting the lowest estimate of the contact population of Tenochtotlan.
5 On this matter see, among many others, Phelan 1956 and Weckmann 1984.
6 As at least Torquemada (1977, 5:238) had the grace to notice, there were early criticisms of Motolinía’s claims. Already Gómara (López de Gómara 1943, 2:281-282) found them hard to believe. To my knowledge none of the High Counters have pointed this out, ignoring Gómara in this instance just as they ignore Bernal Díaz del Castillo in so many other cases (for which see below).
7 For examples of this in their use of post-conquest sources see Slicher van Bath 1978.
8 The dwelling count ploy reaches a crescendo in Cook and Borah 1971—1979, 2:25-38, concerning Yucatan.
9 Innumerable examples could be cited; for starters see Russell 1958.
10 For further criticism of the protein hypothesis see Conrad and Demarest 1984:166-70.
11 For accounts of similarly exaggerated estimates of human sacrifice in India and west Africa and for identical cultural reasons see Davies 1984:220-22; Graham 1965; Igbafe 1979:40-49, 70-72.
12 E.g., Marcus 1977:199-201.
13 In one of the surviving manuscripts Bernal used “1,000,” in another “8,000.”
14 For details see Henige 1986b.
15 Cook and Borah (1971-1979, 1:388) emulated Dobyns’ twist here, translating “se alcanza” as “may be found in the records” when “alcanzar” simply means to “find” or “glean,” without implying the existence of any records.
16 In fact, with his usual consummate misreading of the sources, Dobyns (1973:293) asserts that Las Casas estimated the aboriginal population of Hispaniola at four million, as well as that the only other “materially lower” figure was that of a “polemicist” (to the High Counters of course Las Casas was not a polemicist!) who was contemporary with Las Casas. Neither claim is true.
17 For a good overview of this issue see Joralemon 1982.
18 This roseate view is typified in Borah’s criticism of Angel Rosenblat’s work (Borah 1968:476) because it suggested that “scholars can disagree endlessly about which source is reliable, which ought to be discarded, which interpretation of the text is correct, and so on.”
19 For a detailed critique of the High Counters’ high conversion factors for tribute records see Rosenblat 1967:25-71 and Slicher van Bath 1978, to which—need I say it?—the High Counters have never responded in kind. Also see Doybns, Snow, Lamphear, and Henige 1989.
20 In a recent study of the contact population of Hawaii, Stannard (1989) has embraced the premises and methods of the High Counters, and with them the same brand of scattershot arithmetic, tendentious reliance on dubious comparative material, and studied indifference to critical literature in the field.
Alcina Franch, José. 1973. “Juan de Torquemada.” Handbook of Middle American Indians 13:256-275.
Anonymous. 1869. “The Later Life of De Foe.” British Quarterly Review 50: 483-519.
Baum, L. Frank. 1910. The Emerald City of Oz. Chicago: Reilly & Britton.
Blázquez Garbajosa, Adrián. 1985. “Las ‘Cartas de Relación de la Conquista de Mexico’: Política, Psicología y Literatura.” Bulletin Hispanique 87:5-46.
Borah, Woodrow W. 1964. “America as Model: The Demographic Impact of European Expansion Upon the Non-European World.” In Actos y Memorias del XXXV Congreso International de Americanistas (Mexico, 1962), 3: 389-397. Mexico City: Instituto Nacional de Antropología.
Borah, Woodrow W. 1968. Review of Angel Rosenblat, “La población de América en 1492.” Hispanic American Historical Review 48:475-477.
Borah, Woodrow W. 1976. “The Historical Demography of Aboriginal and Colonial America: An Attempt at Perspective.” In The Native Population of the Americas in 1492. William M. Denevan, ed. 13-34. Madison: University of Wisconsin Press.
Borah, Woodrow W. and Sherburne F. Cook. 1963. “The Aboriginal Population of Central Mexico on the Eve of the Spanish Conquest.” Ibero-Americana 45. Berkeley: University of California.
Borges, Jorge Luis. 1974. Obras completas. Buenos Aires: Emece Editores.
Brady, Robert. 1684. Introduction to the Old English History. London.
Butler, Samuel. 1970. Characters. Charles W. Daves, ed. Cleveland: Case Western Reserve University.
Butlin, N.G. 1983. Our Original Aggression. Aboriginal Populations of Southeastern Australia, 1788-1850. Sydney: Allen & Unwin.
Commager, Henry Steele and Elmo Giordanetti. 1967. Was America a Mistake? An Eighteenth-Century Controversy. New York: Harper & Row.
Conrad, Geoffrey W. and Arthur A. Demarest. 1984. Religion and Empire: The Dynamics of Aztec and Inca Expansionism. Cambridge: Cambridge University.
Cook, Sherburne F. 1946. “Human Sacrifice and Warfare as Factors in the Demography of Pre-Columbian Mexico.” Human Biology 18:81-102.
Cook, Sherburne F. 1948. “The Population of Central Mexico in the Sixteenth Century.” Ibero American 31. Berkeley: University of California Press.
Cook, Sherburne F. 1973. “Interracial Warfare and Population Decline Among the New England Indians.” Ethnohistory 20:1-24.
Cook, Sherburne F. and Lesley B. Simpson. 1948. “The Population of Central Mexico in the Sixteenth Century.” Ibero-Americana 31. Berkeley: University of California Press.
Cook, Sherburne F. and Woodrow W. Borah. 1960. “The Indian Population of Central Mexico, 1531-1610.” Ibero-Americana 44. Berkeley: University of California Press.
Cook, Sherburne F. and Woodrow W. Borah. 1966. “On the Credibility of Contemporary Testimony on the Population of Mexico in the Sixteenth Century.” In Summa Anthropologica. Homenaje a Roberto Weitlaner. 229-239. Mexico City: Instituto Nacional de Antropología e Historia.
Cook, Sherburne F. and Woodrow W. Borah. 1971-1979. Essays in Population History: Mexico and the Caribbean. 3 vols. Berkeley: University of California Press.
Cortés, Hernán. 1958. Relaciones de Hernán Cortés a Carlos Vsobre la invasion de Anáhuac. Eulalia Guzmán, ed. Mexico City: Libros Anáhuac.
Crosby, Alfred W. 1986. Ecological Imperialism. The Biological Expansion of Europe, 900-1900. Cambridge: Cambridge University.
Darwin, Charles. 1903. More Letters of Charles Darwin. Francis Darwin, ed. 2 vols. New York: D. Appleton & Co.
Davies, Nigel. 1984. “Human Sacrifice in the Old World and the New: Some Similarities and Differences.” In Ritual Human Sacrifice in Mesoamerica. Elizabeth H. Boone, ed. 211-226. Washington: Dumbarton Oaks Research Library.
Díaz del Castillo, Bernal. 1982. Historia verdadera de la conquista de la Nueva España. Carmelo Sáenz de Santa Maria, ed. Madrid: Instituto Gonzalo Fernández de Oviedo.
Dobyns, Henry F. 1966. “Estimating Aboriginal American Population: An Appraisal of Techniques with a New Hemispheric Estimate.” Current Anthropology 7:395-416, 440-444.
Dobyns, Henry F. 1973. “Review of Cook & Borah, Essays in Population History, I. Hispanic.” American Historical Review 53:292-296.
Dobyns, Henry F. 1976. Native American Historical Demography: A Critical Bibliography. Chicago: Newberry Library.
Dobyns, Henry F. 1983. Their Number Become Thinned: Native American Population Dynamics in Eastern North America. Knoxville: University of Tennessee.
Dobyns, Henry F., Dean R. Snow, Kim M. Lamphear, and David Henige. 1989. “Commentary on Native American Demography.” Ethnohistory 36: 285-307.
Ellenberger, D.F. 1912. History of the Basutos, Ancient and Modern. J.C. Macgregor, ed. London: Caxton.
Enzensberger, Hans M. 1974. Introduction in Bartolomé de las Casa, The Devastation of the Indies: A Brief Account. New York: Seabury Press.
Fernández-Armesto, Felipe. 1982. The Canary Islands After the Conquest. Oxford: Oxford University.
Fernández de Quirós, Pedro. 1609/1866. “Memorial dado a S.M. por el capitán Pedro Fernán dez de Quiros. In Colección de Documentos inéditos relativos al descubrimiento, conquista y organización de la antiguas posesiones españ-olos de América y Océania. Madrid.
García, Genaro. 1902. Dos antigos relaciones de la Florida. Mexico City: Aguilar
Gerbi, Antonello. 1973. The Dispute of the New World: The History of a Polemic, 1750-1900. Pittsburgh, PA: University of Pittsburgh.
Graham, James D. 1965. “The Slave Trade, Depopulation, and Human Sacrifice in Benin History.” Cahiers d’etudes africaines 5:317-334.
Harner, Michael. 1977. “The Ecological Basis for Aztec Sacrifice.” American Ethnologist 4:114-135.
Harris, Marvin. 1977. Cannibals and Kings. New York: Random House.
Henige, David. 1978a. “On the Contact Population of Hispaniola: History as Higher Mathematics.” Hispanic American Historical Review 58:217-237.
Henige, David. 1978b. “Response to Rudolph Zambardino.” Hispanic American Historical Review 58:707-712.
Henige, David. 1986a. “If Pigs Could Fly: Timucuan Population and Native American Historical Demography.” Journal of Interdisciplinary History 16:701-720.
Henige, David. 1986b. “Primary Source by Primary Source? On the Role of Epidemics in New World Depopulation.” Ethnohistory 33:293-312.
Henige, David. 1986c. “Measuring the Immeasurable: The Slave Trade, West African Population, and the Pyrrhonist Critic.” Journal of African History 27: 295-313.
Hollingsworth, T.H. 1969. Historical Demography. London: Hodder and Stoughton.
Hopkins, Donald R. 1983. Princes and Peasants: Smallpox in History. Chicago: University of Chicago Press.
Igbafe, Philip A. 1979. Benin under British Administration. London: Longman.
Joralemon, Donald. 1982. “New World Depopulation and the Case of Disease.” Journal of Anthropological Research 38:108-127.
León-Portilla, Miguel. 1979. “New Light on the Sources of Torquemada’s Monarquía Indiana.” The Americas 35:287-316.
Leonard, Irving A. 1948. Books of the Brave. Cambridge: Harvard University Press.
López de Gómara, Francisco. 1943. Historia de la Conquista de Mexico. Joaquín Ramírez Cabañas, ed. 2 vols. Mexico City: Editorial Pedro Robledo.
Losack, Torn. 1987. “The Kindreds, Houses, and Population of the Elves During the First Age.” Mythlore 51:34-38, 56.
Marcus, Raymond. 1977. “La conquête de Cholula: conflit d’interpretations.” Ibero-Amerikanische Archiv 3:193-213.
Mendieta, Jeronimo de. 1870. Historia Ecclesiastica Indiana. Joaquín Garcia Icazbalceta, ed. Mexico City: Antigua Librería.
Motolinía, Toribio de Benavente. 1941. Historia de los Indios de la Nueva España. Mexico City: Chavez Hayhoe.
Motolinía, Toribio de Benavente. 1971. Memoriales o libro de las cosas de la Nueva España y de los naturales de ella. Edmundo O’Gorman, ed. Mexico City: Universidad Autónoma Nacional de Mexico.
Ortiz de Montellano, Bernard R. 1983. “Counting Skulls: Comments on the Aztec Cannibalism Theory of Harner-Harris.” American Anthropologist 85:403-406.
Owst, Gerald R. 1933. Literature and Pulpit in Medieval England. Cambridge: Cambridge University Press.
Phelan, John L. 1956. The Millennial Kingdom of the Franciscans in the New World. Berkeley: University of California Press.
Price, Barbara J. 1978. “Demystification, Enriddlement, and Aztec Cannibalism: A Materialist Rejoinder to Harner.” American Ethnologist 5:98-115.
Rabelais, François. 1965. Pantagruel. V.L. Saulnier, ed. Geneva: Droz.
Rosenblat, Angel. 1967. La población de América en 1492. Mexico City: Colégio de México.
Russell, Josiah C. 1958. Late Ancient and Medieval Population. Philadelphia: American Philosophical Society.
Sahlins, Marshall. 1978. “Culture as Protein and Profit.” New York Times Review of Books (23 Jan):45-49.
Slicher van Bath, B.H. 1978. “The Calculation of the Population of New Spain, Especially for the Period Before 1570.” Boletín de Estudios Latino-Americanos y del Caribe 24:67-95.
Stannard, David E. 1989. Before the Horror: The Population of Hawai’i on the Eve of Western Contact. Honolulu: University of Hawaii, Social Science Research Institute.
Stuart, Paul. 1987. Nations Within a Nation: Historical Statistics of American Indians. Westport, CT: Greenwood Press.
Tapia, Andrés de. 1866. “Relación hecha el Señor Andres de Tapia sobre la con quista de Mexico.” In Colección de Documentos para la Historia de Mexico 2:554-594. Joaquín García Icazbalceta, ed. Mexico City: J.M. Andrade.
Thornton, Russell. 1988. American Indian Holocaust and Survival: A Population History Since 1492. Norman: University of Oklahoma Press.
Torquemada, Juan de. 1977. Monarquía Indiana. 5 vols. Mexico City: Universidad Autónoma Nacional de Mexico.
Weckmann, Luis. 1984 La herencia medieval de México. 2 vols. Mexico City: Colégio de México.
Zambardino, Rudolph A. 1980. “Mexico’s Population in the Sixteenth Century: Demographic Anomaly or Mathematical Illusion?” Journal of Interdisciplinary History 11:1-27.