According to the guidelines of morphology, the linguistics branch concerned with the internal structure of words, a morpheme is the very smallest meaningful linguistic unit in the grammar of a language. In written language, morphemes are composed of graphemes, or the smallest units of typography. In oral language, however, morphemes are composed of phonemes, or the smallest units of speech.

A morpheme can be either a word or it can be just an element of a word. For example, the word “technique” is both a word and a morpheme because it cannot be broken down into any smaller meaningful parts. A more complex example is the word “unkindly,” which consists of three morphemes. The morphemes are “un” which means not, “kind” which means benevolent, and “ly” which means like. None of these morphemes can be broken up into smaller parts without losing all semantic meaning.

A morpheme is categorized depending on how it combines with other morphemes in order to create a word. A free morpheme, which is often called an unbound morpheme, is a linguistic unit that is able to stand alone as a word without another morpheme attached to it. The word “cat” is a good example of a free morpheme.

Bound morphemes, on the other hand, are sounds or a combination of sounds which must be bound to a free morpheme in order to create a word. Most prefixes and suffixes are considered bound morphemes. For example, the letter “s” in the word “dogs” is a bound morpheme because it does not have any semantic meaning without the free morpheme “dog.”

In the English language, morphemes may also be classified into the following types: root, stem, or affix. A root, sometimes called a base, is the morpheme which gives the word its meaning. For example, the root morpheme “bird” gives the word “birds” its meaning, which is a particular type of animal. A stem is the root of a word combined with any affixes. An affix is a morpheme that comes at the beginning, called a prefix, or at the end, called a suffix, of a root morpheme.

Many people believe that morphemes are the same as syllables, but this would be incorrect. The word “apple,” for example, has two syllables but only one morpheme. “Apple” cannot be broken up into "app" and "le," which are two separate syllables, because “app” and “le” have no semantic meaning on their own.