A further issue to be addressed in Figure 2 concerns the nature and extent of any
bias that different sources might introduce into the populations reported. While there is
no strong indication of bias, there is enough difference in the tendencies observed that
questions can be raised. In particular, the tendency for Christian Missionary, WCD,
Government and especially other sources (mostly entries from the World Almanac) to
report higher population figures might be a concern. As mentioned earlier, these sources
may have a tendency to report ethnic populations in place of language groups. Moreover,
governments tend to favor official and national language groups in their accountings,
which means that smaller groups are likely to be neglected. Finally, these sources are
almost certainly using different counting methods. And given the linguistic survey
requirement of visiting areas for the languages under survey, there may be a tendency for
the language groups surveyed in this way to be under-counted. Those undertaking the
survey tend to have fewer resources than census-takers do, and are unlikely to report
counts including people from villages or areas they have not learned about, even if they
speak one of the languages under survey.
Figure 4. Log
10
population size by decade of citation among language entries.
Figure 4 does for year of citation what Figures 2 and 3 did for region and source.
The first two boxes in the figure have a somewhat deviant shape, on account of the small
number of data points (two for the 1920 category, three for the 1960 category), which
causes the confidence intervals to be somewhat larger than their inter-quartile ranges.
Because of the small number of points they concern, they should not be interpreted. The
remaining categories are interpretable, however, and they show variation over the
different decades of citation that is less extreme than that of region and source. The one
category that appears to be somewhat distinct from the others is the decade centered on
1980, whose reported figures are somewhat smaller than the others. This may reflect the
extensive use of Wurm and Hattori (1981), which is cited 135 times, or in 6.7% of the
entries in our sample. While this is clearly an important reference, over-reliance on a
single such reference, now a quarter-century old, should be avoided if possible. Newer
sources of information about the languages cited in this and other publications of similar
age should be sought.
3.3. Locations
Location information for the Ethnologue’s language entries is found in its maps section.
These can be studied for the information they contain and compared with other maps,
from both linguistic atlases and other sources of geographic information.
3.4.1. Global and regional maps
The very first map in the Ethnologue maps section (pp 674-5) is a global map of language
locations. Each language is represented by a red dot, where the placement of the dot is
intended to represent the geographic center of a living language population. In areas of
high language density, such as equatorial Africa, the Himalayas, Southeast Asia, etc., the
locations are fairly interpretable. In places of relative sparsity, it may be less so, as in
North America, where the eastern coast is nearly devoid of any dots, or in Oceania, where
several dots appear in the open ocean between islands where the corresponding languages
are presumably spoken.
The discrepancies are partly the fault of this form of presentation. Each language
has been assigned a single geographic center, regardless of size, and regardless of how
widely distributed its speakers may be. Consequently, heavily populated areas such as the
New York city area, which have no languages primarily spoken there, with the possible
exception of Yiddish, are simply empty spaces on this map. Moreover, languages whose
center of population may have moved in the 20
th
century, such as Yiddish, are
represented in their ancestral location(s), if at all. Finally, each language is represented as
having a single location, when in fact, many have several discontiguous locations in
which they are spoken (e.g. Romani, of which several varieties exist, all of which are
spoken by minority enclaves in various countries in Europe). The idealization used for
this map is hence misleading in certain ways.
At the same time this map does communicate reasonably well where the main
areas of linguistic diversity remain in the world. These areas lie primarily at or near
equatorial latitudes on all continents, with the highest concentrations of points in Papua
New Guinea, Southeast Asia, Western Africa, the Himalayan mountain range, and
Southern Mexico. Other smaller concentrations of diversity can be found through the
Andes in South America, Northern Australia, the Caucasus mountains, Southern China,
the Southern Indian state of Kerala, etc.
Comparing this map with the maps of Wurm (2001) is quite revealing about the
state of health of linguistic diversity, worldwide. The large number of points in Arctic
North America, compared to the apparent sparsity of points in the same region of the
Ethnologue map suggests that an unusually high proportion of those languages are
endangered. Similar observations hold for most of the regions where points are sparse on
the Ethnologue global map: Europe, Northeast China, Siberia, the Brazilian Amazon
basin. One area that shows an unusually high concentration of endangered languages that
is somewhat denser in the Ethnologue global map is Northern Australia. Other areas of
rather dense language diversity, such as Papua New Guinea and the Himalayas, show
surprisingly small numbers of endangered languages, given the number of languages
spoken there. These are facts calling for some kind of explanation. It is possible that
future surveys in the Himalayan region or in Papua New Guinea will show more
endangered languages. Certainly there are forces operating in the latter region that would
suggest a process of language shift is a concern for many individual groups (Kulick 1992;
Mühlhäusler 1996). At the same time, the aggregate picture suggests greater language
endangerment elsewhere.
Another interpretation that can be made from this map, since languages are
located on the map in their ancestral homes, is that the existing linguistic diversity in
regions of low language location density, such as Australia, North America and Europe,
comes from more recently relocated groups, i.e. immigrant populations, rather than from
indigenous languages. This is an important fact that also requires careful study, although
additional data, such as the language questions in the long form of the US census, will
need to be brought to bear on it.
One may also compare the Ethnologue global map and the endangered languages
maps with locations of displaced persons. For example, the UNEP published a map
locating displaced persons worldwide at the end of 2000, including those not under the
protection of the UNHCR (Rekacewicz 2001). The most striking point of comparison is
in the southern Sudan, where most of Sudan’s 134 living languages are found — 21 of
which are endangered, according to Wurm (2001). Current estimates of the number of
displaced persons and refugees in the Sudan run around 6 million. Clearly the threat to
these languages of Sudan’s ongoing instability is very great. Other notable regions of
concern are Angola, Southern Columbia, the Caucasus region, and Myanmar, all of
which have islands of somewhat greater linguistic diversity than their surroundings,
together with a disproportionate share of displaced persons. Hence internal displacement
of people appears to be a major threat to many areas of local linguistic diversity.
This examination illustrates that the global and regional maps of the Ethnologue
are informative, although in a somewhat roundabout way. Locating modern, living
languages in their ancestral home locations, while adequate for areas of relative linguistic
diversity, does not reveal an accurate picture of the linguistic makeup of metropolitan
areas, especially where one or more large languages are the primary varieties used.
Dispersals of people from their ancestral homes, whether as immigrants, guest workers or
refugees, are also important factors bearing on an accurate representation of linguistic
geography.
Dostları ilə paylaş: |