On the Idea of the Indigenous

Author(s): John Brown Childs and Guillermo Delgado‐P.
Source: Current Anthropology, Vol. 40, No. 2 (April 1999), pp. 211-212
Published by: The University of Chicago Press on behalf of Wenner-Gren Foundation for
Anthropological Research
Stable URL: http://www.jstor.org/stable/10.1086/200005
Accessed: 04-08-2016 06:43 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.

The University of Chicago Press, Wenner-Gren Foundation for Anthropological
Research are collaborating with JSTOR to digitize, preserve and extend access to Current Anthropology

This content downloaded from on Thu, 04 Aug 2016 06:43:15 UTC
All use subject to http://about.jstor.org/terms

These alliances were premised on a profound sense of affiliation among distinct cultures and regions based on both their resistance to further European expansion and on their ‘‘nativeness. Be´teille is concerned not with the thinking of indigenous peoples but with a distant how best to ‘‘define’’ them.21 on Thu. should adopt a detached social science perspective unaffected by the alleged emotionalism of certain anthropologists whose ‘‘state of moral excitation’’ is responsible for the very idea of the ‘‘indigenous. Rather. does open up ground upon which mutually respectful interactions of indigenous peoples and anthropologists can develop (Varese 1996).156. Such defining. arrangements between Seneca and archaeologists have been made which allow for excavation in certain areas. however.and 19th-century examples such as the strategic efforts by leaders including Joseph Brant (Mohawk). have handsomely benefited from their interactions with indigenous peoples. 13 vii 98 The title of Andre´ Be´teille’s essay. Rather. morally excited or not. the Andean Pachakutiq (‘‘Turn of Times’’) movements. or overstressing their pristine-isolated status).’’ The development of the current concept of the ‘‘indigenous’’ is not. the development of the term ‘‘indigenous’’ by and for the people to whom the term applies is itself but a chapter in a centuries-long story of resistance to colonial disruptions. some stretching from the Gulf of Mexico to southern Canada. but also identity and power. indigenous activists in the Americas adopted the Haudenosaunee ‘‘Iroquois’’ strategy of the early 1920s. the American Indian Movement.’’ nor is it the result of anthropological interference. Merrill College. while it does not come from anthropology (which historically has often contributed to reducing ‘‘indigenous peoples’’ to a class-defined rural peasantry. especially during the 20th century.S. Meso-Mayan repositioning that privileges Indigenous-Mayan rebuilding over colonial nationstates of Mesoamerica.A. producing research that only in limited cases has been returned to native communities. new generations of native intellectuals have been formed through intense interaction be- 1. both in the Americas and around the world. 95064.jstor. merely a recent ‘‘awakening of old memories. this concept is being forged by numerous communities of indigenous peoples themselves. Consequently. there have been anthropologists who have consistently accompanied such endeavours. Calif. At the ancient Seneca historical site of Ganondagan in western New York.p. is another significant chapter in the long history of their resistance to European domination. the development of the concept of ‘‘indigenous. For example.org/terms . In the Americas. Granted. Tecumseh (Shawnee). As a consequence of this.Discussion and Criticism On the Idea of the Indigenous j oh n b r o w n c hi l ds a n d gu i l l e r m o d e l g a d o . he argues. when a delegation traveled to be heard at the League of Nations. Permission to reprint items in this section may be obtained only from their authors.’’ linked to ethnic subjectivity by peoples who say that they have adopted this term. 04 Aug 2016 06:43:15 UTC All use subject to http://about. as Be´teille has it. In the early ’80s.’’ However. Pontiac (Ottawa). U. the term ‘‘indigenous’’ is not. and finally the legitimation of hemispheric Native American studies in universities. University of California. We would suggest the ‘‘alterNative’’ title ‘‘Indigenous Peoples’ Ideas. This self-defining.’’ As Gregory Evans Dowd (1992) points out. This is to say that official history has almost consistently narrated the ‘‘native’’ in terms of barbarism and against extensive uprooting and displacement. as in the Barbados Meetings of anthropologists and natives.’’ as can be seen in movements as diverse as that of the Ghost Dance. and Gero´nimo (Apache) to create broad alliances of diverse native peoples. however. 1 tween certain anthropologists and indigenous peoples. Since then indigenous peoples have been insistent in promoting a charter of human rights in which collectivity and self-determination are stressed. thus adding 211 This content downloaded from 202. anthropologists have contributed significantly to making indigenous voices heard. we have come to understand how histories have been hysterias.92. Most anthropologists. these alliances involved ‘‘not just strategy. starting with a reevaluation of historical narratives and the insistent questioning of exclusionary grand narratives. One thinks of North American 18th. This declaration has finally been drafted after solid work by the UN’s Working Group on Indigenous Peoples.’’ From the 17th century to the present there have been numerous overlapping efforts based on a sense of what today is being called ‘‘indigenous. Santa Cruz. the result of any anthropological impetus. the 500-year Pan-Indian movement to reunify the People of the Eagle (North) and the People of the Condor (South) (easily involving some 50 million indigenous people). is indicative of an approach that is open to question. ‘‘The Idea of Indigenous People’’ (CA 39:187–91). Wild Cat and John Horse (Seminole/Black Seminole). often devoid of ethnic and gender specificities. and in several cases.

we have d o w d.S. 04 Aug 2016 06:43:15 UTC All use subject to http://about. s t e f a n o.92. A spirited resistance: The North American Indian struggle for unity. 2301 EB Leiden. If the actual number equals the equilibrium number. Similarly. soberanı´a y globalismo. is ‘‘perhaps the most important and least recognized need of the human soul. one can find the values of wt . London: Unwin Hyman. is a function of St and the actual number of people Nt . but they can probably represent them as co-workers. Quito: Abyayala. If the actual number of people is above the equilibrium value. The Netherlands. NY: Chauncy Press. s i m o n e. 109) gives four equations to describe the behavior of these variables with respect to each other and to time. Pueblos indios. recognition of multicultures enriches. They can no longer represent them. 1986. and therefore we could write S without the index t. Wood labels this equilibrium value St . are basically functions of wt . These are Wood’s equations 2–5. (2) dt ⫽ δ0 ⫺ δ1 ln wt ⫹ δ2 bt . Box 11090. 1989. 1954. They can. but I will stick to Wood’s general notation. ‘‘A word is nothing until it is given form and set to work in the world. (1) bt ⫽ β0 ⫹ β1 ln wt ⫹ β2 dt . Being in a secure position. (3) 1/Nt ∂Nt /∂t ⫽ bt ⫺ dt. articulating complex processes of ethnogenesis (Varese 1996). Well-being at equilibrium is called θ. w.’’ the Haudenosaunee Mother of Nations said to The Peacemaker who helped to create the Five Nation Confederacy (Wallace 1986). criticizing certain aspects of the mathematical model he uses to describe the dynamics of Malthusian systems of population. that is. a nd r e´ v a n d o k k u m Foundation for Theoretical Research in Anthropology. Substituting and solving 4. In a given Malthusian system St will usually be constant (pp. finally. After all. bt .21 on Thu. For the sake of clarity I will repeat them here: wt ⫽ θ(St /Nt)κ. The need for roots: Prelude to a declaration of duties towards mankind. This content downloaded from 202. but to suggest that anyone who feels rooted is the equivalent of a Nazi is as wrong as saying that democracy must be bad because at certain moments it condoned and produced slavery and genocide. and Latin/o American anthropologists continue to accompany indigenous peoples’ struggles at their request. multiplying both St and Nt by the same arbitrary factor yields the same value of wt as without such a factor. since the condition ‘‘bt ⫽ dt whenever Nt ⫽ St’’ yields θ ⫽ exp [(τ0 ⫺ σ0)/(τ1 ⫹ σ1)]. w a l l a c e. respectively. On the Theory of Preindustrial Population Dynamics References Cited where κ is a given shape parameter which is ‘‘determined by the details of the curve of declining marginal productivity of labor. Saranac Lake. New York: Routledge.156. Conflict in the archaeology of living traditions.O. (4) and. 1745–1815. g r e g o r y e v a n s. 1992. Editor. The white roots of peace. contribute to it precisely because indigenous peoples are defining themselves. the number of people will decrease over time in the direction of the equilibrium. I will focus on the Malthusian part of Wood’s article. Below the equilibrium the number will increase in that direction.jstor. 108.212 c ur r e n t an t hr o p o l o g y to knowledge about Seneca/Haudenosaunee history and philosophy. As is illustrated by the case of the environmentalists. Furthermore. The idea of ‘‘claims of blood and soil’’ that Be´teille employs in criticizing indigenous intellectual selfdetermination is misleading. w e i l. 110). monocultures collapse.’’ as the anti-Nazi French writer Simone Weil (1952) said.. In the ‘‘Malthusian systems’’ concerned. ‘‘Well-being. dt . v a r e s e. Such anthropologists are not the cause of the indigenous renaissance. 1996. however.’’ Note that wt is independent of scale.’’ wt . p a u l a. Wood (p. that number of people will remain stable. U. and Nt as functions of t for given St ⫽ constant.’’ The self-definition of ‘‘indigenous’’ by and for indigenous peoples is related to the presence of strong communities that can ideally and materially offer sustenance to their members while also being able to reach out to others from their positions of strength. European. Editor. The ways in which native peoples are giving form to the word ‘‘indigenous’’ so that it can do good work in the world should be cause for celebration rather than condemnation. while protecting sacred ground and burial places. Baltimore: John Hopkins University Press. there is an equilibrium number of people if the means of production of the people involved remain constant. Malthus and Ester Boserup. r. bt and dt respectively.org/terms (5) . The birth and death rates. P. Malthusian systems with fixed productivity parameters. l a y t o n. ‘‘To be rooted. 8 vii 98 In his interesting article on population dynamics. indigenous peoples can welcome the strategic contributions of allied efforts by outsiders to carry out full decolonization (Layton 1989). The parameter θ can be eliminated in the results. The imagery of ‘‘blood and soil’’ harks back to Nazi claims of Aryan superiority. Wood (CA 39:99–135) offers proposals for a demographic theory combining the ideas of Thomas R.

The question remains whether one can use such formalisms as if they were only mechanical devices without content. an instance of birth has the effect of increasing the death rate through Nt . since land is part of the definition of St (p.3 Negative feedback in equation 3 can occur with reduced suicide rates for people with children (Durkheim 1967[1897]:207–8) or (though with a time lag) a delay of death in old age when people are being cared for by their grownup children. Lee (1993:7) states: ‘‘High mortality creates more widows and widowers. Another construction in 2 and 3 is debatable. If equation 1 is used in equations 2 and 3 (or. In China it was often considered improper for a widow to remarry. which of κ ⬍ 1. There is another problem with the system 1–4. Finally. Thus the necessity of including positive values for β2 and δ2 is far from clear on intrinsic grounds. this will reduce rather than increase fertility (Wrigley 1969:15).g. Any instance of death has the effect of reducing Nt (4). This is not what we would expect. in equations 11 and 12). Proceed as follows: Take the κth root on both sides of equation 1/κ 1/κ 1. If mathematical convenience is an objective. (10) which is superexponential. σ1 ⫽ (β1 ⫺ β2 δ1)/(1 ⫺ β2 δ2).) Finally. for example. Its removal does not alter the behavior of wt relative to St and Nt . He only writes that they were introduced ‘‘purely for mathematical convenience’’ (p.92. but this would require some explanation. The information that The fact that the parameter θ is entirely determined by the necessity that the system 1–4 (together with the mentioned condition) be solvable makes it wholly artificial. for it means that an increase in Nt has a greater impact on the change of bt and dt when population pressure is not yet felt than when such pressure is actually beginning to be felt. One must still consider the meaning of the resulting expressions. To give one Boserupian disgression: The ‘‘acceptable standard of living’’ of which Wood (p. the parameter κ is introduced by Wood without much of an interpretation of its function and value (e. as w′t .jstor. or κ ⬎ 1 is more likely to be the case in what situations). 3.org/terms . for that matter. θ* ⫽ θ . indicated. which again has an increasing effect on the birth rate (1 and 2). the inclusion of cross-references seems to me in some cases redundant. Since θ is made up of the β and δ parameters. κ ⫽ 1.) Wood offers no justification for using logarithms in 2 and 3. leaving β2 and δ2 out of the analysis will greatly enhance the workability of the equations without much damage. Here Wood neglects the explanatory possibilities of his own model. This suggests that κ (defined by Wood as being dependent on aspects of production) can be interpreted as having something directly to do with the birth and death rates. Then define w* t ⫽ wt . who generate additional marriages and higher fertility. the cross-reference construction can be criticized on the ground of its supposed meaning. the cross-reference of dt in bt and vice versa. But negative correlations can easily be found: For example. If bt and dt are written as functions of Nt . (6) dt ⫽ τ0 ⫺ τ1 ln wt . giving the impression that θ is given externally to the model. θ does not provide any extra information. resulting in 1.’’ The argument ‘‘generate additional marriages’’ is not generally valid.2 Wood’s initial equations (here 1–4) lead to the following expression of Nt : Nt ⫽ St (N0 /St)exp [⫺ (σ1 ⫹ τ1)κt]. 109) mentions only examples of positive correlations between bt and dt . Then for instance ‘‘β1 ln wt’’ reduces to ‘‘β1κ ln w* t ’’ ⫽ This content downloaded from 202. Wood ignores this. the absolute values of the partial derivatives ∂bt /∂Nt and ∂dt /∂Nt decrease for increasing Nt. 109. namely. Wood (p. Especially in Wood’s (p. 109) example of land’s becoming available for new couples (on the death of someone) the theoretical duplication is rather obvious. Mathematically they can be eliminated immediately. perhaps in order to make the parameter κ manageable (the differential equation 4 would otherwise become rather awkward). (8) τ1 ⫽ (δ1 ⫺ δ2 β1)/(1 ⫺ β2 δ2). 109) should be replaced by an equivalent one about the β and δ parameters. Thus the logarithms in 2 and 3 create problems with respect to the theoretical justification of the shape of the Nt curve and with respect to the partial derivatives ∂bt /∂Nt and ∂dt /∂Nt . 2. Furthermore. we would expect them to increase (absolutely) with increasing Nt. elimination immediately reduces 2 and 3 to the simpler 11 and 12.1 I suggest that θ be left out of the modeling. I submit that storage of κ in 2 and 3 does not alter the behavior of the system 1–4 if we just introduce new 4 parameters β* 1 ⫽ β1κ and δ* 1 ⫽ δ1κ. Wood’s comment that ‘‘θ can be thought of as fixed by biology and thus constant for all systems of production’’ (p.21 on Thu. (A fortunate corollary of this is that in Wood’s equation 18 [p. Now. if the κth root of θ is taken. (11) (12) σ0 ⫽ (β0 ⫹ β2 δ0)/(1 ⫺ β2 δ2). especially in the Ming period (Linck-Kesting 1985:146–52). 116] the complicated λ parameters can be removed. (9) Again. Lee (1993:7) also gives only positive examples. the logarithm makes it possible to multiply κ by β1 and δ1 (or σ1 and τ1). 108). my emphasis). 4. (N0 is obtained by setting N0 ⫽ Nt⫽0.156. without weighing them against negative ones. 111) speaks can without θ-like parameters be directly expressed as a function of St and Nt. Equation 1 then becomes w* t ⫽ θ*(St /Nt). Perhaps there are good reasons for defining bt and dt as in 2 and 3. In India a widow was subject to sati. If the partial derivatives for different values of Nt differ at all. because there are already positive feedback relations between birth and death through Nt .. it is apparent that these rates depend logarithmically on Nt . if one of the spouses of a married couple dies. which meant that she had to die when her deceased husband was cremated (Rao 1975: 233). a construction resulting from the logarithms in 2 and 3. (7) τ0 ⫽ (δ0 ⫹ δ2 β0)/(1 ⫺ β2 δ2). Analogously.Volume 40. 04 Aug 2016 06:43:15 UTC All use subject to http://about. Number 2. April 1999 213 where bt ⫽ σ0 ⫹ σ1 ln wt .

θ. Spiegelman 1970 [1955]:408 and Luther et al. The two alternative systems do not add anything to Wood’s general principles concerning Malthusian systems. (14) dt ⫽ δ0 ⫺ δ1 wt.’’ especially the logistic variant. Cohen (1995:343–44) provides a differential equation with an St-like parameter which can change over time. otherwise we would expect to find terms var(κ) and var(θ). 1987:22.org/terms . 394). 1 ⫹ (St /N0 ⫺ 1) exp [⫺ (β1⫹ δ1)Stt] site direction. but this θ would be subject to the same restrictions as Wood’s original θ. which makes wt independent of scale. Cohen calls his equation ‘‘the equation of Malthus. we work in the oppo‘‘β* ln θ*(St /Nt)’’..214 c ur r e n t an t hr o p o l o g y κ is supposed to represent will then be conserved in 1/κ β* 1 and δ* 1 and a new θ* ⫽ θ . found independently by Pearl and Reed (e. Pearl and Reed used the logistic in order to predict asymptotes of Nt (thus. dt . the minus sign is necessary in order to obtain the right value for var(wt). zero. The shape of that differential equation resembles that of 16 (with 13. though not uniformly so across all households. 14.g. The systems have constant partial derivatives ∂bt /∂Nt and ∂dt /∂Nt . Nt)/[E(St)E(Nt)]’’ should be ‘‘⫺ 2 cov(St. var(St) ⫽ var(Nt) ⫽ cov(St. and Nt be given as follows: wt ⫽ St ⫺ Nt . Since in such a case wt is identical for all households.’’ It is a pity that Wood does not explain why he chose to deviate from this ‘‘equation of Malthus. Let wt . Nt)/[E(St)E(Nt)]}. That it should be minus is clear when one takes into consideration that the formula should be able to account for the situation in which all households have identical κ and θ and are in demographic equilibrium though with different St ⫽ Nt. (19) The study of variances. The highest value of var(wt) should be reached when St and Nt are anti-congruently distributed across the households. One result of this is that St appears in the exponent of Nt in 17. This means that the conclusion Wood draws from his equation 15. κ. the logistic. 04 Aug 2016 06:43:15 UTC All use subject to http://about. ‘‘the variance [of wt] is inversely related to the 6. getting rid of the logarithms (unjustified). wt ⫽ 0 in equilibrium situations. namely. (21) is invalid. thus making theoretical experimentation possible. for basic explanation. 16). Substituting and finally solving 16 lead to Nt ⫽ St . The condition ‘‘bt ⫽ dt whenever Nt ⫽ St’’ gives β0 ⫽ δ0.7 In the case that there is an overall equilibrium. and furthermore the member ‘‘⫺ cov(St . They are designed to show how a particular set of general principles can be translated into different formalisms with different consequences if a few changes are made within the scope of the formalism. St and Nt work in opposition to each other. θ (artificial). 114–17 Wood investigates the effects of the variances across households of St and Nt on the variance of wt. in the system 1–4 we would consequently get a differential equation with Nt multiplied by β0 and δ0 rather than by β1 and δ1. dt . dt . without β2. δ2.6 This means that Wood’s equation 15. A characteristic feature of alternative 1 is that wt is scale-dependent. To avoid the zero value of wt one could add an additive θ in 13. Alternative 1. Therefore I propose two alternative systems. thus when cov(St. The shape of 21 suggests this. e. More generally. var(wt) ⫽ 2[κθ/E(St)]2[var(St) ⫹ 1]. and Kish 1940 and. for when all households are at demographic equilibrium var(wt) will simply be zero. as explained above. In 13. and κ (storable in β1 and δ1) would make the modeling at least more apparent and probably also more realistic. Here St functions as a weighting factor of St ⫺ Nt. Nt) is maximally negative. when representing Malthusian systems in his own equations.g. Nt)’’ should still be included in the equation.5 The shape of 17 resembles that of a well-known formula. when all households have identical κ and θ. and 15 substituted). bt. wt ⬍ 0 in situations of population overshoots. However. β2 and δ2 (ambiguous). Two logistic alternatives to Wood’s model.jstor. and Nt be given as follows: wt ⫽ (St ⫺ Nt)/St and bt . which proved to be a premature enterprise (Cohen 1995: 345 n. 5. (16) Substituting and finally solving 16 lead to St Nt ⫽ . and Nt as in alternative 1.92. This content downloaded from 202. (15) 1/Nt ∂Nt /∂t ⫽ bt ⫺ dt . and E(St) ⫽ E(Nt). as is required for the Malthusian situations under study and analogously to the asymptotical behavior of Nt in 10. see his equation 14: var(wt) ⫽ [κE(wt)]2 ⫻ {var(St)/E(St)2 (17) (18) ⫹ var(Nt)/E(Nt)2 (20) ⫹ 2 cov(St. bt . In the models discussed here (Wood’s as well as mine) this is not a problem. Let wt. Alternative 2.21 on Thu. we look for the best-fitting models. var(Nt) will in general not be equal to var(St). On pp. and logarithms but with wt defined basically as a function of the difference between St and Nt rather than their proportion. and wt ⬎ 0 if the population is less than the equilibrium level. with St given. Dropping the asterisks yields the original equations but without κ. Logistic functions form a subset of the total set of solutions. St).156. Nt)/[E(St)E(Nt)]’’ (with minus). since St is considered to be expressable in terms external to the set of parameters in 1–4 and 13–16. Again. 7.. The member ‘‘⫹ 2 cov(St. In summary. which appears to be the shared asymptote of bt and dt and therefore can be read off directly. Reed. Nt). 1 ⫹ (St /N0 ⫺ 1) exp [⫺ (β1 ⫹ δ1)t] Nt is stable if N0 ⫽ St and will asymptotically tend to St if N0 ≠ St. (13) bt ⫽ β0 ⫹ β1 wt . namely. Pearl. In other words.

It can then immediately be seen from equation 20 (Wood’s 14)— whether with plus or minus cov(St. Littleton and Attenborough (p. In the case that the random disturbance terms are both identical with zero. but as it stands the argumentation is unconvincing. E(St) in itself plays no effective role therein. Nt)—that if we multiply all attached St and Nt by an arbitrary factor. κlike parameters could be reintroduced in 1. Also. 11. Wood. it is simply due to reorderings of the quotients (St /Nt). Malthus himself (1989[1798]:81) paid attention to negative effects of increasing scale for the ‘‘lower classes of society.jstor. If there is any change in the variance across households when the overall level of demographic saturation increases. alternative 1 confirms Cowgill’s assertion (p. Thus in alternative 1. Wood’s 18 should reduce to his 14 (20 in this comment). If all the particular St’s and Nt’s are multiplied by the same factor a. and cov(St. 116).92. this parameter can have an alternative use in equations 2 and 3. This is another example of how the particular translation of general principles into a formal model affects the outcomes. suggesting an interpretation that productivity has something to do with the β1 and δ1 parameters. 116 Wood introduces demographic stochastic variation and then writes: ‘‘Periods of demographic growth will usually be characterized by widening variation in well-being. Number 2.’’ Cohen (1995:341) states: ‘‘Along with human population. The results of changes in St or Nt for some household can be accounted for without having to determine the new E(wt) first. In other words. (22) It can easily be seen that var(wt) is. theoretical as well as empirical. From 13 we derive var(wt) ⫽ var(St) ⫹ var(Nt) ⫺ 2 cov(St .9 8.156. to be able to choose between scale-dependent and scale-independent parameterizations. April 1999 215 square of the mean value of demographic saturation’’ (p. Differences in var(wt) can occur only if the quotients St /Nt change. we finally have a multiplication by 1. though not uniformly so within each household. var(wt) will be multiplied by a factor a2. which it does not. chooses only one parameterization and presents results with respect to (especially) var(wt) without commenting that such results may be contingent on the particular parameterization chosen. A wt parameter which couples nondemographic indicators with demographic ones would be more useful. dependent on scale. and cultural habits (pp. This removes the confusion that E(wt) raises with respect to var(wt). Leaving the θ. Consider a situation in which there is overall equilibrium. 122) that ‘‘social surpluses tend to be extracted from households that have little power. as we have seen.10 The difference between alternative 1 and alternative 2 is an example of how a single alteration in a model can have a profound impact on the results of the modeling (cf. things simply held invariant in the Malthusian models under study. For alternative 2 we get as a rough approximation of var(wt): var(wt) ⫽ var(St) E(Nt)2 /E(St)4 ⫹ var(Nt)/E(St)2 ⫺ 2 cov(St. Nt). Unfortunately.’’ effects that are counterbalanced only by changes in social organization. technology. is unfounded. a. In their comment on Wood’s article. 125) suggest that ‘‘wellbeing’’ and Wood’s ‘‘wt’’ are not synonymous. now with more concrete significance. var(Nt).Volume 40. 115–16). (23) which is independent of scale. 22 and 23 are less complicated with respect to the variables contained than 20. equation 1 does not contain a real hypothesis (neither do its derivatives 13 and 18. like wt itself. Attach to all households the same κ and the same θ and arbitrary values for St and Nt. thus with E(St) ⫽ E(Nt) but in general with Nt ≠ St for each household. 9. and δ2 parameters out of the model10. because at least in equilibrium situations (even if the equilibrium is not uniformly distributed across households) its increase is annulled by the increase in var(St). one can also calculate var(wt) for a particular society and do the arbitrary multiplication test to see that var(wt) is independent of scale. the comment of Littleton and Attenborough [pp. 23 does not contain a reference to E(wt). β2.8 In alternative 1. inventing a satisfactory wt would require a whole new article). This may be so. Introducing St as a weighting factor (see 13 and 18) has effects with respect to the scale-dependency of wt and var(wt). var(wt) will increase with increasing scale rather than decrease as Wood has it (pp. Nt will decrease and E(wt) will increase for all households. It is interesting to investigate how the variance of wt will react to scale differences here. Perhaps the difference between 23 and 20 is not very dramatic. The point is that wt is essentially made up of demographic parameters and is consequently somewhat tautologically defined with respect to the internal dynamics of the system 1–4. and I agree with them. This leaves the reader with the impression that the results presented stem from the general Malthusian principles. since all households will eventually reach their respective equilibrium levels. Wood’s equation 18 is highly obscure. but it will require a great deal more research. 84–86).’’ This is more in line with my alternative 1. according to 1–4. Later on p. E(St)2 will increase by a factor a2. In this sense. The point is that both alternatives represent general Malthusian principles. Here the increase of scale has an ambiguous effect. wt is scale-dependent.21 on Thu. Where all households have arbitrary population overshoots. but at any rate. but with ⌺(St) ⫽ ⌺(Nt). 125–26]). Nt) E(Nt)/E(St)3. Yet a sentence later he says that there is an inverse relationship between var(wt) and E(Nt). 04 Aug 2016 06:43:15 UTC All use subject to http://about. Since E(wt) remains the same. Though Wood brings in the κ parameter in equation 1. the inequality in the distribution of global income has grown in recent decades’’ while there has been ‘‘an immense improvement in average well-being’’ (1995:344). however. This content downloaded from 202. Here a possible explanation would be that in a society of larger scale there is more for the rich to accumulate relative to (and probably at the expense of) the poor. Like 22. say. E(wt) will generally not equal θ. but so will the var’s and the cov (by the distributivity of multiplication). Of course. but var(wt) will decrease. Nt).11 Conclusions.org/terms .

In fact. f r i t z r e u t t e r. Adams 1998).’’ so to speak (Geertz 1973:7. e. They place a heavy burden on Wood’s model. Science 269:341–46.’’ in Geschlechtsreife und Legitimation zur Zeugung. pp. s p i e g e l m a n. j a m e s w. 102) view that empirical data need theory to be understood. 1987. a . Alternatives 1 and 2 show that a change in the use of just one parameter can affect important predictions of a model. current anthropology 39:125–26. formal theories might be developed—something that has sometimes proved difficult to argue in the anthropological discipline (cf. Comment on: A theory of preindustrial population dynamics. that it is indeed possible to prefer one construction over another on the basis of general considerations. Le suicide: Etude de sociologie.21 on Thu. 1940. ‘‘Suicide in India. e m i l e. r o n a l d d. Wood. 1967(1897). 04 Aug 2016 06:43:15 UTC All use subject to http://about. since they are doing the real demographic job. d . pp. This content downloaded from 202. Population and history. k i s h. g e o r g e l. first.k e s t i n g. 1973. l i t t l e t o n. d u r k h e i m. l e e. 1998. second. m o r t i m e r. I fully share Wood’s (pp. The analysis of var(wt) which Wood presents is confusing. are exactly not congruent with Wood’s logarithmic functions in 2 and 3. Introduction to demography. r a o. by James W.’’ in Suicide in different cultures. r a y m o n d. His article provides a good example of how general and.org/terms . 85– 176. A little more attention to the meaning of the mathematics—more ‘‘thick description. a n d r. c o w g i l l. Population growth and earth’s human carrying capacity. Cohen’s model is theoretically at least as sound as Wood’s and mathematically more convenient. 1998. Gewo¨nliche Differentialgleichungen: Analytische und numerische Behandlung. current anthropology 39:99–135. Investigation of the β1 and δ1 parameters would have been more appropriate. k l a u s n i e d e r d r e n k. economy. that performing thought experiments is useful for revealing more of the meaning of proposed models and. London: World University Library. Braunschweig: Vieweg. c o h e n. The logistic curve and the census count of 1940. Wood (p. An essay on the principle of population. with an inquiry into our prospects respecting the future removal or mitigation of the evils which it occasions. or A view of its past and present effects on human happiness. a. Comment on: A theory of preindustrial population dynamics. Edited by Norman L.92. 1989(1798). Cambridge: Harvard University Press. r e e d.jstor. It would have been nice if attention had been given to the contrast with constructions leading to logistic-like solutions such as Cohen’s (1995). h . a t t e n b o r o u g h. current anthropology 39:122. Freiburg: Verlag Karl Alber. 1993. c l i f f o r d. 2. Farberow. A theory of preindustrial population dynamics: Demography. and their usefulness is questionable. especially 10C. Cambridge: Cambridge University Press. v e n k o b a. 1969. 1975. 1998. 231–38. Accidental and systematic change in population history: Homeostasis in a stochastic setting. and well-being in Malthusian systems. Unfortunately. 1998. Vol. after Ryle)—would have been helpful. Baltimore: University Park Press. a n d j o s e p h f. These figures. l i n c k . References Cited a d a m s. 100. w o o d. 1970(1955). What I have tried to show here is. As to the shape of equations 1–3. 1995. New York: Basic Books. his philosophy of model building (‘‘purely for mathematical convenience’’) leaves many questions unanswered. but no justification is given. Science 92:486–88. 1985. The interpretation of cultures: Selected essays. a n d h a r r y y s e r e n t a n t. g e e r t z. w o l f r a m. current anthropology 39:137. j o e l e. g u d u l a. Edited by Ernst Wilhelm Mu¨ller. Most important. he does not investigate to what degree this analysis depends on the particular parameterization chosen.216 c ur r e n t an t hr o p o l o g y ing process would do little if any harm. Wood. t h o m a s r o b e r t. figure 1B is the standard Pearl–Reed situation (with bt ⫽ β0).156. m a l t h u s. 130) himself states in his reply to Konigsberg that his figure 1B or 10C is likely to represent preindustrial populations. l u t h e r. r i c h a r d n. Paris: Presses Universitaires de France. by James W. Explorations in Economic History 30:1–30. w r i g l e y. On causation and the struggle for a science of culture. j. p e a r l. more particularly. l o w e l l j. Wood’s decision to define wt as in 1 and to use logarithms in 2 and 3 may be justifiable. . ‘‘China: Geschlechtsreife und Legitimation zur Zeugung.