“Suffix” or “Ending”—Clarifying the Terminology

Aug 20, 2015 by

A reader of my earlier post “Some Observations on Morphophonological Adaptation of English-derived Loanwords in Russian Slang” pointed out that my use of the term “suffix” for the plural marker in Russian is incorrect; the term “ending” should be used instead. Although I have already responded to the reader in private, I will also clarify this potential terminological confusion here.

The Russian linguistic tradition draws a 3-way distinction between “prefixes”, “suffixes”, and “endings”. “Prefixes” in this system are, as one would expect, morphemes that attach to the beginning of the root. Both “suffixes” and “endings” attach to the end of the root, with the “endings” attaching—as their name suggests—at the end, that is outside any “suffixes” that may be present. (The only exception appears to be the morpheme ‑sja, which attaches outside any “endings” encoding agreement on verbs in person/number/gender; but see Nesset 1998a,b.) Back in early middle school, we had to use different notation for “suffixes” and “endings” when doing morphological analysis: “suffixes” were marked with an “accent grave”-like diacritic and “endings” were put into little squares. (Yes, we did do this sort of thing as early as the 4th grade!)

So what is the difference between “suffixes” and “endings” in the Russian model? According to the textbook definitions, a “suffix” is “a functional morpheme, attaching after the root (immediately or after another suffix) and serving to form new words or their non-syntactic forms”, whereas an “ending” is “a morpheme serving to change word forms and to express grammatical meanings: number, gender, case, person” (translation mine). The “endings”, we also learn, serve to connect words to the rest of the sentence. In a word like knižka, –k is a “suffix” (a diminutive, to be more precise), while -a is an “ending”, expressing declension class, gender, number, and case.

The Western/Anglophone linguistic tradition, however, uses a different system, one that I used in the abovementioned post. Instead of one 3-way distinction, two independent, orthogonal 2-way distinctions are made: prefixes vs. suffixes and inflectional vs. derivational morphemes. Prefixes and suffixes in this system differ by the locus of attachment: before or after the root. Inflectional and derivational morphemes differ by their function: the latter serve to form new words, whereas the former serve to make new forms of words and express grammatical meanings. Thus, the Russian “suffixes” correspond to “derivational suffixes” in the Anglophone model, while “endings” are “inflectional suffixes”. (Both, however, can be referred to as simply “suffixes” in English, when their function is either unimportant or clear from the context.)

suffix_prefixOne reason I used the terminology from the Anglophone model in the post in question is not to confuse (the majority of) my readers, who are familiar with this model but not the Russian one. I do, however, also think that the Anglophone model is also more adequate than its Russian counterpart in that it clearly separates form (prefix vs. suffix) from function (inflection vs. derivation)—a distinction that is generally very important in the study of language. Moreover, the Anglophone model, which distinguishes four types of morphemes where the Russian system distinguishes three, is more adequate for the description of a broad range of languages and, ironically, of the Russian language itself. The Russian model does not distinguish inflectional and derivational morphemes attaching before the root, in the same way it distinguishes inflectional and derivational morphemes attaching after the root. From the cross-linguistic perspective, the Anglophone model allows to draw more precise distinctions between prefixes serving different functions. This allows for a better description of prefixing languages. Thus, one could hardly use the term “ending” (which in Russian, as in English, is cognate with “end”) to designate an inflectional prefix, for instance, in Swahili. Moreover, the Russian model appears to imply that the difference in function—derivation vs. inflection—is significant only in the case of suffixes but not of prefixes. There are no good reasons to think that this is true. Almost 300 languages in the 969-language WALS sample use prefixation to encode inflectional categories, including such Bantu languages as Zulu and Khosa, Guaraní in South America and Yaqui and Chipewyan in North America, Ket in Western Siberia, Berber languages in northwestern Africa, and many others (see the WALS map on the left).

When it comes to describing the Russian language itself, distinguishes “suffixes” from “endings” but using only one term for prefixes obscures the generalization that Russian, like English, places all the inflectional information after the root, whereas derivation can be done by a wider range of morphological means: both prefixes and suffixes, as well as an occasional infix, reduplication, backformation etc. This similarity between Russian and English is due, of course, to their common Indo-European roots.



Nesset, Tore (1998) Antiphonotactic Suffixation: Russian -sja and Optimality Theory. Scando-Slavica 44: 145-153.

Nesset, Tore (1998) Affiks eller klitikon? Norsk Lingvistisk Tidsskrift 16: 185-206.


Related Posts

Subscribe For Updates

We would love to have you back on Languages Of The World in the future. If you would like to receive updates of our newest posts, feel free to do so using any of your favorite methods below:

  • Ivan Derzhanski

    Well, the traditional Russian terminology obviously wasn’t made with Xhosa or Guaraní (or even Hebrew or Georgian) in mind; its purpose is to highlight the characteristics of the morphological makeup of Russian, so if that language has exactly three types of affixes, then using three terms for them makes good sense. But even in Russian things are more complicated, which is why accounts (even present-day school grammar, it seems) draw a distinction between _словообразовательные суффиксы_ and _формообразующие суффиксы_, the latter being different from _окончания_.

    • Ah yes, these “формообразующие суффиксы” complicates matters further. So neither the form nor the function per se determines the label for a morpheme… Also, the definition of окончание says that it “connects the word to other words in a sentence”, but there is a very peculiar view of what exactly that means. But even if we set the issue of “формообразующие суффиксы” AND cross-linguistic considerations aside, as I mentioned in the post, the terminological distinction between suffixes and endings conceals some interesting properties of Russian itself…

      • Ivan Derzhanski

        The truly outstanding thing about the ending is that, at least _prima facie_, every form of a word belonging to one of the inflecting parts of speech has exactly one. Prefixes and suffixes may be absent altogether, or there may be several of them, but there is always one ending (though it may be null), and it expresses, in portmanteau fashion, the values of several inflexional categories that determine the form’s place in the paradigm. I suppose this is where `connects the word to other words in a sentence’ comes from, since it is the ending that changes under government and agreement.

        • Both “connects the word to other words in a sentence” and “encodes the categories that determine the form’s place in the paradigm” don’t define the ending vs. формообразующий суффикс, at least not without other (not clearly warranted) assumptions. For example, comparative and superlative are considered to be part of an adjective’s paradigm, as are деепричастия for verbs. When it comes to connection with other words, the infinitival marker -ть certainly marks connection to other words, for example requirement for subject in the dative (or no overt subject at all).

          • Ivan Derzhanski

            Well, I’m not saying I’m happy with all this, I’m just trying to understand and explain what their reasoning may have been. Had they thought more about it, they might’ve realised that the appeal to ‘connecting the word to other words’ is problematic, and so is the appeal to the distinction between derivation and inflexion, because that distinction isn’t quite so sharp as school grammar teaches.

            «Рекомендую ребятам не удивляться, встретив новую для себя интерпретацию. Ничего не поделаешь, есть вопросы, относительно которых исследователи еще не пришли к согласию. Главное – быть последовательными и всегда комментировать спорные явления одинаково». (http://russkiy-na-5.ru/articles/158)

          • Great quote, yes.

  • What do you mean by “everyday linguistic jargon” (in Russian or in any other language)? The issues we are discussing here have to do with specialized linguistic terminology, so I don’t see what mass schooling might have to do with it. Except that in the last 70 years (when most people living today have been educated) there has been considerably more language-related education at elementary and secondary levels in Russian schools than in the Anglophone schools, or so it seems.

    • Sean Manning

      Well, most Anglophones can use terms like „noun“ „subject“ and „passive voice“ to talk about language even if their usage is fuzzy. This jargon had been inherited from Latin by the eighteenth century, and it is missing many of the refinements which Anglophone linguists use because it is older than Anglophone linguistics. You write “Back in early middle school, we had to use different notation for “suffixes” and “endings” when doing morphological analysis.” And you imply that Russian linguists use these same terms. So it sounds like Russians with a middling level of education use the same vocabulary to talk about morphology as Rusian linguists use, while Anglophones with the same level certainly do not.

      • Thanks for the clarification, Sean, I see your point now. But I don’t think this difference between the Russians’ use of linguistic jargon and the Anglophones’ has to do with when the mass education system was put in place. People who went through a Russian schooling system (such as myself) use those terms because we were taught them (yes, there was a lot of fuzziness and misinformation too), whereas in my experience most Anglophones my age or younger were not taught this stuff in school, certainly not to the extent that we have. But it has nothing to do with when mass schooling began because as recently as our parents’ generation in both “worlds” learned far more linguistic terminology and fundamentals than later generations have…

  • jenny_brn

    Dear Ms Pereltsvaig,
    Could you tell the gender of an author by reading a paragraph written by him or her? ( Even if he or she deliberately avoided grammatical constructs which could reveal his/her gender). I talk only about English here. I understand that many languages have first person gender specificities that give away the writers gender ( deliberate deception is not the question here).
    Thanks for the attention.

    • Yes, there are studies of differences among men and women in their use of language (spoken or written or both, I am not sure). Not my area of expertise, I am afraid.