e opposite ideas are so clearly
distinguished that, as Father Carochi warns his readers in his Mexican
Grammar, to confound them would not merely be a grievous solecism in
speech, but a formidable heresy as well.

Another example. What can you make out of this sentence, which is
strictly correct by English grammar: "John told Robert's son that he
must help him"? You can make nothing out of it. It may have any one of
six different meanings, depending on the persons referred to by the
pronouns "he" and "him." No such lamentable confusion could occur in any
American tongue known to me. The Chippeway, for instance, has three
pronouns of the third person, which designate the near and the remote
antecedents with the most lucid accuracy.

There is another point that I must mention in this connection, because I
find that it has almost always been overlooked or misunderstood by
critics of these languages. These have been free in condemning the
synthetic forms of construction. But they seem to be ignorant that their
use is largely optional. Thus, in Mexican, one can arrange the same
sentence in an analytic or a synthetic form, and this is also the case,
in a less degree, in the Algonkin. By this means a remarkable richness
is added to the language. The higher the grade of synthesis employed,
the more striking, elevated, and pointed becomes the expression. In
common life long compounds are rare, while in the native Mexican poetry
each line is often but one word.

Turning now from the structure of these languages to their vocabularies,
I must correct a widespread notion that they are scanty in extent and
deficient in the means to express lofty or abstract ideas.

Of course, there are many tracts of thought and learning familiar to us
now which were utterly unknown to the American aborigines, and not less
so to our own forefathers a few centuries ago. It would be very unfair
to compare the dictionary of an Indian language with the last edition of
Webster's Unabridged. But take the English dictionaries of the latter
half of the sixteenth century, before Spenser and Shakespeare wrote, and
compare them with the Mexican vocabulary of Molina, which contains
about 13,000 words, or with the Maya vocabulary of the convent of
Motul, which presents over 20,000, both prepared at that date, and your
procedure will be just, and you will find it not disadvantageous to the
American side of the question.

The deficiency in abstract terms is ge