Historic American Indian Place Names, Geographical Area and Density in New England: A Statistical Note by waabu


More Info
									                     Historic American Indian Place Names,
                  Geographical Area and Density in New England:
                                A Statistical Note


                                    Frank Waabu O’Brien, Ph.D.
                                     Aquidneck Indian Council
                                       Newport, Rhode Island
                                                                                          July 23, 2012


        Professor John C. Huden’s compilation of American Indian place names in New
England published in 1962 continues to be the most comprehensive source for Indian
names and meanings distributed by language/dialect. Gordon Day (1963) indicated that
Huden worked about 50 years on this research project, an effort not since surpassed. The
author has used Huden for a number of years in his own indigenous toponymic research
in Rhode Island, Massachusetts and Connecticut (O’Brien, 2010).
        Recently an electronic copy of Huden’s book was discovered on the Internet in
PDF file format. This source was used to compile a quantitative listing of names and
languages/dialects. We hope others will avail themselves of this rich source of data and
        Table 1 shows the first search performed. Day mentioned about 4,800 names
compiled by Huden. The electronic search by State showed 4,357 names in Huden’s
book1 . In an effort to understand the variability in the number of names across States, a

correlational analysis was performed by relating present day State size and population
density to the number of historic indigenous names. Table 1 shows the raw data and
below the results of the analysis2.

  The data for place name numbers were obtained by an electronic search of the pdf document of Huden’s
1962 book (on pages 15–298 using the Extract Pages Adobe function, by State abbreviation–R.I. Vt., etc).
Area and Density data were taken from the Internet. The R.I. figure of 534 names is the author’s estimate
based on historic data from many sources (O’Brien, 2010). Huden shows 487 names for this State.
  Gordon Day (1963) estimated about 4,800 total New England place names in the Huden work. As a
check on the 28 megabyte electronic pdf file search providing the 4,357 total names in Table 1, a non–
random manual sample of 20 pages showed the following rough estimation:
    Using the relation, ( P  D ) * T  N , where P = average number of entries per pages (17.2), D = average
    duplicate spelling references per page (2.4), T = 284 total pages, to give a total of about N=4,203 New
    England Indian toponyms (a 97.5% match to the actual number). This estimate implies about 15 entries per
    page ( P  D ) whereas Day’s figure assumes roughly 17 per page—does his number represent the inclusion
    of the D value in the total estimate?

Place Names Analysis. F. Waabu <> Aquidneck Indian Council,  July 7, 2010.                                1
                               Table 1. Distribution of Toponyms by State
                                 and Contemporary Demographic Data3

                        STATE              AREA      POPULATION                    NAMES
                                         (Sq. miles)   DENSITY                    (Historic)
                        Maine              35,385         43.0                      1093
                     Massachusetts         10,554        840.1                      1328
                       Vermont              9,623         67.7                       245
                    New Hampshire           9,350        147.0                       265
                      Connecticut           5,543        739.1                       892
                     Rhode Island           1,214       1,006.0                      534
                       TOTAL               71,669        198.2                      4357   F


    Statistical Analysis

    Correlation Analysis for bi–directional hypotheses:

   Pearson Linear Corr. of Area and Names:                = 0.412
   Rank Order Corr. of Area and Names:                   RS = 0.500
   Pearson Linear Corr. of Density and Names:             = 0.359
   Rank Order Corr. of Density and Names:                RS = 0.643

        All bivariate correlations are statistically non significant at the 5% level or beyond.

   Multiple correlation with Names as dependent variable : R = 0.968

            For a population size of N = 6 States, the linear and rank–order correlations are
    not statistically significant due to such a small population size. But the contemporary
    density–level correlation (0.643) shows a moderately large positive relationship. If all 50
    States showed a similar relationship between size and place name quantities, the Pearson
    linear relationship of   0.412 would be significant (p=0.003, 2–tailed) as would the
    rank order measure. The multiple correlation with Area and Density as predictors shows
    an almost perfect relationship ( R = 0.968). Thus, the State variability in Indian place
    names is amply explained when the two most demographically predictive variables are
    combined as a single linear multivariate function.
            A common sense notion seems to be confirmed by these data—the larger the
    geographic space the greater the number of place names likely to be found describing
      Obviously, present day population density is only a proxy for Colonial indigenous density structure which
    is not well established, but we know it was comparatively very small in southern New England. Pre–
    contact estimates range from a low of 0.6 per sq. mile of land area to 6.5 per sq. mile of land area with
    higher densities on the coast (based on data converted from Bragdon, pp. 25–26). Amerindian toponymic
    typology would be related fundamentally to environmental use, travel, hunting and planting patterns, and
    daily survival.

    Place Names Analysis. F. Waabu <> Aquidneck Indian Council,  July 7, 2010.                              2
features &c. The data best show a 3rd degree polynomial fit which the author cannot
explain theoretically. Perhaps an analysis of livable area and prevalence of varied
geographic features in each State would be relevant aside from the density data which
showed the highest relationship. The sparse but large States Vermont, New Hampshire
and Maine are especially anomalous in this regard. For example, Maine comprises about
50% of the total New England area but only about ¼ of the total place names in all 6
States. Many explanations could be offered for this fact including inadequate compilation
and documentation of Indian names.

        Toponymic Density

        The concept of “Toponymic Density” (see Hunn, 1994) can be addressed by the
data in Table 1. This is simply the number of place names in a given area which previous
research by Hunn and others found to be highly correlated with population density.
These measures are probably less questionable than those of the foregoing analysis. The
overall ratio of New England Historic Names per sq. mile of total area is 0.06, and varies
from a low of .03 in all three large but sparsely populated northern States (Vt., N.H. and
Me.) to a relatively large value of 0.44 in highly densely populated Rhode Island.
        The Pearson linear correlation between modern population density and name
density is substantial,  = 0.836, while the relationship between area and indigenous
name density is negative, –0.562, which is about the same level as the correlation
between area and population density, –0.622, both of which make perfect sense.
Overall, the multiple correlation provided a perfect fit (R = 1.00) with Names as the
dependent variable and Area, Population Density, and Name Density as the three
independent variables. Hence, the variability in Names across all New England States is
soundly explained by an interrelated mix of contemporary and historic demographic
variables. Are these geographical and density quantities meaningful?
        To exemplify this finding for Rhode Island, a look at the 45 cm. x 36 cm. Map of
the Colony of Rhode Island (Rider, 1904, reprinted in O’Brien) shows a widely dispersed
pattern of indigenous toponyms throughout the entire State. The historic name density is
much higher visually in the coastal regions of the State comprising nearly 75% of the
total toponyms, and where 21st century demographic density averages the highest. A
correlation among the 5 Rhode Island counties of Population and Name Density
continues to be positive (  0.595) —these data come from Huden which, as indicated,
undercounts the Rhode Island toponyms, a difference which could skew a correlation
coefficient for such a small population. Earlier we reported that the Native population
was also highest on coastal regions of the Atlantic seaboard. Moreover, historical
documentary evidence points to large Wampanoag and Narragansett Tribe villages in
those regions (Trigger, 1978, and Bragdon, 1996). Thus, these data do make sense.
Apparently, the sparsely populated wooded hinterlands were primarily Native hunting
and foraging grounds while the coastal regions were dwelling, fishing and planting
grounds. These data are amenable to further theoretical analysis.
        We note that in Rhode Island only about            rd   of the total historic place names
exist in current U.S. Government data bases (O’Brien, p. 10). Two–thirds of the historic
Indian place names were replaced by non Indian names or retired altogether. It would be

Place Names Analysis. F. Waabu <> Aquidneck Indian Council,  July 7, 2010.                    3
of interest to analyze this trend for other New England States to determine the correlative
cultural loss and identity of American Indian history.

    Languages and Dialects in New England Place Names

        Table 2 shows the dominant languages and dialects comprising the New England
place names complied by Huden. As is well known, all of the languages in southern New
England are extinct (Goddard, “Eastern Algonquian Languages,” in Trigger, 1978).
Trigger (1978) and Goddard (1996) were used to suggest language/dialect names in the
e–search. Some names in the Huden document are combined (“Pequot–Mohegan”),
some entries contain no linguistic information, and some categories may have been
missed, but about 97% coverage is achieved compared to the number of discoverable
place names in Table 1; for this percentage we add into the Table 2 totals the 47 Rhode
Island place names missed by Huden.

        Table 2. Distribution of Languages/Dialects in New England Place Names

LANGUAGE/DIALECT NUMBER                       LANGUAGE/DIALECT                    NUMBER
Abnaki                   778                  Tunxis                             29
Wampanoag                587                  Delaware                           22
Narragansett             502                  Pocumtuck                          20
Natick                   413                  Chippewa                           17
Nipmuck                  404                  Wappinger                          16
Mohegan                  228                  Hammonassett                       13
Niantic                  206                  Podunk                             12
Mahican                  165                  Pocasset                           9
Malecite                 150                  Montauk                            8
Pennacook                131                  Nauset                             5
Wangunk                  81                   Oneida                             4
Mohawk                   79                   Huron                              3
Paugusset                78                   Onondaga                           2
Micmac                   72                   Seneca                             2
Quinnipiac               72                   Shawnee                            1
Pequot                   50                   Sokokis                            1
Siwanoy                  29                   TOTAL                              4189
  NOTE: Rhode Island is underestimated by 47 names per footnote 1 above.

Place Names Analysis. F. Waabu <> Aquidneck Indian Council,  July 7, 2010.              4
                                References and Sources

Afable, Patricia O. & Madison S. Beeler (1996). "Place Names". Pp. 185–199 in Ives
Goddard (ed.), Handbook of North American Indians, vol. 17. Washington, DC: Smithsonian

Bragdon, Kathleen J. (1996). Native People of Southern New England, 1500–1650. Norman,
Oklahoma: University of Oklahoma Press.

Day, Gordon M. (1963). Review of Indian Place Names of New England. American
Anthropologist, 65(5): 1198–1199.

Goddard, Ives (Ed.). (1996). Handbook of North American Indians, vol. 17. Washington,
DC: Smithsonian Institution.

Huden, John C (1962). Indian Place Names of New England. New York: Museum of the
American Indian, Heye Foundation, vol. XVIII.
[ http://archive.org/details/indianplacenames00hude ].
HU                                                        UH

Hunn, E. (1994). Place–Names, Population Density, and the Magic Number 500. Current
Anthropology, 35(1): 81–85.

O’Brien, Frank Waabu (2010). Understanding Indian Place Names in Southern New
England. Colorado: The Bäuu Institute Press.

Rider, S. S. (1904). Map of the Colony of Rhode Island: Giving the Indian Names of
Locations and the Locations of Great Events in Indian History with Present Political
Divisions Indicated. In The Lands of Rhode Island as They Were Known to Caunounicus and
Miatunnomu When Roger Williams Came in1636. Providence, Rhode Island: Sidney S.

Snow, Dean and Kim M. Lanphear (1988). European Contact and Indian Depopulation in the
Northeast: The Timing of the First Epidemics. Ethnohistory 35(1): 15–33.

Trigger, Bruce G. (Ed.). (1978). Handbook of North American Indians, vol. 15. Washington,
DC: Smithsonian Institute.


     Place Names Analysis. F. Waabu <> Aquidneck Indian Council,  July 7, 2010.        5

To top