1. Prepositions, indicating relations of objects of thought; 2. Conjunctions, indicating relations in the thought itself;

3. Auxiliaries, used to help out inflections;

4. Expletives, indicating some rhetorical relation;

5. Interjections, indicating a modification of thought by some feeling.

§ 9. The medium through which thought is communicated, as before stated, is articulate sound. In order that it should serve as such a medium, it is evident, the sound must be accepted, by both speaker and hearer, as symbolizing the thought. In other words, the sound must be identified, in some way, with the thought, so as to be recognized by both as expressing it. This may be in several different ways.

First, certain articulate sounds are natural expressions of certain thoughts, especially such as are generally embodied in feelings. Thus, contempt and scorn find a natural expression in nasal articulations; affection and endearment in labials. Such associations of sentiment with sound might give origin to words, or, in combination with other associations, determine their form, so that speaker and hearer should recognize them as symbolizing the thought. Some theorists go so far as to attribute the origin of speech exclusively to this principle of association, or of natural expression Theirs is the so-called Interjectional theory of language. It is founded only in partial truth.

Secondly, many objects in nature have a sound peculiar to them which may be more or less perfectly imitated in articulation. The wind whispers, the crow croaks, the horse neighs. So, feelings expressed in interjections are imitated in articulated sounds to form words; thus, the feeling of pain expressed in the interjection ah, is imitated in the word ache. A theory has hence originated which founds language in the imitation of natural sounds the Imitative

or Onomatopoetic theory. It is, like the first named, only of partial truth. They both point to an association or identification of the thought with the sound, as the indispensable condition of introducing an articulate sound to symbolize the thought. These two grounds of association are undoubtedly principles of wide application in the forming and transforming of words; but the more correct view is that word-formation may be grounded on any kind of identification of the sound with the thought possible in human experience, not on natural ejaculatory expression alone, or on imitation of sound alone, or on both conjointly to the exclusion of other grounds of association.

The formation of language implies other identifications than those of the articulate sound with the thought to be expressed. After speech has begun with ejaculations, and imitations of natural sounds, the words thus introduced, both through themselves as sounds and also through the objects which they symbolize, can be identified with other thoughts indefinitely, and thus furnish the necessary condition for the indefinite growth of language. Language is too far advanced from its beginnings to justify much reliance on the etymology of primitive words; yet there is nothing improbable in supposing that ba being taken to symbolize the imperfect articulations of a child, the word might, with slight changes, adopted perhaps for the very purpose of meeting the change in meaning, be applied to the utterance itself; or to the child that makes it; to any person or thing that makes a similar sound, as babbler, babbling brook; to the place where imperfect or unintelligible utterances are made, as Babel; in short, to any thought that can by any accident of quality, effect, condition of place or time, or relation of any kind, in whole or in part, in respect of sound or meaning, be associated with it or any of its derivatives or modifications. The general fact is, that words are formed freely on the condition of any identification of the sound with the object of thought, immediate

or remote.

§ 10. The primitive words of language, as has been already observed, would naturally be, for the most part, although not of necessity universally, single utterances monosyllables. Those languages we have accordingly ranked as being of the first or lowest gradation, the words in which are mainly monosyllables. They are of course characterized as having no accentuations, no inflections, no formative elements. No language is absolutely and strictly of this character; but in some, as the Chinese, this type prevails and gives character to them. We have recognized a great advance in the progress of language when it comes to admit freely combinations of words into one under a single accent. Languages in which this type prevails we have marked as of the second gradation. They are called agglutinative languages. We have recognized as the last stage in the development of language, that in which not merely modifications of objects of thought, as in the agglutinative type, but relations in the thought itself through the formative elements, so called, are freely expressed. This type has been denominated the Inflectional.

It should be observed, that these stages are only stages of development, and are not necessarily exact chronological stages. It by no means follows that a language is older than another because it is less developed, or has less of an inflectional character. A very highly inflected language has developed itself almost at once out of a monosyllabic or agglutinative type; and highly inflected languages rapidly wear off the nferely formative parts of words on the intermixture of tribes speaking different dialects. "Turanian languages," says Prof. Max Müller,1 meaning by this designation those generally classed as of the second or agglutinative stage, are so pliant that they bend themselves to endless combinations and complexities, unless a national literature or a frequent intercourse with other tribes act as safeguards against dialectical schism. Tribes who have no

66

1 Report on Turanian Languages in Bunsen's Outlines, i. 482, 483.

literature and no sort of intellectual occupation seem occasionally to take a delight in working their languge to the utmost limits of grammatical expansion." The English, although a very recent language, has little of the inflectional element; while some of the oldest languages which have a literature rank among the highest in this respect.

§ 11. Language, as the communication of thought through articulate sound, is ever changing. It changes with the ever varying thought of the community that speak it; it changes in the kind of thoughts expressed, and the number of thoughts, in the kind and number of words, and of the relations expressed through the various forms; it changes in the sounds, through all the influences that can affect articulation, either subjectively and immediately through facility of utterance, or objectively and mediately through agreeableness of effect on the ear.

If social thought is active and inventive, the dialect will grow in its vocabulary and in its formative elements with comparative rapidity. If social thought be stagnant, the dialect will be stagnant. Its vocabulary may change from mere mental inertia of retention, as sheer necessity maj occasion the introduction of new words; but its formative elements, if there be any, will tend to die out.

A literature, even if only oral, as in traditions, legends songs, and the like, but especially if written, is the grea conservative force in language, preserving words, preserv ing grammatical forms.

So, likewise, large communities, speaking the same lan guage, if the communication be active, are more conservative than small communities. Small wandering tribes characterized as they are by poverty of thought as of goods if without a literature, must have meagre vocabularies, and few or no grammatical elements. Their dialects may change entirely in a few generations. It has been ascertained that even in England, the entire vocabulary of some of the peasantry contains less than three hundred words; and in

some heathen tribes, missionaries have found a dictionary to become useless after a period of ten years. In the early state of the race, when families or tribes were wandering from place to place, with no literature, even although they may have proceeded from one stock, and originally have spoken the same dialect, a few years would suffice to obliterate the primitive language among them, and there would be almost as many dialects as separate tribes or families. If, in some cases, the dialect should happen through the mere stagnancy of the thought to become crystallized, so as to be transmitted unchanged from generation to generation, in other cases, where great activity of thought in social directions, especially if characterized by an imaginative and inventive cast, happened to prevail, the language which would naturally embody this social activity of thought would become rich both in vocabulary and in formative elements.

Still further, in a nomadic condition, as small roving tribes, speaking different dialects, met with one another in friendship or in strife, even if these dialects were inflected to a greater or a less degree, the object-words would of necessity come most into use; the formative elements would be dropped, and the resulting dialect, if the tribes continued together, would be a lapse back to the more primitive stages would be at first, at least, more agglutinative, or more monosyllabic, until from this relatively primitive germ, a new language-forming movement should start. Such is the teaching of all history. Such especially is the teaching of the history of the English tongue.

Language is ever changing. It changes with the social thought of the community that speak it, with the kind, the objects, and the directions of thought, and with the degree of social activity. It begins with the simplest articulate utterances accidentally associated with the object of thought common to those that form it. Words denoting objects come soon to be used to denote relations of thought. Notion-words become form-words. Words at first mono