Using the terms in this sense, Humboldt presents the following terse formula, as his definition of Inflection: “_Inflection is the expression of the category in contrast to the definition of the idea._”[13-*] Nothing could be more definitive and lucid than this concise phrase. The inflectional or formal elements of language are usually derived from words expressing accessory ideas. Generally, they are worn down to single letters or a single syllable, and they usually may be traced back to auxiliary verbs and pronouns. Often various accessories are found which are not required by the main proposition. This is a common fault in the narratives of ignorant men and in languages and dialects of a lower grade. It is seen in the multiplication of auxiliaries and qualifying particles observed in many American languages, where a vast number of needless accessories are brought into every sentence. The nature of the relations expressed by inflections may be manifold, and it is one of the tasks of philosophic grammar to analyze and classify them with reference to the direction of mental action they imply. It is evident that where these relations are varied and numerous, the language gains greatly in picturesqueness and force, and thus reacts with a more stimulating effect on the mind. § 8. THE DEVELOPMENT OF LANGUAGES. Humboldt believed that in this respect languages could be divided into three classes, each representing a stage in progressive development. In the first and lowest stage all the elements are material and significant, and there are no true formal parts of speech. Next above this is where the elements of relation lose their independent significance _where so used_, but retain it elsewhere. The words are not yet fixed in grammatical categories. There is no distinction between verbs and nouns except in use. The plural conveys the idea of many, but the singular not strictly that of unity. Highest of all is that condition of language where every word is subject to grammatical law and shows by its form what category it comes under; and where the relational or formal elements convey no hint of anything but this relation. Here, only, does language attain to that specialization of parts where each element subserves its own purpose and no other, and here only does it correspond with clear and connected thinking. These expressions, however, must not be understood in a genetic sense, as if historically one linguistic class had preceded the other, and led up to it. Humboldt entertained no such view. He distinctly repudiated it. He did not believe in the evolution of languages. The differences of these classes are far more radical than that of sounds and signs; they reach down to the fundamental notions of things. His teaching was that a language without a passive voice, or without a grammatical gender can never acquire one, and consequently it can never perfectly express the conceptions corresponding to these features.[14-*]