Derivational Morphology: Roots, Categorizers, and Compounding

Derivational morphology is how words are formed. It works through four basic mechanisms: roots, categorizers, derivational affixes, and compounding.

Roots

A root is a morpheme that contributes phonological and lexical-semantic information to a word; roots carry the lexical information and are the best place to start. The study of roots has advanced sharply in recent linguistics, much of it drawing on Semitic languages, especially Hebrew, where a root is very visibly a sequence of three written consonants. The broader claim is that all languages have roots, even if their morphological systems do different things with them.

Roots may be bound or free. A bound morpheme necessarily occurs with another morpheme, whereas a free morpheme has a meaning standing on its own. The sentence supplies both: dermatologist contains the root derma- or dermat-, which comes from Greek (the -t- surfaces in the genitive) and cannot stand alone in English, so it is bound; girl, from which girly can be formed but which also stands alone, is free. In Hebrew and Greek, roots are always bound, because they must co-occur with another morpheme to be interpreted. In Hebrew, for instance, a root cannot be interpreted without its vowels; the language was written without vowels but never spoken without them.

A key consequence is that similarity in root does not predict similarity in meaning. The word journal (in the latest journal articles) differs sharply in meaning from journey, though both come from the same root, because words drift over time. Sometimes words sharing a root do stay close in meaning, as with journalism and journalist. The logical error to avoid is: root X yields these two words, therefore they share this component of meaning. The real question is when two words from the same root share meaning, and what component the shared root contributes. Synchronically one can usually say they share some component, but drift can make the relationship hard to trace.

Categorizers

Roots are acategorial: they have no grammatical category on their own, and they always combine with categorizers. A categorizer is a morpheme that turns a root into a noun, verb, or adjective.

In English, categorizers may be inaudible, a so-called zero categorizer. The root kick has no evident category until placed in context: a kick is a noun (marked by the article a), while kicked is a verb (marked by tense -ed). In Hebrew, categorizers are realized as vowel patterns or templates, as with melek, "king," versus malak, "to reign." In Greek, categories are normally revealed by inflectional morphology, so that a noun ending signals a noun and an -ō (omega) ending signals a verb. In every case a root must have a categorizer to be interpreted.

The implication is that a word's meaning is the semantics of the root plus the semantics of the categorizer. If a root consistently carries some idea, say something to do with being right, then determining what its various derived words mean becomes a matter of working out what each categorizer contributes: the difference between the adjective form, the verb form, and the noun form built on the same root.

Derivational affixes

A derivational affix is a morpheme that turns a root or a word into a noun, verb, or adjective. On this definition a categorizer is a kind of derivational affix; the label "derivational affix" is reserved especially for cases where one word is turned into another word of a different category.

The sentence offers dermatology becoming dermatologist, where the -ist suffix transparently yields "someone who does or studies dermatology." It also offers breed and breeding: since breed stands alone, it can be treated as a root combining with -ing to make a noun, or, since breed is also a verb, as a verb embedded under the noun breeding. Which analysis holds is an empirical question tied to the interpretation of breeding.

Derivational affixes carry meanings and may or may not be compatible with certain readings. Greek apokalyptō, "I reveal," versus apokalypsis, "revelation," shows the -sis ending making a noun from a verb while retaining a verbal idea; the same reasoning underlies the argued relationship between pistis and pisteuō. English has been studied most fully here, particularly for nominalization, the process of creating new nouns. In English Nouns: The Ecology of Nominalization (2016), Lieber catalogues the affixes and their meanings. The suffix -ation, for example, takes a verb base such as examine to yield examination, which allows an event interpretation, a result or referential interpretation ("the examination arrived in the mail"), and readings for agents, instruments, patients, locatives, and paths. The point is not that every -ation word carries all these meanings, but that the suffix is compatible with them, with the base word constraining which surface. Likewise -er yields prayer (something produced) and player (an agent), the reading depending on the base. Doing this work for Greek or Hebrew remains largely open, since it requires very large corpora of the sort available for English.

Compounding

Compounding is the combination of two lexemes to form a new lexeme. The sentence supplies well read, formed from well and read.

The relationship between the two elements varies. A can opener is an instrument that opens cans, but a dog walker is a person who walks dogs, not an instrument. A greenhouse is not a green house at all: the combination is non-compositional, much like an idiom, and today denotes anything with the right heat- and moisture-holding properties. By contrast blue-green simply means something between blue and green. Languages vary in how readily they compound: Greek compounds readily and English is very good at it, while Hebrew is quite poor, with few clear examples. One possible case is tsalmavet, literally rendered "the shadow of death" by placing two components side by side, though, as with greenhouse, the true meaning must be found by identifying both components and examining how the compound is actually used.