68 THE LANGUAGE OF IMMIGRANT PEOPLES
tongue, we have an index of the keenness of the respective foreign stocks in learning the
English language. As an index it is crude, and the question arises as to what other factors
besides pure differences of “crigin” are involved and how far they interfere with its use as
a measure of speed in acquiring the language. If we compare the percentages in Column
3 with the proportions of the respective stocks living in cities 25000 and over by the
method of rank correlation, a coefficient of + +04 is found, which indicates that whether a
people is predominantly rural or urban has little or no effect on the readiness with which
they learn English.
Some relationship appears, however, between the percentage of a stock North American
born (a crude index of length of residence among English speaking people) and the proportions
 of those speaking foreign languages who had learned English. The rank coefficient
 was found to be 4 -29.
More significant correlation would be expected were comparison made with the length
of time the foreign born sections of the different “origin” groups had been in Canada.
Unfortunately data as to length of residence of immigrants are not available by origins, but
in Chapter II a table was shown, giving the average number of years the immigrants who had
arrived from foreign countries since January 1, 1901, had been resident in Canada at the
date of the last census. Such a table, however, has its limitations. In the first place, it
sould be prepared only for those nationalities whose geographical boundaries had not been
radically changed during the last two decades, and in the second place, as has already been
pointed out, origin and birthplace are in many cases by no means identical. However, keeping
 in mind the necessity of caution in comparing such data, if the origins be ranked according
 to the percentage of those of foreign mother tongue who had learned English outside
the home and the average for the upper and lower half of the table be compared with the
corresponding figures for length of residence of immigrants by country of birth, there appears
to be a year’s difference between the upper and lower group. Further, the coefficient of
correlation by the method of rank differences was found to be + +37 and would probably be
somewhat higher were the data on immigration available by origins. The implication is
that the length of time the new arrivals have been in Canada is a significant factor in determining
 the proportions who have learned to speak English.
The correlation, however, is only moderate, which implies that date of arrival, though
exerting an appreciable influence on the data in Column 3, is far from adequate to account
‘or the differences between the figures for the various origins. One is driven, therefore, to
the conclusion that differences in origins are of major importance in the acquiring of English.
The several stocks show by no means the same keenness nor aptitude. Many examples of
such differences are revealed by a detailed comparison of Column 3 and Column 8. For
instance, it is obvious from the data that the Danes, Swiss and Greeks learn English comparatively
 rapidly, while the Austrians, Roumanians, Hungarians and Icelanders are slow in
acquiring it.
The above conclusion, namely, that speed in acquiring English is largely a matter of
origin, is confirmed by the appearance of a marked relationship between the tendency to
ntermarriage with the British stocks and the percentage learning English. Omitting the
igure for the Dutch, where the percentage learning to speak English is unduly reduced by
the Mennonites in the Prairie Provinces, and that for the Hebrews, where a very small
proportion intermarrying is coupled with a very large percentage learning English for
occupational reasons,—omitting these two exceptional cases—a coefficient of correlation by
the method of rank differences of + +77 is found to exist between the two series, Were the
figures on intermarriage in Column 6 representative of both sexes, the correlation would
orobably be higher. It is evident, therefore, that those stocks which tend to intermarriage
with the British learn English most rapidly.
Turning now from the question of speed in acquiring English to the proportions of the
various peoples unable to speak either English or French at the date of the last census,
one finds, as in the former case, no correlation with the proportions in cities 25,000 and
over. That length of residence among English and French speaking people has a bearing
on the percentages of foreign stocks unable to speak either of these languages is shown by