¹ The Atlantic Monthly, April, 1915, p. 528.

² Cf. especially Eduard Meyer, Zur Theorie und Methodik der Geschichte, pp. 37 ff., who remarks justly that the different factors contributing to the sentiment of nationality must be investigated for each nationality separately. Yet there is no impropriety in isolating, for purposes of observation, and surveying the rôle of one particular factor,—such as language, the obvious relation of which to national sentiment has not of course been overlooked by writers on nationality, but the relative importance of which is variously estimated. This is what is attempted in the following—not exhaustively, for this would mean a review of all linguistic and political history, but by means of illustrations selected from ancient and modern times. We are not concerned with any doctrinaire thesis of the ideal relationship between language and nationality,—nor, again, with the problem whether what seems to many the economic folly of preserving minor languages from extinction is not more than offset by the sometimes extraordinary impulse to educational progress which attends a linguistic-national revival. We are concerned here only with the question of what the relationship between language and national sentiment has been as a matter of historical fact.

³ And not improperly, if we understand by this the physical descent of, at the very least, a considerable portion of a nation, and especially its social descent. In the reaction against the careless confusion of language and race, there is now perhaps less danger of overestimating than of underestimating the historical bearing of linguistic evidence. For it is still a truism that language implies a people speaking it. It does not pass from one people to another without human agency. In the majority of cases of racial mixture it is the language of the numerically superior element which survives, so that here linguistic descent reflects also the physical descent of the majority. The mere physical domination of a small body of invaders, forming only the ruling class, is not sufficient to impose their language upon the masses. Witness the fate of the Franks or the Normans in France, the Swedish founders of the Russian state, the Asiatic Bulgars, the Manchus in China. In those cases where a minority has imposed its language upon the majority, as for example the Romans in Gaul, this is all the more evidence of the dominance of this minority in social organization. The linguistic descent then reflects a degree of physical descent by no means inconsiderable, but especially what it is fair to call the main line of social descent.

⁴ The Turkish identification of religion and nationality and the centuries-long grouping of all the Christian subjects in one “millet” enabled the Greeks, who were in complete control of church and education, as well as of commerce, not only to maintain but even to extend their language and nationality. The well-to-do Bulgarians and Wallachians spoke Greek and called themselves Greek, and in many cases this meant a permanent accession to Greek nationality. The movement was halted by the nineteenth century revivals of the other Balkan peoples, notably the Bulgarian, which was directed first and foremost against the Greek domination of the church. Although the Greeks are still inclined to claim what they call the “Bulgarian-speaking Greeks,” namely those who ad-here to the Greek patriarch instead of the Bulgarian exarch, this is hopelessly untenable. And, indeed, the undisputed claims of Greek nationality on the linguistic basis offer a goal sufficiently advanced for the most ambitious politician.

⁵ It is to be noted that the Greek language in contrast to Latin was as slow to attain that degree of uniformity which we expect in a “national language,” as was Greek nationality to effect political unity. The particularistic tendency manifested in the system of rival city-states had its counterpart in the linguistic situation. Until a late period there was no Greek language in the sense of a single standard language used throughout Greece, but rather numerous local dialects, each paramount in both spoken and written form within its own domain. These dialects, however, had sufficient resemblance to be in the main mutually intelligible; and furthermore, while no one dialect prevailed exclusively even in literature, the works of the Greek poets, whether in Aeolic, Ionic, Attic or Doric guise, were familiar to all Greece. So, in spite of the long continued diversity in dialect, the Greeks could not fail to realize that they were of one tongue, that all others were “barbarians.” And eventually the dialect of Athens became, first the recognized language of prose literature, and, later in a modified form, the common language of the Greek world.

⁶ See p. 56.

⁷ This is one illustration of a fact which, while not in conflict with the general thesis of the important relation between language and nationality, we do not wish to be accused of slighting,—namely that the linguistic situation itself is not something absolutely fixed, wholly immune from manipulation or independent of external factors. For example, it is owing to the political history of the Netherlands that Dutch is a distinct language, while the similar Low German of Northern Germany is only a dialect of German,—that is, according to the conventional use of “language” and “dialect.” For these terms are incapable of rigid definition by purely linguistic criteria. The German “dialects” of Westphalia and Bavaria differ from each other much more than do the Danish and Swedish “languages,” or Servian and Bulgarian. The popular recognition of a language as such applies primarily to a standard or literary language and the evolution of such is the result of a centralizing process, the manner and extent of which is determined by historical factors. Thus the High German of Luther's bible translation, itself based upon what had already become established as the official language of the Saxon and imperial courts, was finally adopted by the Low German speaking population of North Germany, but never in the Netherlands, though, from the purely linguistic point of view, its use there would not have been materially more difficult or artificial. Another literary language, based upon the native Low German speech had grown up here, and, owing to the political independence, had gained the status of a national language, with all its powers of resistance. The dominance of the speech of Paris marched with the political consolidation of France, and just as the South was not permanently united with the North till after the final extinction of the English claims, so it long remained outside of the linguistic centralization. An ordinance of 1539 requiring the use of French in all the courts brought a formal protest from the Province against being forced to use a language which must be learned like Latin. Had the South of France remained permanently detached from the North, Provençal would have the same status as a language with French. Had the absorption by Castile of the other kingdoms constituting Spain extended to that of Portugal, its standard language would be the same Spanish, based on the dialect of Castile. There would be Portuguese dialects coördinate with Spanish dialects, but no Portuguese language in the conventional sense. But since, owing to the course of history, Portuguese did gain the status of a national language and has maintained it for centuries, its position is of course established beyond the liability to reversal, and the relation to national sentiment the same as elsewhere. Such questions as to “language” or “dialect,” one nationality or two, have long since been settled in most of Europe, despite sporadic movements like the Provençal or Catalan revivals. But we must recall the historical basis of their settlement, when we face such questions as whether Slovak is a distinct language or only a dialect of Bohemian, whether Ruthenian, or more broadly, Little Russian is a distinct language or only a dialect of Russian, whether the Slavic speech in Macedonia is Bulgarian, or Serbian, or just “Macedonian Slav.” For it is folly to imagine that it is only necessary to apply to the philological expert for a categorical answer. The philologist knows that Little Russian is more closely allied to Great Russian (the standard Russian language) than to any other of the Slavic tongues, and at the same time, differs from it much more than the various Great Russian dialects among themselves. But he cannot determine whether, in the course of historical events, its final relation to Russian will be parallel to that of Low German of North Germany to German, or to that of Dutch to German. The philologist knows that the Slavic speech of Macedonia has many points in common with Bulgarian, others in common with Serbian, and he may chart these in intricate “isoglossal” lines. But Serbian and Bulgarian philologists will continue to differ in the interpretation of the same facts; and if the unprejudiced conclusion is that the relation to Bulgarian is the closer, such a scientific dictum is of far less consequence than the prevailing Macedonian sentiment mainly the result of Bulgarian propaganda through their school system in favor of Bulgarian language and nationality. An extended period of either Serbian or Bulgarian control of the schools would secure the dominance of either Serbian or Bulgarian, but the latter with less resistance.

For illustration of an exceptionally conscious design in the creation of a standard language, note what is said about Croatian, p. 63.

⁸ One of Vuk's orthographic reforms was the adoption of one letter from the Latin alphabet. This met with violent opposition, as a fatal and unpatriotic concession to Rome. For even such an external feature of language as the form of the letters employed may become an element of religious and national sentiment. Bismarck is said to have refused to read German books printed in Roman type. In Roumania it was not the greater simplicity of the Latin alphabet that enabled it to oust the Cyrillic (one cannot imagine such a change in Russia), but the “Latin race” sentiment. We have noted the former Russian regulation that Lithuanian books could be printed only in Russian characters. When the Turks made some concessions to the Albanians regarding the written use of their language, they still insisted that Turkish characters should be employed. And while this factor is now eliminated, the two rival systems of writing Albanian in the Latin alphabet are under popular suspicion of being agencies of Austrian or Italian propaganda respectively.