A typical charactersitic of inflectional suffixes is that they are productive (can occur on many words) and obligatory (are required by virtue of certain grammatical constraints). Simplistic though this may be, I was wondering about the suffix -er in English. This suffix is typically derivational insofar as it changes the category of the word bake>baker, for example. But is it not also highly productive? Fish>fisher, sleep>sleeper, mouse>mouser, eat>eater, smell>smeller, read>reader etc etc etc.

On what basis is this form truly derivational then?

  • In this particular case, -er is a derivational suffix. I'm unaware of a 20-21 century linguist who'd analyze it otherwise. On the other hand, if you want to become familiar with how derivation vs. inflection is treated in modern linguistic morphology, start with an excellent - and concise - summary in Bauer 2003, Chapter 6. This stuff is always discussed in undergrad. – Alex B. Jul 8 '14 at 3:33

-er a derivational suffix because it changes the word class to which the entire expression belongs. That is what defines derivational affixes.

bake is a verb, but bak-er is a noun. (I assume the stem bak because the final letter e is not pronounced.)

Productivity is not a sufficient criterion for the distinction of inflection and derivation.

English uses zero-conversion, i.e. the possibility that one expression belongs to more than one word class, e.g. The dog wants to run, and a run. That means that one must take care when looking at the verb/noun distinction in English.

  • 1
    @mateuz Globally, yes. There's room to quibble, though. At least productive is a scalable notion. The question is then from which point onwards we want to view an affix as productive. I'm not sure about obligatory: in what way is -er obligatory? – Thomas Gross Jul 6 '14 at 12:55
  • 1
    @mateuz re productivity. I don't think I'd employ this notion concerning the distinction inflection/derivation. Often closed word classes have paradigms different from open classes, but does that mean that affixes for closed word classes are less productive? If you insist on using the notion I'd suggest the opposite approach: an affix may not be unique. – Thomas Gross Jul 6 '14 at 13:43
  • 1
    @mateuz re typology. The standard approach is always to view inflectional suffixes as following derivational suffixes. It's bak-er-s, but not bak-s-er, etc. However, the distinction between inflection and derivation remains artificial if it doesn't head gradience. Clearly, derivational affixes can grammaticalize into inflectional ones, and that means that there must be a greyzone. It's not a black/white issue. – Thomas Gross Jul 6 '14 at 13:48
  • 4
    The really important thing about inflections is that they're Paradigmatic. That is, they form paradigms, although in agglutinative languages the paradigms tend to be one-dimensional; nevertheless, they are paradigms, and that's what defines inflection. That's also why English has so little inflection; there are almost no productive paradigms left in English. Plenty of non-productive ones, like this, though. – jlawler Jul 6 '14 at 18:23
  • 1
    @mateuz, why do you think the plural isn't part of a paradigm? – Alex B. Jul 6 '14 at 22:45

I don't think you've said the correct distinguishing factors between derivational and inflectional affixes.

The primary factor I think is that derivational affixes often change the part of speech of a lexical item, and inflectional affixes don't. What this means is that the concept of the lexical item changes substantially. There is a huge difference between the concept of running, a physical action, and the concept of a runner, a person who regularly runs because of it being their habit or job. These conceptual differences mean that derived words are often listed separately in dictionaries. Additionally, semantic change will affect the related parts of speech separately. Consider the difference between a lecture, to lecture and a lecturer. To lecture has changed it's meaning so that it now almost always means to tell someone off with a long speech. A lecturer retains the original meaning of someone who teaches through speeches. A lecture has both meanings. But note that not all derivational affixes do change the part of speech: the suffix -ship turns the noun friend into another noun, friendship.

Inflectional affixes generally don't communicate real-world or dictionary information. They instead communicate semantic categories. There's no hard and fast rule about whether something will be a semantic category or not, but we all know the really common ones: plurality, tense, agreement markers etc. No English dictionary would have separate entries for dog and dogs because the concepts are too similar, which is a good indicator that the plural marker is inflectional. No English dictionary will have separate entries for jumped and jumping, so those tense/aspect markers are inflectional.

Sometimes it's a little bit hard to decide whether an affix is derivational or inflectional. Which one is dis- in disclose?

  • 1
    This is a great answer. Thank you. One quick follow-up. You wrote "The primary factor I think is that derivational affixes often change the part of speech of a lexical item, and inflectional affixes don't" -- but perhaps it is more accurate to simplify this, "... often significantly change the meaning of the word to which they attach", e.g. bake>baker>bakery. This way, both -r and -ery can be considered as derivational rather than inflectional. What do you think? – Teusz Jul 7 '14 at 5:59
  • Yes that's a good thought. – curiousdannii Jul 7 '14 at 6:50
  • So, a good way to determine if a suffix is inflectional, is with reference to: (1) Can it attach to many words? (2) Is it grammatically necessary? (3) Does it have little/no semantic effect on the meaning of the word to which it attaches? [this last one seems hard to elaborate -- what do you say?] – Teusz Jul 7 '14 at 8:16
  • 1
    No I don't think any of those are necessarily accurate, but it's hard to come up with a single universally acceptable definition. It may be that there isn't a binary difference but more of a spectrum with prototypical derivational and inflectional affixes at each end. – curiousdannii Jul 7 '14 at 8:28

I am not convinced that either one of these answers in correct. Most Indo-European languages have suffixes that make verbs into participles (i.e. adjectives), like English go > going, Latin amo > amans, Greek legō > legōn. All grammars describe these as inflectional elements. They are not listed separately in dictionaries. I have the feeling that the distinction between “inflections” and “derivational suffixes” is largely artificial.

  • But what isn't a convention in linguistics? Even concepts like noun and verb are arguably conventions. Anyway, I see your point. Perhaps there's no real distinction between inflection and derivation... (though as the examples you gave are part of paradigms, apply productively, they clearly have inflectional aspects, no? perhaps they are just LESS inflectional than others) – Teusz Jul 7 '14 at 18:11
  • 1
    Derived forms are normally not listed in dictionaries if they are predictable. Eg in Australian languages, wherein causative is often a highly productive and predictable process, derived causative verbs will often not be listed in dictionaries unless there is something idiosyncratic (i.e. unpredictable) about them. – Gaston Ümlaut Jul 10 '14 at 5:16
  • While participles behave similarly to adjectives, they are not generally considered to be the same part of speech in linguistics. From what I remember, in traditional Latin grammar participles were a word class of their own, and in modern grammatical analyses of English participles are considered to be verbs. Relevant question about a similar topic: Is the {-ing} of the gerund a verbal inflectional suffix? – brass tacks Dec 27 '16 at 9:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.