Concept
A synthetic language is a language, which is statistically wise characterized by a higher morpheme-to-word ratio. In contrast to analytic languages, which break up concepts into separate words, synthetic languages combine (synthesize) them into a single word. The syntactic role a word may have in a sentence, such as a subject or an object, is assigned to the word by adding affixes (characteristic for fusional languages, a subtype of synthetic languages). In the present-day English, once a fusional language, only a few remnants of its fusional origin are retained: for example, the role of an object is assigned to the word who by adding affix m to it (resulting in whom); a different tense is assigned to a word by adding affixes, such as -ed and -ing, to a verb; a possessive role is assigned to a word by adding an apostrophe and 's' to it; either a comparative form is assigned by adding affix -er (resulting in faster) or it changes a verb to a noun (resulting in teacher). Instead of using affixes to assign individual words their syntactic roles in a sentence, the analytic languages predominantly use auxiliary verbs and word order, instead. Combining two or more morphemes into one word is used in agglutinating languages, instead. Subsubcategories include polysynthetic languages (most of them in an agglutinative subcategory, with an exception of Navajo and other Athabaskan languages that are often categorized as fusional), and oligosynthetic languages (only found in constructed languages). Derivational and relational synthesis are opposite ends of a spectrum. In derivational synthesis, many whole nouns, verbs, affixes, etc, are synthesized into new words with a new concrete meaning. In relational synthesis, affixes are attached to a root word to assign it a syntactic role in a sentence. A single root can be involved in both kinds of synthesis, each of them more or less frequently.