SECTION I. .

THE PRINCIPLE OF DEVELOPMENT IN LANGUAGE, AND THE POSSIBILITY OF DISCOVERING EVEN THE MOST REMOTE HISTORICAL AFFINITY BETWEEN LANGUAGES BEYOND

THE GRAMMATICAL FORMS.

A.

OUTLINES OF A METHOD OF RESTORING THE GENEALOGY OF MANKIND BY A PHILOSOPHICAL AND HISTORICAL COMPARISON OF LANGUAGES.

SUPPOSING we wished to find out the position of AngloSaxon in the general system of Germanic languages, and by that means its place in time and its importance in the internal development of that branch of languages, how should we set about it?

In the first place, we should evidently have to discover when and where it branched off from the language of the home country; and secondly, when it first appears in an independent shape, that is to say, different on the one hand from the old mother tongue, on the other from the language of the country which in its own home either grew out of the common mother tongue or the languages of the country.

It is notorious that the Anglo-Saxon grew out of the emigration of the Angles and Jutes, Saxons and Frieses, who about the middle of the fifth century sent colonies to England, and founded there a kingdom. But as regards the second question, Anglo-Saxon was notoriously an independent tongue, in both these respects, towards the middle of the eighth century.

By a similar method we arrive at a similar result in regard to the Icelandic. It was originally nothing but the language spoken in Norway when, towards the end of the ninth century, a number of high-spirited Norsemen fled to the Northern island from the tyranny of Harold (Fair-hair). The Edda contains the record of that ancient Norman tongue, and the writings of the fifteenth century exhibit to us the Icelandic exactly as it is spoken and written at this day; different from the Edda language, as well as from the new Scandinavian idioms which had grown up in the meantime, but still much nearer than they are to the common mother tongue.

Lastly, owing to the immigration and dominion of German races from the fifth century downwards, we find the Romanic languages, Italian, Provençal, French, Spanish, Portuguese, in neighbouring countries, each grown into such an independent shape within six centuries, that none of the above-mentioned peoples could understand each other or the Latin. Still less, however, did they understand the Wallachian, which had been formed contemporaneously in Dacia by the military colonies of the Romans.

Now as the results of all these inquiries are the same, and such indeed as are in harmony with reason, we must suppose them to exhibit an organic law of formation, which is applicable in all stages of affinity, but under a different formula. With this proviso, therefore, we may from our present point of view lay down the following laws as applicable to the organic formation of language:

First One language developes itself out of another without any violent influences; the new one, however, must so far be said to be different, when it is just as unintelligible to those who speak the old language as it would be to those who speak other living offshoots of it.

Secondly Foreign elements find their way into the dictionary in a formation of that kind, as single words, but do not enter into the grammar.

Thirdly: Every stage of language, in such a series of development, becomes poorer in grammatical forms than the older one was, although it may be richer in words, owing to composition or the adoption of foreign terms.

Now the question is, whether we can show that a similar organic development exists beyond the comparison of those languages whose grammatical forms are essentially homophonous. For such languages alone have hitherto been made the subject of scientific comparison.

The results of the researches instituted in this century, which may be called the age of Indo-Germanic philology, are at this stage incontrovertible and universally admitted.

The preceding phenomena, which are clearly verified according to the laws of formation enunciated above, in every branch of one great family of languages, especially in the formation of the Romanic from the Latin, and modern Persian mixed with Arabic from the old Parsee, and which are patent when we compare the Gothic and German, these same phenomena are likewise found to exist upon a scientific comparison of the various branches of the main stems of human language. The Iranian languages, from India to Iceland and Lithuania, are identical in their grammatical structure and roots; as are the different Semitic dialects.

The latest researches, indeed, have proved, or at all events rendered it highly probable, that such organic languages as are neither Iranian nor Semitic represent, for Asia and Europe, earlier, but still already organic, stages of forms of speech; the Turanian of the Iranian, Khamism or Egyptian of the Semitic. Those who will not admit this to the same extent as ourselves in regard to the Turanian, which is the most

extensive and varied of all the families of language,

will still find it difficult to dispute the fact as to its most developed branches. The discussion of the question must certainly be of the strictest kind. The similarity must be proved from the oldest formations in the existing stem: for instance, from the Gothic and Icelandic as to the German, from the Vedic and Zend as to the Ario-Indian, and the old Sclavonic as to the Slave. Such primitive forms may then be compared with those Greek and Latin forms which can be proved to be the oldest and the original. We therefore agree with our predecessors in saying that what constitutes the closest affinity is the similarity, not only of the roots, but also of the mode of expressing the grammatical forms. . But I submit that it is not the only one capable of being tested scientifically. If (as cannot be disputed), wherever a history of a language exists, the grammatical forms turn out invariably to be old complete roots worn down, which by conventional coinage have been used gradually, first as particles, then as affixes and suffixes, and lastly, not unfrequently as pure endings; according to the same analogy of a rational coinage of formation, there must also have been a time when those formative words were still complete roots, and betokened either a thing or a quality. It accordingly may, and therefore will, have happened, that races or peoples speaking languages the roots of which are common to all have separated, and developed them independently, before such a conventional coinage had taken place.

These languages would consequently come within the category of the most remote degree of affinity as formative languages, if, upon comparing their concrete roots, we should meet with common elements, not casually introduced from foreign sources, but pervading the whole original system of roots. The evidence, however, that this agreement is historical, and consequently original, and that it is due to their cognate character,

must be the same as in the case of the closest affinity; that is to say, there must be an analogy both in the points of concordance and those of deviation.

It will also sometimes happen that the germ of grammatical affinities will be found together with a prevailing analogical agreement in the radicals. Where the materials used in the expression of the forms (particles) are entirely different, nothing can be proved beyond a community of type, which is not necessarily a consequence of historical connexion. But there must be a method for ascertaining the fact whether the materials too are of identical origin. We must extend, not alter, the historical method of research.

Even in comparing the different Turanian languages, we have found the necessity of adopting such a method for proving the more remote historical affinity, by means of the affinity in the common type of the grammatical forms, where there was but very slight resemblance in the material used for the expression of it. William Von Humboldt himself, in the then state of linguistic science, even thought it possible that this similarity of grammatical type might be explained by a certain internal organic analogy alone. This view, however, is now highly improbable; for there is too palpable evidence of this common material in the above languages to allow the supposition of there being merely an ideal, and not at the same time an historical, connexion in every instance where we have the means of testing it. The similarity of the pronominal stems in the TuranoFinnish and the Ario-Germanic tongues cannot be explained in any other way; either as accident or as a natural necessity, or as an inorganic sporadic introduction of foreign words. The one is more impossible and senseless than the other. This in itself This in itself proves the historical connexion of the Turanians and Iranians as matter of fact, and the only question is how far the circle of Turanian languages is to be extended.