Systen

Kent V.

Chapter 20 KENT V. FLANNERY

tant tha
came (
Archaeological Systems Theory
besides
involve(
and Early Mesoamerica I beR
food-eol
began ~
B.C., car
ual chili
tems, n:
seasonal
that not
during d
result ot
AS WORK on the early periods of Mesoamer a culture adapted to a particular environ ouslyexi
ican prehistory progresses, and we learn more mental zone: "oak woodland," "mesquite that the 1
about the food-collectors and early food-pro grassland," "semitropical thorn scrub," "trop to see
ducers of that region, our mental image of ical forest," and so on. New data suggest, first, change l
these ancient peoples has been greatly modi that primitive peoples rarely adapt to whole In the
fied. We no longer think of the preceramic "environmental zones" (Coe and Flannery apply, OJ
plant-collectors as a ragged and scruffy band 1964: 650) . Next, as argued in this article, it ecosystel
of nomads; instead, they appear as a prac appears that sometimes a group's basic "adap Vayda (
ticed and ingenious team of lay botanists who tation" may not even be to the "micro-envi modifical
know how to wring the most out of a super ronments" within a zone, but rather to a small archaeol.
ficially bleak environment. Nor do we still series of plant and animal genera whose Highland
picture the Formative peoples as a happy ranges crosscut several environments. complex
group of little brown farmers dancing around Another model badly strained by our new tems wh
their cornfields and thatched huts; we see data is that of culture change during the tran over a pc;
them, rather, as a very complex series of com sition from food-collecting to sedentary agri 8000 B
petitive ethnic groups with internal social culture. Past workers often attributed this to proach VIi
ranking and great preoccupation with status, the "discovery" that planted seeds would and "sea
iconography, water control, and the accumu sprout (MacNeish 1964a: 533), or to the re as a moe
lation of luxury goods. Hopefully, as careful sults of a long series of "experiments" with change.
studies bring these people into sharper focus, plant cultivation. Neither of these explana The fu
they will begin to make more sense in terms tions is wholly satisfying. We know of no regulator
of cotpparable Indian groups surviving in the human group on earth so primitive that they back" pn
ethnographic present. are ignorant of the connection between plants and cow:
Among other things, the new data from and the seeds from which they grow, and this tions ove
Mesoamerica strain some of the theoretical is particularly true of groups dependent (as cybernetb
models we used in the past to view culture and were the highland Mesoamerican food-col processes
culture change. One of these was the model of lectors) on intensive utilization of seasonal systems 11
plant resources. Furthermore, I find it hard ity at big!:
Reproduced by permission of the author and the to believe that "experiments with cultivation" as anyone
Anthropological Society of Washington from were carried on only with those plants that
Betty J. Meggers, editor, Anthropological A rcheol
terns thee
ogy in the Americas (Washington, D.C.: Anthro
eventually became cultivars, since during the English s:
pological Society of Washington, 1968), pp. 67 food-collecting era those plants do not even Let us
87. The bibliographical references have been seem to have been the principal foods used. pattern oJ
placed in the general bibliography for this volume. In fact, they seem to have been less impor- cultivator.
222
 Systems Theory and Early Mesoamerica
Kent V. Flannery
223
tant than many wild plants which never be
came domesticated. Obviously, something
besides "discoveries" and "experiments" is
involved.
I believe that this period of transition from
food-collecting to sedentary agriculture, which
began by 5000 B.C. and ended prior to 1500
B.C., can best be characterized as one of grad
ual change in a series of procurement sys
tems, regulated by two mechanisms called
seasonality and scheduling. I would argue
that none of the changes which took place
during this period arose de novo, but were the
result of expansion or contraction of previ
ar environ ously existing systems. I would argue further
"mesquite that the use of an ecosystem model enables us
Db," "trop to see aspects of this prehistoric culture
~est, first, change which are not superficially apparent.
lit to whole In the course of this essay I will attempt to
I Flannery apply, on a prehistoric time level, the kind of
is article, it ecosystem analysis advocated most recently by
ISic "adap Vayda (1964) and Rappaport (1967), with
llicro-envi modifications imposed by the nature of the
rto a small archaeological data. Man and the Southern
era whose Highlands of Mexico will be viewed as a single
Ids. complex system, composed of many subsys
I'J our new tems which mutually influenced each other
I the tran over a period of over seven millenia, between
lltary agri 8000 B.C. and 200 B.C. This systems ap
II:d this to proach will include the use of both the ::fu.st"
!Ids would and "second" cybernetics (Maruyama 1963)
. to the re as a model for explaining prehist9ric culture
cots" with change.
~ explana The first cybernetics involves the study of
ow of no regulatory mechanisms and "negative feed
: that they back" processe~ whifhprQroQl~ equilibrium,
ICen plants and counteract deviation from stable situa
'. and this tions over long periods of time. 'I:h~ se(;ond
ndent (as cybernetics is the study of "positive feedback"
food-col processes which amplify deviations, causing
t seasonal systems to expand and eventually re1,lch stabil
Id it hard ity at higher levels. Because I am as distressed
dtivation" as anyone but the esoteric terminology of sys
lants that tems theory, I have tried to substitute basic
luring the English synonyms wherever possible.
not even Let us begin by considering the subsistence
ods used. pattern of the food-collectors and "incipient
ss impor cultivators" who occupied the Southern High
lands of Mexico between 8000 and 2000 B.C.
The sources of our data are plant and
animal remains preserved in dry caves in the
Valley of Oaxaca (Flannery, Kirkby, Kirkby,
and Williams, 1967) and the Valley of Tehu
can (MacNeish 1961, 1962, 1964a). Rele
vant sites are Guila Naquitz Cave, Cueva
Blanca, and the Martinez Rock Shelter (near
Mitla, in the Valley of Oaxaca), and Mac
Neish's now-famous Coxcatlan, Purr6n, Abe
jas, 1 Riego, and San Marcos Caves, whose
food remains have been partially reported
(Callen 1965; Smith 1965a). Tens of thou
sands of plants and animal bones were re
covered from these caves, which vary between
900 and 1900 meters in elevation and occur
in environments as diverse as cool-temperate
oak woodland, cactus desert, and semi-tropi
cal thorn forest. Because most of the material
has not been published in detail as yet, my
conclusions must be considered tentative.
Preliminary studies of the food debris from
these caves indicate that certain plant and ani
mal genera were always more important than
others, regardless of local environment. These
plants and animals were the focal points of a
series of procurement systems, each of which
may be considered one component of the total
ecosystem of the food collecting era. They
were heavily utilized - "exploited" is the term
usually employed-but such utilization was
not a one-way system. Man was not simply
extracting energy from his environment, but
participating in it; and his use of each genus
was part of a system which allowed the latter
to survive, even flourish, in spite of heavy
utilization. Many of these patterns have sur
vived to the present day, among Indian
groups like the Paiute and Shoshone (Stew
ard 1955: 101-21) or the Tarahumara of
northern Mexico (Pennington 1963), thus
allowing us to postulate some of the mecha
nisms built into the system, which allowed the
wild genera to survive.
Each procurement system required a tech
nology involving both implements (projec
tiles, fiber shredders, collecting tongs, etc.)
and facilities (baskets, net carrying bags, stor
STUDY OF HUNTER-GATHERERS Systems TIr
Kent V. Flar

224

age pits, roasting pits, etc.). In many cases, The maguey, a tough and phylogenetically major impol'
these implements and facilities were so similar primitive monocotyledon which thrives on is little else a
to those used in the ethnographic present by marginal land even on the slopes of high, during the II
Utoaztecan speakers of western North Amer cold, arid valleys, is unbearably bitter when discovery tho
ica that relatively little difficulty is encoun raw. It cannot be eaten until it has been after sending
tered in reconstructing the outlines of the roasted between 24 and 72 hours, depending its natural fe
ancient procurement system. on the youth and tenderness of the plant harvested we:
Literally hundreds of plant species were involved. ' already, and
used by the food-collectors of the Southern The method of maguey roasting described pollen. Thus
Mexican Highlands. There were annual by Pennington (1963: 129-30) is not unlike on the hillsil
grasses like wild maize (Zea) and fox-tail that of the present-day Zapotec of the Valley spite of the
(Setaria), fruits like the avocado (Persea) of Oaxaca. A circular pit, 3 to 4 feet in diam ceramic fooi
and black zapote (Diospyros) , wild onions eter and of equal depth, is lined with stones weed out the:
(Allium), acorns and pinyon nuts, several and fueled with some slow-burning wood, System 2.
varieties of pigweed (Amaranthus) , and many like oak. When the stones are red-hot, the pit Organ cacti.
other plants, varying considerably from region is lined with maguey leaves which have been at Tehuacam
to region because of rainfall and altitude dif trimmed off the "heart" of the plant. The which appeal
ferences (Callen 1965; Smith 1965b, and maguey hearts are placed in the pit, covered very commOi
personal communication). However, three with grass and maguey leaves and finally a sold under 1
categories of plants seem to have been espe layer of earth, which seals the roasting pit tuna, but the
cially important wherever we have data, re and holds in the heat. After one to five days, fruit of the i
gardless of altitude. They are: depending on the age and quantity of maguey, ubiquitous CI
1) The maguey (Agave spp.), a member the baking is terminated and the hearts are slopes. Mosl
of the Amaryllis family, which is available edible: all, that is, except the indigestible toward the e
year-round; 2) a series of succulent cacti, in fiber, which is expectorated in the form of a on altitude, I
cluding organ cactus (Lemaireocereus spp.) "quid" after the nourishment is gone. Evi be peeled 31
and prickly pear (Opuntia spp.), whose fruits dence of the roasting process can be detected the year.
are seasonal, but whose young leaves are in maguey fragments surviving in dessicated The colle.
available year-round; and 3) a number of human feces from Coxcatlan Cave (Callen place before:
related genera of tree legumes, known locally 1965:342). fruit to mus
as mesquites (Prosopis spp.) and guajes (Lu The Zapotecs of the Valley of Oaxaca, like competition
caena, Mimosa, and Acacia), which bear most Indians of southern Mexico, recognize rodents, wlM
edible pods in the rainy season only. that the best time to cut and roast the maguey The fruits 31
System 1: Maguey Procurement.-Maguey, is after it has sent up its inflorescence, or can caves COl
the "century plant," is most famous today as quiote. The plant begins to die after this have been ".
the genus from which pulque is fermented event, which occurs sometime around the the stem (II
and tequiia and mezcal are distilled. In pre sixth or eighth year of growth, and a natural tion). The l!
historic times, when distillation was unknown, fermentation takes place in the moribund fruits transp
the maguey appears to have been used more plant which softens it and increases its sugar they cannot 1
as a source of food. Perhaps no single plant content. The sending up of this inflorescence the fruit can
element is more common in the dry caves of is a slow process, which can culminate at any nington 196:
southern Mexico than the masticated cud or time of the year. The large numbers of quiote gins to rot. :
"quid" of maguey (Smith 1965a:77). It is not fragments in our Oaxaca cave sites indicate harvest of 0
always realized, however, that these quids that the Indians of the preceramic food-col done quicld)
presuppose a kind of technological break lecting era already knew that this was the best petition fro~
through: at some point, far back in prece point in the plant's life cycle for roasting. age.
ramic times, the Indians learned how to make The discovery that maguey (if properly The harvt
the maguey edible. processed) can be rendered edible was of no matter hi
 Systems Theory and Early Mesoamerica
Kent V. Flannery

225

rlogenetically major importance, for in some regions there appear to diminish the available stands of
:a thrives on is little else available in the way of plant food cactus nor reduce subsequent generations of
IfJes of high, during the heart of the dry season. And the tuna and pitahaya-for the seeds from which
bitter when discovery that maguey was best for roasting the plant is propagated almost inevitably sur
it has been after sending up its inflorescence and starting vive the human digestive tract and escape in
:s, depending its natural fermentation meant that the plants the feces, to sprout that very year. It is even
Df the plant harvested were mostly those that were dying possible that such harvests are beneficial for
already, and had long since sent out their the prickly pear and columnar cacti"in afford
mg described pollen. Thus the maguey continued to thrive ing them maximum seed dispersal. This is
:is not unlike on the hillsides of the southern highlands in only one example of the self-perpetuating
M the Valley spite of the substantial harvests of the pre nature of some of the procurement systems
feet in diam ceramic food-collectors: all they did was to operating in preceramic Mexico.
I with stones weed out the dying plants. System 3: Tree Legume Procurement.
Ihling wood, System 2: Cactus Fruit Procurement. Mesquite is a woody legume which prefers
I-hot, the pit Organ cacti of at least four species were eaten the deep alluvial soil of valley floors and
:It have been at Tehuacan and Oaxaca, and their fruits river flood plains in the highlands. During
1& plant. The which appear late in the dry season-are still the June to August rainy season it bears hun
-pit, covered very common in Mexican markets. Most are dreds of pods which, while still green and
IIDd finally a sold under the generic terms pitahaya and tender, can be chewed, or boiled into a kind
~ roasting pit tuna, but the best known "tuna" is really the of syrup (called "miel" in the Oaxaca and
eta five days, fruit of the prickly pear (Opuntia spp.), the Tehuacan Valleys).
,of maguey, ubiquitous cactus of Mexican plains and rocky Such use of mesquite extended from at
~ hearts are slopes. Most cactus fruit appears some time least the Southern Mexico Highlands (where
t indigestible toward the end of the dry season, depending we found it in caves near Mitla) north to the
lac form of a on altitude, but the tender young leaves may Great American Southwest, where it was evi
I gone. Evi be peeled and cooked during any season of dent at Gypsum Cave and related sites (Har
l be detected the year. rington 1933). Guajes, whose edible pods
ill dessicated The collecting of cactus fruit had to take mature in roughly the same season, character
ave (Callen place before the summer rains turned the ize hill slopes and canyons, and were abundant
fruit to mush, and had to be carried on in in both the Mitla and Tehuacan caves (C.
Oaxaca, like competition with fruit bats, birds, and small Earle Smith, personal communication).
iD, recognize rodents, who also find the fruit appetizing. The amount of food available when mes
,the maguey The fruits are spiny, and some of the Tehua quite and guajes are at the peak of their pod
rescence, or can caves contained wooden sticks which may bearing season is truly impressive. Botanist
Ie after this have been "tongs" for use in picking them off James Schoenwetter, standing outside one of
,ilI'ound the the stem (MacNeish, personal communica our Mitla caves in 1966 during the optimum
ad a natural tion). The spines can be singed off and the mesquite-guaje season, personally communi
c moribund fruits transported by net bag or' basket, but cated to us his suspicion that a family of four
IeS its sugar they cannot be stored for long. By sun-drying, Indians could have collected a week's supply
nflorescence the fruit can be saved for several weeks (Pen of legume pods there "practically without
mate at any nington 1963:117-18), but eventually it be moving their feet."
:IS of quiote gins to rot. It is worth noting, however, that The pod-bearing pattern of mesquite and
ites indicate harvest of most of these wild fruits must be guaje demands a seasonal, localized, and
ic food-col done quickly and intensively because of com fairly intensive period of collecting. The pods
was the best petition from wild animals, rather than spoil can be hand-picked, and probably were trans
laSting. age. ported in the many types of baskets and net
[if properly The harvesting and eating of cactus fruits, carrying bags recovered in the Oaxaca and
ible was of no matter how intensive it may be, does not Tehuacan caves (MacNeish 1964a:533;
STUDY OF HUNTER-GATHERERS Systems Tho
Kent V. Flar

226

Flannery, unpublished data). Both pods and their home ranges, and where ambush hunters deviation fro
seeds can be dried and stored for long periods, can wait for them. In other words, they are discuss only
but they must be picked at the appropriate susceptible to daylight hunts, on foot, by men "scheduling:
time or they will be eaten by animals, like armed with nothing more sophisticated than man by the D
deer, rabbit, and ring-tailed cat. an atlatl or even a fire-hardened spear, such selves; "scm
Mammals were an important year-round as used by the Chiapanecs of the Grijalva which resol",
resource in ancient Mesoamerica, where win Depression (Lowe 1959:7). On top of this, systems.
ters are so mild that many animals never they can stand an annual harvest of 30 to 40 The most
hibernate, as they do at more northern lati percent of the deer population without dimin american ye;
tudes. Deer, peccary, rabbits, raccoons, opos ishing in numbers (Leopold 1959: 513). Ar to May), w.
sums, skunks, ground squirrels, and large chaeological data (Flannery n.d.) suggest (June to SeJ
pocket gophers were common in the pre that the hunters of the Tehuacan and Oa nual rain fa)
historic refuse (Flannery n.d.). However, xaca Valleys did not practice any kind of mals of the II
wherever we have adequate samples of ar conservation, but killed males, females, season, or p
chaeological animal bones from the Southern fawns and even pregnant does (as indicated the semiarid
Highlands of Mexico, it appears that the fol by skeletal remains of late,-term foetuses). like the pod
lowing generalization is valid: white-tailed This does not seem to have depleted local folia), as WI
deer and cottontail rabbits were far and away deer populations in any way. In fact, by thin cacti, bear f:I
the most important game mammals in all pe ning the herds during times of optimum the rains bel
riods, and most hunting technology in the plant resource availability, it may even have that same]
preceramic (and Formative) eras was de prevented the starvation of deer during the (Quercus sp
signed to recover these two genera. Our dis heart of the dry season. carpa sp.) b
cussion of wild animal exploitation will there System 5: Cottontail Procurement. -I have so their sea
fore center on these animals. already discussed the ecology of Mexican winter and I
System 4: White-Tailed Deer Procure cottontails in a previous paper (Flannery These, diffen
ment. - The white-tailed deer, a major food 1966) and will only recapitulate briefly here: to the plant
resource in ancient times, continues to be cottontails are available year round (though and seed itse
Mesoamerica's most important single game most abundant in the rainy season) and can is most am.
species. Part of its success is due to the wide best be taken by means of traps or snares. mined the c
range of plant foods it finds acceptable, and Throwing sticks are also effective, and the cultural ban
its persistence even in the immediate vicinity Indians of northern Mexico use a figure-four Indians had.
of human settlement and under extreme hunt rock trap or "deadfall" (Pennington 1963: 90 week or two
ing pressure. White-tailed deer occur in every and Plate 12)., In the Tehuacan caves there would take)
habitat in Mesoamerica, but their highest were fragments of whittled sticks and fiber harvest furi.
populations are in the pine-oak woodlands of loops or slip knots which may'be trap frag by birds, rOIl
the Sierra, Madre. The tropical rain forests, ments (Mac~eish 1964a:533 and personal MacNeisi
such as those of the lowland Maya area, are communication); similar fragments showed some of the:
the least suitable habitats for this deer. Within up in one of our Oaxaca caves in 1966. The acted to sea
Mesoamerica proper, highest prehistoric pop best feature of cottontail trapping is that the in areas wi
ulations would have been in areas like the only investment of labor is in the manufacture were availa::
mountain woodlands of the Valley of Mexico,. and setting of the trap; it works for you while large group
Puebla, Toluca, Oaxaca, and Guerrero. you go about other tasks. And cottontails are bands," pro
These deer have relatively small home such prolific breeders that no amount of lated famili
ranges, and although they often spend part trapping is likely to wipe them out. During the:
of the daylight hours hiding in thickets, they The ecosystem in which the hunters and edible plan:
can be hunted in the morning and evening collectors of ancient Mexico participated in mented intc
when they come out to forage. Deer have cluded many regulatory mechanisms, which been indivic
known trails along which they travel within kept the system successful, yet counteracted scattered ou
 Systems Theory and Early Mesoamerica
Kent V. Flannery

227

lush hunters deviation from the established pattern. I will ing resources too meager to support a macro
lis, they are discuss only two of these- "seasonality" and band.
ioot, by men "scheduling." "Seasonality" was imposed on The seasonally restricted nature of re
licated than man by the nature of the wild resources them sources made it impossible for groups to re
spear, such selves; "scheduling" was a cultural activity main large all year, and effectively counter
lie Grijalva which resolved conflict between procurement acted any trends toward population increase
top of this, systems. which might have been fostered by the in
oof 30 to 40 The most important divisions of the Meso tensive harvests of the rainy-season macro
Iaout dimin american year are a winter season (October bands. Thus populations never grew to the
~:513). Ar a
to May), which is dry, and summer season point where they could effectively over-reach
1-) suggest (June to September), when most of the an their wild food resources. MacNeish (l964a:
.. and Oa nual rain falls. Many edible plants and ani Fig: 4) postulates that as late as 3000 B.C.
,., kind of mals of the area are available only during one the population of the Tehuacan Valley was
, females, season, or part of a season. For example, in no higher than 120-240 persons, in an area of
ill indicated the semiarid highlands of Mexico some plants 1400 square miles.
~ foetuses) . like the pochote or kapok tree (Ceiba parvi So many possibilities for exploitive activity
~F3 local folia), as well as many species of columnar were open to these. ancient Mesoamericans
by. thin cacti, bear fruit in the late winter just before that it would have been impossible to engage
o optImum the rains begin, so that their seeds will sprout in all of them, even seasonally. It happens
leven have that same year. Other trees, like the oak that there are times of the year when a num
liming the (Quercus spp.) and the chupandilla (Cyrto ber of resources are available simultaneously,
i carpa sp.) bear fruit after the summer season, producing a situation in which there is some
-r-I have so their seeds will lie dormant through the conflict for the time and labor of. the group.
. Mexican winter and sprout during the following year. Division of labor along the lines of sex, with
kHannery These differences, which are of adaptive value men hunting and women collecting, is one
(idly here: to the plant "(allowing each species to flower common solution to these conflicts, but not
.. (though and seed itself during the time of year when it all conflicts are so easily resolved.
t and can is most advantageous), somewhat predeter The solution for more complex situations
snares. mined the collecting schedule of the preagri may be called "scheduling," and it involves a
L and the cultural bands in Mesoamerica: often these decision as to the relative merits of two or
~e-four Indians had to be able to predict to within a more courses of action. Such "scheduling
p1963:90 week or two when the maturation of the,plant decisions" are made constantly by all human
iilves there would take place, and then they would have to groups on all levels of complexity, often with
;-and fiber harvest furiously before the plants were eaten out any awareness that a decision is being
trap frag by birds, rodents, or other small mammals. made.
, personal MacNeish (l964a, 1964b) has shown It is not necessarily true that the lower the
Is showed some of the ways in which human groups re level of social complexity, the fewer the con
1966. The acted to seasonality. During the rainy season, flict decisions, for hunting and gathering
:II that the in areas where many wild plant resources groups of arid America had many scheduling
mufacture were available, they often came together in problems to resolve. Food gathering bands of
you while large groups which MacNeish calls "macro the Great Basin, for example, often depended
IIltails are bands," probably consisting of a series of re on "scouting reports" from relatives who had
!DOunt of lated families (d. Steward 1955: 101-21). passed through certain areas several weeks in
During the heart of the dry season, when few advance. If they noticed an unusually high
Dters and edible plants are available, the group frag concentration of antelope or rabbit in a par
ipated in mented into "micro-bands," which may have ticular valley, or if they saw that a particular
IS, which been individual family units. These small units stand of wild fruit would come ripe within the
nteracted scattered out widely over the landscape, utiliz next two weeks, they would advise other
STUDY OF HUNTER-GATHERERS Systems TII(\j
Kent V. Fl.
1

228

scattered bands of foragers about this resource Assuming that each occupation floor in a tailed. I woI
(Steward 1955:105-6). Often, while they given cave represents the debris of a single follows: ~
descended on the area to harvest that particu encampment, usually dating to a single sea A. In the'S
lar species, new reports would come in from son (an assumption that seems to be borne season, theni
other areas concerning still another resource. out by the quantity and nature of the refuse), of wild plaud
This was not the kind of "hit and miss" pat the combinations of plant and animal remains were intemiil
tern of exploitation one might think, for the observed in a given level tell us something cached as tIIII.
Great Basin Indians had a rough idea that about prehistoric scheduling decisions. Analy to have beeIIt
acorns and pinyon nuts would be available in ses of our Oaxaca caves and MacNeish's Te "all hands'"
the autumn, wild legumes and grasses in the huacan Caves, by roughly the same group of little deer
rainy season, and so on. The outlines of a specialists (MacNeish 1962, 1964a; Flan dians set
schedule, albeit with conflicts, were present; nery n.d.), suggest the following tentative collecting
the "scouting reports" helped resolve con generalizations: conflict wi
flicts and gave precision to the dates of each 1) Dry season camps (October-March), deer hun .
kind of resource exploitation, depending on depending' on their elevation above sea level, B. In
individual variations in growing season from may have great caches of fall and winter have cea
year to year. plants - for example, acorns in the Mitla area, at its best.
These individual variations, which are a or Ceiba pods in the Coxcatlan area-but in male deer
common feature of arid environments, com general they lack the variety seen in rainy selves) faR .
bined with the scheduling pattern to make it season levels. And perhaps most significantly, ing the a
unlikely that specialization in anyone re they have a high percentage of those plants also the
source would develop. This prevented over which, although not particularly tasty, are of the hi
utilization of key plants or animals, and main~ available year-round: maguey, prickly pear be more
tained a more even balance between varied leaf, Ceiba root, and so on. These are the so dry season
resources. Because scheduling is an oppor called starvation plants, which can be eaten warier and
tunistic mechanism, it promoted survival in in the heart of the dry season when little else the moun
spite of annual variation, but at the same time is available. These same levels also tend to year i n .
it supported the status quo: unspecialized have high percentages of deer bone. Some, in evidently mt
utilization of a whole range of plants and fact, have little refuse beyond maguey quids heavily to
animals whose availability is erratic over the and white-tailed deer. the root of~
long run. In this sense, scheduling acted to 2) Rainy season camps (May-September), like sweet ail
counteract deviations which might have re as might be expected, show great quantities uey plant {~
sulted in either 1) starvation, or 2) a more of the plants available at that time of the year: pear to have:
effective adaptation. mesquite, guajes, amaranth, wild avocado, C. By cIii
Thanks to the plants and animal bones pre zapotes, and so on. They also tend to be rich the precerui
served in the dry caves of Oaxaca and Te in small fauna like cottontail, opposum, the late spiI
huacan, 'we can often tell which season a skunk, raccoon, gopher, and black iguana. point he COl
given occupation floor was laid down in. Be Although deer are often present in these Essentially, :
cause of the work of botanists like Earle camps, they frequently represent only a small seasonal aVo
Smith, Lawrence Kaplan, and James Schoen percentage of the minimum individual ani which cfiml
wetter, we know the season during which mals in the debris. Nor are the starvation which were
each plant is available, and hence when its plants particularly plentiful in these rainy ping. He SIC
harvest must have taken place. Even the use season levels. hunting for
of animal resources can often be dated sea 3) What these generalizations suggest, for vests were.
sonally; for example, in the Tehuacan Valley, the most part, is that scheduling gave prefer D.CIiIDI
we studied the seasonality of deer hunting by ence to the seasonality of the plant species rainy se-.
the condition of the antlers, which indicates collected; and when conflict situations arose,
the time of year when the animal was killed. it was the animal ~xp~oitation that was cur- herds at "
picture .....
 Systems Theory and Early Mesoamerica
Kent V. Flannery

229
on floor in a tailed. I would reconstruct the pattern as it, but this cannot be detected in the archaeo
II of a single follows: logical record. The constant evolution of
~ single sea A. In the late dry season and early rainy new bags, nets, baskets, projectile points,
to be borne season, there is a period of peak abundance scrapers, carrying loops, and other artifacts
f the refuse), of wild plant foods. These localized resources from the caves of the Southern Highlands
Eal remains were intensively harvested, and eaten or suggests slow but continual innovation. To
~ something cached as they came to maturity; this appears what extent these innovations increased the
lions. Analy to have been a "macroband" activity. Because productivity of the system is not cleat.
ICNeish's Te "all hands" participated in these harvests, Because the major adaptation was to a
IDle group of little deer hunting was done; instead the In series of wild genera which crosscut several
'964a; Flan dians set traps in the vicinity of the plant environmental boundaries, the geographic ex
ing tentative collecting camp, an activity which does not tent of the ecosystem described above was
conflict with intensive plant harvests the way very great. This adaptation is clearly reflected
bee-March), deer hunting would. in the technological sphere. Implements and
I'Ve sea level, B. In the late fall and winter, most plants facilities of striking similarity can be found in
and winter have ceased to bear fruit, but deer hunting is regions which differ significantly in altitude
Ie Mitla area, at its best. Since this is the mating season, and rainfall, so long as the five basic cate
area-but in male deer (who normally forage by them gories of plants and animals are present. This
;=en in rainy selves) fall in with the does and fawns, mak can be illustrated by an examination of the
llignificantly, ing the average herd larger; and since this is Coxcathin Phase (5000-3000 B.C.) as it is
those plants also the season when the deciduous vegetation represented at Coxcathin Cave, Puebla (Mac
, tasty, are of the highlands sheds its leaves', the deer can Neish 1962) and at Cueva Blanca, Oaxaca
prickly pear be more easily followed by hunters. As the (Flannery, -Kirkby, Kirkby, and Williams
t are the so dry season wears on, however, the deer grow 1967) . . '
an be eaten warier and range farther and farther back into Coxcatlan Cave, type site for the phase,
en little else the mountains. This is the leanest time of the occurs at 975 meters in an arid tropical forest
dso tend to year in terms of plant resources, and it was characterized by dense stands of columnar
lie. Some, in
evidently in this season that man turned most cacti; kapok trees (Ceiba parvifolia); chu
agoey quids heavily to plants available year round, like pandilla (Cyrtocarpa sp.); cozahuico (Sider
the root of the Ceiba (which can be baked oxylon sp.) ; and abundant Leguminosae,
;Cptember) , like sweet manioc) or the heart of the mag Burseraceae, and Anacardiaceae (Smith
It quantities uey plant (which can be roasted). These ap 1965b: Fig. 31). Cueva Blanca occurs at
the year: pear to have been "microband" activities. 1900 meters in a temperate woodland zone
d avocado, C. By chewing roots and maguey hearts, with scattered oaks; Dodonaea; ocotillo
d to be rich the preceramic forager managed to last until (Fouquieria); wild zapote (Diospyros); and
opposum, the late spring growing seasons, at which other trees which (judging by archeological
ICk iguana. point he could wallow in cactus fruit again. remains) may originally have included hack
It: in these berry (Celtis) and pinyon pine.
Essentially, his "schedule" was keyed to the
lilly a small seasonal availability of certain wild plants, In spite of environmental differences, im
!Vidual ani which climaxed at those times of the year plements at the two sites are nearly identical;
: starvation which were best suited for small game trap even the seasonal deer hunting pattern and the
hese rainy ping. He scheduled his most intensive deer size of the encamped group are the same. In
hunting for the seasons when big plant har the past, such identity would have inspired
mggest, for vests were not a conflicting factor. the traditional explanation: "a_similar adap
:ave prefer D. Climatic fluctuations, delays in the tation to a similar arid environment." But as
fl1Jt species
rainy season, or periodic increases in the deer seen above, the two environments are not that
lions arose, herds at given localities probably kept the similar. The important point is that the basic
It was cur
picture more complex than we have painted adaptation was not to a zone or even a bio-'
STUDY OF HUNTER-GATHERERS
230
tope within a zone, but to five critical cate
gories-white-tail deer, cottontail, maguey,
tree legumes, prickly pear and organ cactus.
These genera range through many zones, as
did the Indians who hunted them, ate them,
propagated their seeds, and weeded out their
dying members. This is not to say that bio
topes were unimportant; they played a role,
but they were also crosscut by a very impor
tant system.
Seasonality and scheduling, as examined
here, were part of a "deviation-counteracting"
feedback system. They prevented intensifica
tion of anyone procurement system to the
point where the wild genus was threatened; at
the same time, they maintained a sufficiently
high level of procurement efficiency so there
was little pressure for change. Under the
ecosystem operating in the Southern Mexican
Highlands during the later part of the food
collecting era, there was little likelihood that
man would exhaust his own food resources or
that his population would grow beyond what
the wild vegetation and fauna would support.
Maintaining such near-equilibrium conditions
is the purpose of deviation-counteracting
processes.
Under conditions of fully achieved and
permanently maintained equilibrium, prehis
toric cultures might never have changed. That
they did change was due at least in part to the
existence of positive feedback or "deviation
amplifying" processes. These Maruyama
(1963: 164) describes as "all processes of
mutual causal relationships that amplify an
insignificant or accidental initial kick, build
up deyiation and diverge from the initial con
dition."
Such "insignificant or accidental initial
kicks" were a series of genetic changes which
took place in one or two species of Meso
american plants which were of use to man.
The exploitation of these plants had been a
relatively minor procurement system com
pared with that of maguey, cactus fruits, deer,
or tree legumes, but positive feedback follow
ing these initial genetic changes caused one
minor system to grow all out of proportion to
the others, and eventually to change the whole
ecosystem of the Southern Mexican High
lands. Let us now examine that system.
System 6: Wild Grass Procurement.-One
common activity of the food-collecting era in
the Southern Highlands was the harvesting
of annual grasses. Perhaps the, most useful in
preagricultural times was foxtail grass
(Setaria) (Callen 1965:343), followed by
minor grasses like wild maize (Zea mays),
whi<1h may have been adapted to moist bar
rancas within the arid highland zone (Smith
1965a:95).
We know very little about the nature of
the early "experiments" with plant cultiva
tion, but they probably began simply as an
effort to increase the area over which useful sine.
plants would grow. For example, Smith beans in
(1965a:77-78) has suggested that the pre apart
ceramic food-collectors may have attempted either
to increase the density of prickly pear and breaktlbm.
organ cactus stands by planting cuttings of Startillld
these plants. For the most part, judging by tially)
the archaeological record, these efforts led to positive
little increase in food supply and no change which e
in emphasis on one genus or another, until most p
sometime between 5000 and 2000 B.C.-a Meso
series of genetic changes took place in a few cuItiva'
key genera. It was these genetic changes, act ablecr
ing as a "kick," which allowed a deviation able ge
amplifying system to begin. the greatet
As implied by Maruyama, many of these tion, and'"
initial deviations may have been accidental There cant
and relatively minor. For example, beans more intealj
1) became more permeable in water, making in perfectiDtl
it easier to render them edible; and 2) de well irrigaai
veloped limp pods which do not shatter when 1967; F'1aaI!
ripe, thus enabling the Indians to harvest them Iiams 1967)
more successfully (Kaplan 1965). Equally therefore, ..
helpful were the changes in maize, whose the Spanish 4
genetic plasticity has fascinated botanists for Whattbisc
years. While Setaria and the other grasses re Wild Grassl
mained unchanged, maize underwent a series expense of, I
of alterations which made it increasingly procuremeol
more profitable to harvest (and plant over Moreover, d
wider areas) than any other plant. Its cob by necessibl
increased in size; and, carried around the spring) as w
highlands by Indians intent on increasing its (early fall).
 Systems Theory and Early Mesoamerica
Kent V. Flannery

231

i1nge the whole range, it met and crossed with its nearest rela the spring-ripening wild plants (prickly pear,
lexican High tive, Zea' tripsacum, to produce a hybrid organ cactus) and the fall-ripening crops
'SyStem. named teocentli. From here on its back (acorns, fruits, some guajes). It competed
trement. - One crosses and subsequent evolution, loss of with rainy-season hunting of deer and pec
[lecting era in glumes, increase in cob number and kernel cary. And it was a nicely self-perpetuating
k harvesting row number, have been well documented by system, for the evolution of cultivated maize
llast useful in MacNeish, Mangelsdorf, and Galinat (1964). indicates that no matter how much the Indi
Ioxtail grass Another important process, though some ans harvested, they saved the best seed for
i followed by what less publicized, was the interaction be next year's planting; and they saved it under
~Zea mays), tween corn and beans recently emphasized by storage conditions which furthered the sur
!lfo moist bar Kaplan ( 1965) . Maize alone, although a vival of every seed. Moreover, they greatly
tzone (Smith reasonably good starch source, does not in increased the area in which maize would grow
~ itself constitute a major protein because it by removing competing plants.
ibe nature of lacks an important amino acid-lysine As mentioned earlier, 1) procurement of
pant cultiva which must therefore be made up from an
other source. Beans happen to be rich in ly
"starvation" plants like Ceiba and maguey
seems to have been undertaken by small,
llililply as an
Iwhich useful sine. Thus the mere combining of maize and scattered "microbands," while 2) harvests of
"pIe, Smith beans in the diet of the Southern Highlands, seasonally-limited plants, abundant only for
tbt the pre apart from any favorable genetic changes in a short time-like cactus fruits, mesquite and
fie attempted either plant, was a significant nutritional guajes, and so on-seem to have been under
~ pear and breakthrough. taken by large "macrobands," formed by the
Starting with what may have been (ini coalescence of several related microbands.
t cuttings of
... judging by
~orts led to
tially) accidental deviations in the system, a
positive feedback network was established
Because of this functional association between
band size and resource, human demography
td no change which eventually made maize cultivation the was changed by the positive feedback of early
maize-bean cultivation: an amplification of
ildter, until most profitable single subsistence activity in
iooo B.C.-a Mesoamerica. The more widespread maize the rainy-season planting and harvesting also
a.ce in a few cultivation, the more opportunities for favor
able crosses and back-crosses; the more favor
meant an amplification of the time of macro
band coalescence. MacNeish (l964b: 425)
~anges, act
ra deviation able genetic changes, the greater the yield;
the greater the yield, the higher the popUla
anticipated this when he asked:
t
tion, and hence the more intensive cultivation. Is it not possible as the number of new agri
y of these . doubt that pressures 'for
cultural plants increased that the length of time
in accidental There can be lIttle that the microbands stayed in a single planting
..pIe, beans more intensive cultivation were instrumental area also increased? In time could not perhaps
later, making in perfecting early water control systems, like one or more microbands have been able to stay
well irrigation and canal irrigation (Neely at such a spot the year around? Then with
i and 2) de further agricultural production is it not possible
.atter when 1967; Flannery, Kirkby, Kirkby, and Wil
that the total macroband became sedentary?
cllarvest them liams 1967). This positive feedback system, Such would, of course, be a village.
55). Equally therefore, was still increasing at the time of
Daize, whose the Spanish Conquest. Actually, it may not be strictly accurate to
h>tanists for What this meant initially was that System 6, say that sedentary village life was "allowed"
1:1' grasses re Wild Grass Procurement, grew steadily at the or "made possible" by agricultural produc
lrent a series expense of, and in competition with, all other tion; in fact, increased permanence of the
increasingly procurement systems in the arid highlands. macroband may have been required by the
:I plant over Moreover, the system increased in complexity amplified planting and harvesting pattern.
lant. Its cob by necessitating a planting period (in the An aspect of early village agriculture in
around the spring) as well as the usual harvesting season Mexico not usually dealt with in the literature
ncreasing its (early fall). It therefore competed with both is the extent to which increased concentration
 STUDY OF HUNTER-GATHERERS Systems
KentV.F

232

on maize production made it necessary to re from region to region, depending on whether cultivated
schedule other procurement systems. It is not agriculture could be practiced year-round, o~ maize crOJ
possible in an essay of this length to discuss only seasonally. In the Valleys of Mexico and region of ti
all the subtleties of Formative agricultural Tehuacan, remains of white-tailed deer are is as impc
systems. The basic distinction I would like to abundant in Formative sites (Vaillant 1930, years it is
make is this: given the technology of the 1935b; Flannery n.d.), but wherever we have (AubreyY
Early Formative as we understand it at pres accurate counts on these fragments it is clear cation).
ent, there were regions where maize could be that by far the most intensive deer hunting System
grown only during the rainy season, and re was done in the late fall and winter. Projectile Fowl.-UD
gions where maize could be grown year points and obsidian scrapers of many types Mesoameri
round. All differences in scheduling to be are plentiful in these sites (MacNeish 1962, since we sti
considered in this paper ultimately rest on this Vaillant 1930). On the Guatemalan coast, at dence froo::
dichotomy. -. Panuco, or in the western Valley of Oaxaca, cultivation'
Regions where we postulate that agricul deer remains are absent or rare, and projectile marsh area
ture was practiced only during the rainy sea points nonexistent (Coe and Flannery 1967; beginning'
son include areas with an extremely arid cli~ MacNeish 1954). It has occasionally been have data I
il
mate like the Tehuacan Valley, or higher suggested that the lowland areas had such in lakes of hi
valleys where frosts occur in October and tensive agriculture that hunting was "unneces and lagoon:
continue until April, as is the case in the sary," whereas the highland areas needed deer Water Ii
Valley of Mexico (Sanders 1965: 23 ). Re "as a supplement to their diet." I do not be stricted in
gions where we postulate that agriculture was lieve this is the case; it is more likely a matter resources D
practiced year-round include very humid of scheduling. It so happens that t~ best sea breed in ]II
areas in the frost-free coastal lowlands (such son for deer hunting in the oak woodlands of some twod
as the southern Gulf Coast or the Pacific highland Mesoamerica is late fall, after the geese) spc
coasts of Chiapas and Guatemala), and areas maize crop has been harvested and the frosts marshes 01
in the frost-free parts of the interior where are beginning. This made intensive fall and Alberta, Sa
one of two techniques was possible: 1) in winter deer hunts a logical activity. By con the format
tensive cropping of permanently humid river trast, lowland peoples concentrated on those these duc.lc
bottomlands, such as in the Central Depres wild resources that were availtble year-round series of fe
sion of Chiapas (Sanders 1961:2) or 2) very in the vicinity of the village. Exploitation of only two w
primitive water control techniques like "pot these resources could be scheduled so as not and Centra
irrigation," such as in the western Valley of to siphon off manpower from agricultural ac the Central
Oaxaca (Flannery, Kirkby, Kirkby, and Wil tivities. On the Guatemalan Coast, for exam of the Ceo
liams 1967). ple, the very rich perennial fish resources of Patzcuaro,
What did this mean, region by region, in the lagoon and estuary system were relied pintail (All
terms o( "scheduling"? It meant that, in re upon. Some villages, located near mangrove clypeata), ;
gions of year-round agriculture, certain sea forests, collected land crabs; others, located carolinensis.
sonal activities were curtailed or even aban at some distance from the ma9groves, ignored is resident
doned, and emphasis was placed on those them (Coe 1961b; Coe and Flannery 1967). constitutes
year-round resources that did not conflict with None of these resources cqnflicted with the Ducks corDi
farming schedules. In regions where farming farming pattern. ! the extensi.
was conducted only in the rainy season, the Similar rescheduling of ;\vild plant collect Chiapas aIh
dry season was left open for intensive seasonal ing took place in the highlands. The plants most numeJ
collecting activities. Even exploitation of per that dwindled in impor~ance were the ones the pintail
manent wild resources might be deferred to that ripened during.the seasons when corn (Anas cyar.
that time of year. Let me give a few exam would have to be planted or harvested. Plants americana).
ples: like maguey, whose exploitation could be species likl
The Rescheduling of Deer Hunting.-Deer deferred until the winter, were still exploited (Dendrocyg,
hunting in the Formative differed greatly intensively, and in fact eventually came to be stitute less d
 Systems Theory and Early Mesoamerica
Kent V. Flannery

233

ag on whether cultivated widely in areas where a winter other words, between 97 and 99 percent of
lCaI'-round, or maize crop is impossible. In the arid Mitla the duck population of Mesoamerica is avail
rf Mexico and region of the Valley of Oaxaca today, maguey able only between November and March; by
iIed deer are is as important a crop as maize, and some March or April most of these species are
'aillant 1930, years it is the only crop that does not fail either back in Canada or on their way. This
rever we have (Aubrey W. Williams, Jr., personal communi necessitated an intense seasonal exploitation
ails it is clear cation) . pattern similar to that required by perishable
deer hunting System 7: Procurement of Wild Water seasonal fruits.
fer. Projectile Fowl.-Until now, we have not mentioned It is difficult to compare the relative abun
f many types Mesoamerica's great water fowl resources, dance of waterfowl on the Pacific coast la
tNeish 1962, since we still have no good archaeological evi goons with Lake Texcoco, because the lake
lilian coast, at dence from the food-collecting and "incipient system of the Valley of Mexico was drained
, of Oaxaca, cultivation" periods in any of the lake and by the Spanish, and is now a pale shadow of
Ibd projectile marsh areas where those fowl congregate. But what it was in the Formative. In 1952, an
.....ery 1967; beginning with the Formative period, we do estimated 33,540 migratory ducks spent the
lJonally been have data on wild fowl exploitation from the winter in Lake Texcoco (Leopold 1959: Ta
~ such in lakes of highland Mexico and the swamps ble 4), while the totals for the Chiapas Coast
(lis "unneces and ~ons of the coast. during the same period were over 300,000;
l-eeded deer Water fowl in Mesoamerica are as re some 27,000 of these were in the area be
lk do not be stricted in availability as the seasonal plant tween Pijijiapan and the Guatemalan Coast
~y a matter resources mentioned above. Only four species alone, a stretch of only 100 miles of coastline.
lite best sea breed in Mexico. All the others (perhaps The Early Formative villagers responded
ioo<nands of some two dozen species or more of ducks and quite differently to these populations of win
III. after the geese) spend the summer in the prairie ter waterfowl. Every Formative site report
ltd the frosts marshes of western Canada, principally in from the Lake Texcoco area stresses the abun
lYe fall and Alberta, Saskatchewan and Manitoba. Before dance of duck bones in the refuse. Vaillant
Ily. By con the formation of winter ice in November, (1930: 38) claimed that the animal bones
lid on those these ducks and geese head south down a from Zacatenco indicated "considerable con
t year-round series of four well defined routes, of which sumption of the flesh of birds and deer," and
ilJoitation of only two will be considered here: the Pacific his illustrations of bone tools suggest that
lid so as not and Central Flyways. Ducks coming down bones of waterfowl were well represented.
Icu1tural ac the Central Flyway, terminating at the lakes Worked bird bone also appears at El Arbolillo
I, for exam of the Central Mexican Plateau (Texcoco, (Vaillant 1935b:246-47). Pifia Chan (1958:
1CS0urces of Patzcuaro, Cuitzeo, Chapala), include the ,l7) likewise lists "bones of deer and aquatic
"Were relied pintail (Anas acuta), the shoveler (Spatula birds" from Tlatilco. At Ticoman, bird bones
.. mangrove clypeata) , and the green-winged teal (Anas were also common, and the larger ones ap
IeI'S, located carolinensis). The coot (Fulica americana) parently were ducks (Vaillant 1931). Re
1IJes, ignored is resident year round in Lake Texcoco, but cently, I have had a chance to examine faunal
Iery 1967). constitutes only 3 percent of the water fowl. remains from Tolstoy's new excavations at
=d with the Ducks coming down the Pacific Flyway reach El Arbolillo, Tlatilco, and Tlapacoya, as well
the extensive lagoon-estuary system of the as the Late Formative site of Temesco near
IUlt collect Chiapas and Guatemala coasts. Among the Lake Texcoco (Dixon 1966), and ducks of
The plants most numerous ducks taking this route are the genera Anas and Spatula are abundant,
~ the ones the pintail (Anas acuta) , blue-winged teal confirming Vaillant's impre!,sions.
when corn (Anas cyanoptera) , and baldpate (Mareca On the Guatemalan Pacific Coast, as sug
!ted. Plants americana). There are also a few resident gested by Coe and Flannery (1967) the ex
l could be species like the black-bellied tree duck tensive duck populations were virtually ig
D exploited (Dendrocygna autumnalis) , but they con nored. Although rich in fish and mollusks,
came to be stitute less than 1 percent of the waterfowl. In the Formative middens have yielded not a
STUDY OF HUNTER-GATHERERS

234 PART
single bone of the ducks that flew over our periments," or "genius," but instead enables Archaeola
heads as we traveled upriver to the site each us to treat prehistoric cultures as systems. It
day. Since other birds, like the brown pelican, stimulates inquiry into the mechanisms that
the Study
were sometimes killed and eaten, we assume counteract change or amplify it, which ulti
that ducks must occasionally have been con mately tells us something about the nature of
sumed. But the paucity of their remains is in adaptation. Most importantly, it allows us to
striking contrast to the Lake Texcoco sites. vi~w change not as something arising de novo,
I suggest that in areas where agriculture but in terms of quite minor deviations in one
was practiced year-round, heavy exploitation small part of a previously existing system, Introductit
of winter duck resources would have con which, once set in motion, can expand greatly
flicted with farming, and hence was not because of positive feedback. The inventiOi
practiced. In areas like the Valley of Mexico, The implications of this approach for the archaeology.
where winter frosts prevent agriculture, ducks prehistorian are clear: it is vain to hope for silnultaneous
arrive during the very time of the year when the discovery of the first domestic corn cob, pology since
farming activity was at its lowest ebb, and the first pottery vessel, the first hieroglyphic, domesticatiol
hence they could be heavily exploited. This or the first site where some other major break
was probablJ
may be one further example of the kind of through occurred. Such deviations from the
"scheduling" that characterized the Forma preexisting pattern almost certainly took place outlined the I
tive. in such a minor and accidental way that their of urbanism I
The use of a cybernetics model to explain traces are not recoverable. More worthwhile Near East dIo
prehistoric cultural change, while terminolog would be an investigation of the mutual causal forced into II
ically cumbersome, has certain advantages. processes that amplify these tiny deviations remaining pi;
For one thing, it does not attribute cultural into major changes in prehistoric culture.
resulted in ni
evolution to "discoveries," "inventions," "ex
Braidwood"

Pleistocene b
theatre for ell
here as is C1II
of domestical
cities. Many
outlined.
It is in the
demonstratec
ment and de!
for us an unI
use of domes
techniques fJ
strengths an<
aspects of eJi
In additio!
advent of dO!
become cone:
agriculture a
social organi
and animals