Foreign learners are often taught that Russian verbs always come in aspectual pairs, with the perfective verb formed from the imperfective verb either by changing the stem or adding a prefix. The two verbs are supposed to be synonymous in every way, except that one is perfective and the other is imperfective. If the perfective verb is formed from the imperfective verb by adding a prefix, the fact that the two are synonymous means that the prefix must be "semantically empty", i.e. have no meaning or effect on the meaning of the word, and just denote that the verb is perfective.

Question: Do such verbs actually exist? In other words, are there perfective verbs formed by adding a prefix to an imperfective verb, for which the prefix does not mean anything?

One example that is commonly given is писать and написать. However, some authors argue that even in these cases the prefix isn't actually semantically neutral, but that it "overlaps" with the meaning of the original imperfective verb in such a way that it de facto doesn't affect the meaning. A monograph detailing this hypothesis can be found here or here.

So ideally an answer to this question would in part address the answerer's opinion about the validity of the "overlap hypothesis". I.e. whether it makes more sense as a learner to focus on "aspectual pairs" as opposed to more general "aspectual clusters" (for which there might not be a one-to-one correspondence between imperfective and perfective verbs).

Background: (totally unnecessary to read, unless you're really curious)
Thinking about this question was motivated by this answer to another question.

That answer mentions biaspectual verbs, which challenge to some extent the idea of Russian verbs as always belong to aspectual pairs. On the other hand, most research shows that the majority of biaspectual verbs are foreign borrowings, so the argument could be made that most simply have not been present in the language long enough to develop full aspectual characteristics.

However, what seems to me even more problematic to the aspectual partner model are imperfective verbs which don't have a single "all-encompassing" perfective partner. Instead there exist multiple perfective versions, each formed by adding a different prefix, and with a different meaning. I have asked at least two questions already on this site about these verbs, here (курить) and here (смеяться), which also motivates in large part this question.

Such verbs can be interpreted somewhat so as to rectify the aspectual partner model, namely by interpreting the imperfective verb as actually being biaspectual, and forming imperfectivizations from each of its various hyponymous perfectivizations, and interpreting each resulting imperfective/perfective pair as a derived aspectual pairs. However, as with the examples of посмеиваться and засмеиваться, this interpretation seems somewhat forced, since the imperfectivizations are often rarely to never used or don't have the same original meaning as the perfective verbs they were supposed to be derived from. See also this question.

The linguist Janda writes, for example here, that Russian verbs are more naturally interpreted as belonging to "aspectual clusters" than "aspectual partnerships", in contradiction to the grammar usually taught to foreign learners.

The biggest seeming counterexample to such an idea are those imperfective verbs who, unlike the examples of курить and смеяться mentioned above, form perfective verbs using so-called "empty prefixes", i.e. prefixes devoid of any semantic content, but just signal that the verb is to have the perfective aspect instead of the imperfective aspect. For example, писать and написать. Such examples seem to also be the greatest motivation behind the "aspectual pair" model usually taught to foreign learners, since the model works most neatly and cleanly for such verbs.

In order to explain this, Janda and others have advanced the "overlap hypothesis", which claims that most, if not all, of so-called "empty prefixes" for perfectivizing verbs actually do have semantic content -- it is just that the semantic content of the prefix "overlaps" with the meaning of the original imperfective verb in such a way that it does not restrict the meaning of the resulting perfective verb, i.e. leads to a synonymous rather than hyponymous verb, despite not being semantically void/empty. They have written about this in a monograph explained in detail here, with some explanation of the methodology found here. Other sources explaining the hypothesis: (1)(2)(3).

More related questions: (1)(2)

  • 1
    My personal idea is that there is no 100% neutral or 100% specific prefixes/suffixes. Some of them feel more generic than others, some are less, but with maybe few exceptions there is no 0% or 100% black and white. So, learning "aspectual pairs" may be a good intermediate step, to convert "English-thinking" persons to "Russian-thinking". To help them getting the gist of it using few selected verbs, which are close to 100% neutral. But it would be fleeting and transitional. As soon as they get a gist, it would better to immediately switch to learning clusters. – Arioch May 30 '17 at 15:22
  • 1
    Granted, English perfect tenses also do not seem to me truly semantically neutral. In 1980-s school, we were taught that Perfect sense is used when something is done and "the result of the deed is obvious". Whatever that should mean. But another teacher about 5 years ago told me different perspective, she said that when something is merely done (complete, perfected) it should be conveyed by Definite/Simple tense, and Perfect tenses are only used, when the completion of an activity is explicitly tagged with some, usually explicit, timeframe. No time tag => no Perfect tense. Not 100% neutral. – Arioch May 30 '17 at 15:27

Complicated question. I read, that linguists even haven different opinion on whether it is correct to consider perfective and imperfective verbs as different forms of the same verb or as different verbs with some similar meaning...

I think for beginner learners "aspectual pairs" are much more relevant. They need to understand situations in which they use this or that aspect, as well as to learn a big list of new words (perfectives). If you offer them "clusters" it will just make too much mess for them.

Although, for some verbs (from my point of view) there is sense to learn several (but not all) variants from the beginning (for example: говорить - поговорить/сказать), as they are quite common: Я поговорил с врачом. Он сказал, что теперь я здоров. Or "я уже поел" vs. "я съел пиццу".

Yes, prefixes often have some meaning, but we often have other imperfective pairs for these ones. писать - написать, приписывать - приписать, переписывать - переписать, дописывать - дописать, выписывать - выписать, вписывать - вписать, записывать - записать etc.

Clusters are good for advanced learners to better understand these different meanings, to make their speech richer. For beginners it will just be a shock, while in most situations they can express their idea with just писать or написать.


Well, first of all, there is a bunch of dual-aspect verbs, i.e. perfective and imperfective at the same time. See, for example, wiktionary:Двувидовые глаголы. They are prefixless, though.

Next, there are some perfective verbs which do not add anything to the meaning of its imperfective pair, except the completeness of the action. The most obvious example, in my opinion, is делать - сделать (and, yes, с is a prefix here).

But the existence (or non-existence) of such "neutral" verb doesn't change anything about the ability to produce "specialized" perfective versions. Say, поделать and заделать are nothing worse than покурить and закурить.

So the answer is yes, if we really need an "empty" perfective verb then some of its prefixed form may take such part as well. But "an aspectual cluster", or whatever you may call it, still exists, while "an aspectual pair", in my opinion, might only be a special case of it.

In order to explain this, Janda and others have advanced the "overlap hypothesis", which claims that most, if not all, of so-called "empty prefixes" for perfectivizing verbs actually do have semantic content

Yes, that seems reasonable, as an "empty" prefix may also be "non-empty" when "pasted" to a different verb.

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.