PS-- Fuchs: I would consider it a great favor if you could summarize your findings and your methodology in one succinct post, using the sources you have already showed us, along with your conclusions.
Don't be succinct... my opinion of course.
I fear it will be "not so succinct ...".
Preface
------------------------------------------------------------------------
I would once again like to emphasize that almost nothing of the stuff on Indian population data I will try to summarize here is my own original research.
I will rely almost exclusively on the data presented in the two sources:
Kingsley M Bray, “Teton Sioux: Population History, 1655-1881,” Nebraska History 75 (1994): 165-188.
John S. Gray: "Centennial Campaign: The Sioux War of 1876".
Bray and Gray trawled a boatload of primary sources and filtered out about a dozen useful estimates for the Lakota population between 1800 and 1870.
In addition, they have analyzed the counts and "census" data taken on the Sioux reservations and from surrenders to the military between 1870 and 1890 and condensed those into usable numbers.
What I noted when working through those works, and the few primary sources I had access to, is the central importance of the ratio of people per Indian lodge for understanding the population data.
In virtually all pre-reservation estimates lodge numbers were enumerated, not numbers of people. In many cases this continued long into the reservation period.
(Gray is extremely annoying by using "numbers of people" and "numbers of lodges" interchangeably, using his own preferred conversion ratio of 7 (8 for the Cheyenne) regardless of circumstances. Usually without noting what was actually given in the primary source.)
To derive the population number, you had to multiply the people per lodge ratio with the number of lodges. Both numbers were known only roughly, especially the ratio of people per lodge.
Multiply two numbers which are estimates yields results with considerable uncertainty.
The one reliable method to overcome this uncertainty is to assemble as many independent estimates as possible, that are founded on actual data, and are about as likely to be too low as to be too high.
(An Excel file with a compilation of those data, as well as comments regarding my adjustments and calculations is attached to the post)
------------------------------------------------------------------------
Numbers
------------------------------------------------------------------------
First of all, it's critically important to define who is referred to with those populating numbers:
The Western Sioux, or Plains Sioux or Teton Sioux or
Lakota, expressions that are usually used synonymously, specifically their tribal divisions
Brulé/Sichangu
Oglala
Hunkpapa
Miniconjou
Blackfeet/Sihasapa
Sans Arc
Two Kettle
roughly in descending order of population.
NOT included are the Eastern/Central Sioux or
Dakota, Santee, Yankton and Yanktonnai Dakota and the Cheyenne.
Those will be handled summarily at the end of the section.
------------------
historic estimates
------------------
For the Lakota, there are a total of 10 different primary accounts referenced in Bray and Gray, giving estimates for number of lodges and/or the total population through the years 1833 to 1869.
For all of those, numbers for the seven separate tribal divisions are estimated, and than added up to a total number.
After minimal adjustments to fill in a few missing numbers, the averages of those 10 accounts are (rounded numbers):
Brulé.............3500
Oglala............2750
Hunkpapa.......2250
Miniconjou......2250
Sihasapa........1600
Sans Arc........1400
Two Kettle......1000
---------------------
Sum..............
14750 ( standard deviation +/-2400)
While the numbers for the tribal divisions are fluctuating considerably between estimates, the total are relatively consistent. This follows naturally from two facts:
- those tribal division are rather loosely defined, and the "allegiance" of specific bands to specific divisions was not static.
- the bottom-up accounting tend to average out even relatively large individual mistakes made in those estimates, as long as there isn't an underlying bias.
(Notably "out of whack" is the Culbertson estimate from 1850, which has by far the most outlying numbers.)
Bray tried to condense those estimates (and a few more even older ones) into a consistent image of Lakota population changes. He made some significant adjustments to the original numbers, but all of those are properly noted in the paper. From his reconstruction, a number of
15200would follow for 1876.
I tried to do something similar to the Bray method, but making a slightly different set of adjustments for obvious inconsistencies in the raw data, as well as using slightly lower estimates for pre-1860 people per lodge ratios. My most plausible extrapolation for the 1876 number would be
15400with a considerably reduce spread of the single data points compared to the raw data. Still, I might have introduced unknown errors with my judgment calls, and slight changes for the people per lodge ratio will change the result considerably.
To summarize the analysis of the estimates from historic accounts, they point to a most likely value for the Lakota population in 1876 of about 15000, and a near-certainty of being between 10 and 20000. ----------------
military counts
----------------
A fairly straightforward method of getting a handle on the number of Lakota number in 1876/77 is to add up the results from military counts on the reservations in 1876/77 and of the surrendering Indians after the LBH battle.
Those counts were much more stringent than the usual "census" procedure employed by the civilian Indian agents, and we should expect fairly accurate results from those.
Number of Lakota counted on the reservations in 1876 (from Gray):
Brulé.............4387
Oglala............2336
Hunkpapa........333
Miniconjou........346
Sihasapa..........645
Sans Arc..........170
Two Kettle........840
-------------------------
Sum...............9057
Connell gives a number of 11660 in SotMS, but includes the Yanktonnai Dakota, which according to Gray account for 2755. So we can take this number as an accurate representation of the Reservation counts.
The total number of Lakota surrendered up to 1881 is 7395 according to Gray. To this the number remaining in Canada (about 400 according to Bray) has to be added, and the number fled from the Agencies in 187
7 of about 1400 (Gray) has to be subtracted for a total of about
15450according to the military counts.
This result is in close agreement with the historic estimates. It is likely though, that these numbers add up so conveniently a little bit by accident.
The demographics from Red Cloud Agency are slightly suspect (too many children), but could be explained by a plausible number of unaccounted for young men and young couples that sneaked out and in again without getting counted.
For Spotted Tail the number of children appears impossibly high though (>60%), and no realistic amount of young men and couples appear to be possible to make up the difference. So there is likely still some overcounting there.
Still, the error range for those military counts should be fairly tight, in the 14-17000 range, with a likely value of 15500.-------------------
1890s census data
-------------------
Finally, the probably most reliable source of data for the Lakota population in 1876, the census results from the late 1880s and early 1890s.
By that time census procedures were established that yielded consistent results; plausible demographics and stable numbers for the respective agencies.
Demographics and population numbers have a high inertia for a pre-modern society, so only small intrinsic changes would be expected between 1876 and 1890. And as far as I'm aware, no major extrinsic changes were affecting the Lakota on the reservations, pretty much no one left the reservations permanently, and no catastrophic disease losses were reported.
As a reference, the numbers for the "tame" reservations pretty much stayed constant in that period, so it is very likely that the number of Lakota in 1890 is indeed a very good approximation for the number in 1876.
The 1890 census data, as per Gray:
Brulé.............5260
Oglala............4500
Hunkpapa.......1740
Miniconjou......1330
Sihasapa.........760
Sans Arc.........770
Two Kettle.......970
-----------------------
Sum .............
15330(+300-400 In Canada)
Bray gives an average for the 1887-1895 period of
16000If there was a change from 1876, it would likely have been a slight decline due to increased desease load and decreased quality of food, and not to forget the several hundred that stayed in Canada.
Accordingly, i would derive a value for the 1876 Lakota population of about 16000, with a likely range of 15500-17000 from the 1890 census data.
--------------------------------------
Lakota population in 1876, conclusion
--------------------------------------
According to the three different methodologies for estimating the Lakota population in 1876, we arrive at a most plausible range of 15000-17000, with a near-certainty range of 14000-20000.
A value close to 16000 would be my "best guess" from the data shown in this post.For reference:
Lakota reservation "census" numbers (Bray, rounded):
1875.... 35000
1881.... 22000
1890.....15500
Cheyenne
----------------------------------
In short (Gray, Sweet Medicine):
Cheyenne surrendered in 1877 ............1478
Northern Cheyenne census 1889-90.....1400
Likely excess death numbers from the 1876/77 fighting, Oklahoma exile diseases and flight back to to North: about 150
30 Cheyenne lodges turned up in Oklahoma in 1876/77, 24 of those could have been at the LBH, considering their arrival date in Oklahoma, corresponding to about 150-200 people. It's not entirely clear if those are already included in the ~1500 surrendered as per Gray.
Edit: missed the about 25 lodges that stayed at Red Cloud Agency throughout.
Likely range for total number of Cheyenne in the North: 1500-17001650-1900
Dakota
-----------
Santee + Yanktonnai: most likely less than 500 were out in the North country in 1876, though of course the total Dakota number would have been in the ballpark of 10000.
Attachments: