Why singer, but actor? Why doctor, but dentist? Why customer, but client? There seems to be no logic to the variation in endings for agent nouns.
An agent noun, a word that identifies a person’s occupation or profession, place of origin or residence, or other association, or a device that performs a task, generally signals its function with a suffix. However, various endings are possible, and though the spelling of such words, because of their ubiquity, is often obvious, no set of rules is always reliable.
Some people posit that the -er ending is found in words derived from Germanic sources, while the -or ending denotes Latinate words, but that’s not always true. (For example, adapt is from Latin, but its agent noun is adapter, not adaptor.) Alternatively, the -or ending appears in words referring to figures of authority (director, governor, inspector), but employer, manager, and teacher are counterexamples.
(Occasionally, an agent noun has more than one spelling, as with adviser and advisor. It’s common practice in such cases to always prefer, for consistency, the first option listed in the dictionary, although in this case, doing so means that a writer will write adviser yet will spell the adjectival form advisory.)
Another ending for agent nouns, -ist, might at first seem to consistently correspond to a positive value judgment, in that we associate it with scientist and words for those in scientific specialties (biologist, physicist), as well as powerful people (industrialist, philanthropist), but note the neutral bicyclist and the negative bigamist. Likewise, notice the variable status of people identified by the terms cineast (also spelled cineaste), enthusiast, and pederast, as well as the tongue-in-cheek neologism ecdysiast.
Some endings for agent nouns are often associated with pejoratives, though there are exceptions. For example, beggar, burglar, and the archaic pedlar end with -ar, but so do registrar, scholar, and vicar. Bastard, coward, and drunkard, but steward and wizard. Mongrel and wastrel, but colonel and sentinel. Less common endings that might be associated with positive or negative words are -ant (accountant, but vagrant) and -ent (client, as well as agent itself, but indigent).
Other rare examples are -aster (poetaster), which originally neutrally denoted partial resemblance but came to be exclusively derogatory; -ista, which is employed in lightheartedly pejorative terms such as fashionista (there’s also barista, borrowed from Italian, which in turn borrowed bar from English); and -nik (beatnik, peacenik), which in Russian or Yiddish is neutral but acquired a negative connotation in American English during the Cold War.
As mentioned in the definition of “agent noun” above, that term also refers to inanimate objects. However, no consistent classification is possible for this subgroup: Why washer and dryer, but refrigerator? As revealed in these examples — and the many terms for human agents referred to above — agent nouns follow no rules of etymology or connotation.
3 thoughts on “Agent Nouns”
I wonder why we can use “misanthrope”, but not “philanthrope”. Why is it “critic” instead of something like “criticist”? Why is it “judge” and not “judger”? …”cook” and not “cooker”? Some agent nouns have no suffixes and simply take on the infinitive verb form as their own. Interesting…. What about “thief” and “chief”? …seemingly no suffix and no verb indicated in any root in these words. How mysterious!
Not very well researched, this… There are certainly rules to agent noun endings, but like all language rules there are irregularities. For example, ‘-or’ endings do indeed imply a Latin origin. That does not, however, mean that a word deriving from Latin can never end on ‘-er’. Latin itself was not strictly regular, let alone it’s derivations in English. ‘-ist’ just means ‘someone who does this’; the ‘-ard’ in coward and wizard derives from Germanic and means something like ‘(someone’s) nature’, compare for Dutch ‘aard’ (also in ‘lafaard’ or ‘dronkaard’); and I bet a case can be made in the -er vs. -or thing for a derivation via French or directly from Latin.
Yes, I tend to agree with Nienke’s comment and disagree with that last statement of Mark’s text.
Agent nouns DO follow rules of etymology, the big problem is to know where these words come from – and certainly in their origins they already have different endings. For instance washer and dryer do not come from Latin, as refrigerator does. So, why should we expect that these words would have the same -er or -or endings? Just because they “do things”? That’s quite a naive simplification for a system as complex as a natural language.
Maybe a good conclusion would be that there’s no easy rule (at least in English) that relates directly these “agent suffixes” and their original language, because even in their different origins these endings do not follow one unique rule.