Terminology sometimes refers to specialized language, but it also can refer simply to language that requires more than one word as its signifier. Are terms then in this second sense not also words?
There are two ways that words can contain more that one signifier and still – incontrovertibly – be called words: the compound word (i.e. sailboat, birdbrain, loveshack) and the hyphenated word (i.e. half-baked, gold-digger, mind-boggled). These both leave unaccounted the so-called term (i.e. fish tank, human history, credit cards). Terms refer to a single idea or thing, so in this sense they are cleanly nouns, and yet they are simultaneously not words but are instead a kind of word-set, something universally compounded in usage but never formally linked whether typographically or grammatically. They are paradoxically two-word words.
While this contradiction in the definition of “word” may or may not hold, the OED points to a crack in the foundation of “term” as well:
13 a: A word or phrase used in a definite or precise sense in some particular subject, as a science or art; a technical expression
13: b. In wider application: Any word or group of words expressing a notion or conception, or denoting an object of thought; an expression (for something). Generally with qualifying adj. or phrase.
In the first definition, terms are something specified, technical, scientific and exacting; yet in the second, the thing is general, vague, whimsical, and inexact.
The word “signifier” is a linguistic term that appeals to the ambiguities and tensions of our common conceptions and definitions for the word “word.” Presumably, a signifier refers to what is being used to signal the thing subjected to discussion (signified) whether it is through a hyphenated word, compound word, utterance, term, word-set, sentence, sneeze, or hiccup.
These gaps in clarification highlight just how much grammar (which is a kind of mutant field of philology) and Linguistics (which is both a pseudo-anthropology and a pseudo-neuroscience) are distinctly separate disciplines. Besides their socio-political differences (Grammar is insular, Linguistics is cross-cultural), they have fundamentally different conceptions of what builds our symbolic networks. Each is helpfully insufficient, and both curiously incompatible. Moreover, both the slipperiness of the terms and their porously abstract conceptualizations that attempt to (with varyingly degrees and success) buttress as well as conceal the mechanistic virtues (or even sense) of either further supports the counterintuitive claim that language is far more material and physically resonant than semioticians or grammarians have historically highlighted when configuring the Whats and the Hows of language.