From a morphological point of view, Slavic aspect is expressed derivationally. Aspectual verb pairs like Polish imperfective (IPV) robić and perfective (PV) zrobić ‘do’, or Czech PV odhalit and IPV odhalovat ‘reveal’, are formed using a number of derivational affixes, sometimes with slight changes to the verb stem, and in some few cases with suppletive forms. Thanks to this, the aspectual opposition permeates almost the entire verbal paradigm including participles, the infinitive, and the imperative. While the inventory of aspectual morphology is remarkably similar across Slavic languages, the way in which the grammemes [PV] and [IPV] are employed in certain domains (iteration, performative speech acts, Historical Present etc.) shows considerable inner-Slavic variation. Eckert (1984) and Stunová (1993) for example, compare Czech and Russian aspect, and Dickey (2000) compares aspect in all major Slavic languages for several phenomena (but not the imperative), concluding that Slavic aspect use can be divided roughly into an Eastern type (Russian, Ukrainian, Belarusian, Bulgarian and Macedonian) and a Western type (Czech, Slovak, Slovenian and the Sorbian languages), with Polish and Bosnian/Croatian/Serbian in a transitional zone.

Aspect use in the imperative has been described among others by Padučeva (22010), Lehmann (2008), Wiemer (2008), and by Benacchio (2010) in a comparative monograph comprising all major Slavic standard languages. In general, aspect use in the imperative follows what may be called „canonic“ aspect functions as they are described e.g. in the AG-80 or Lehmann (2009) for Russian or in Dickey (2000) for Slavic languages in general: the PV is used for achievements and accomplishments, the IPV for states and activities. Cf. the following PV examples:

The IPV is used also in open iterations, as in general advice, cf. the following:

Note that Czech and Russian differ in that Czech also allows for the PV in iteration, being able to focus on the perfective micro-event rather than the macro-level of iteration, whereas Russian allows the PV only in the so-called summary meaning (Stunová 1993, Dübbers 2015).

Finally, the IPV is also regularly used under negation, the negated PV being possible only in non-volitional contexts (cf. Wiemer 2001 or Lehmann 2009).

Up to this point, aspect usage in the imperative has not been very surprising. Let us now turn to a new set of examples.

In (5) and (6), Russian allows the use of the IPV aspect although the situation described by the verb is neither an activity, nor iterated, nor negated. The PV is just as possible in this context. The IPV is said to make the statement more soft and polite in (5), more rude in (6). These pragmatic effects of politeness/rudeness are the focus of Benacchio (2010) and are well known and described for Russian (cf. also Padučeva 22010, Wiemer 2008, Lehmann 2008). Both Padučeva and Benacchio also explain, in different ways, how the effects of positive politeness vs. rudeness arise in the situational context. Whether this pragmatic use of the IPV is also found in Polish or Czech is less clear. Eckert (1984) notes that the IPV is used in “certain standard etiquette forms of polite address” and also “to add politeness to an order expressed by verbs rendering a concrete movement” (139) in Russian, but not in Czech, that is to say, she acknowledges a pragmatic difference, but does not point out the possible rudeness of the IPV. According to Benacchio (2010), the positive-politeness effect is completely unavailable in Czech and the rudeness effect is also very limited, possibly exclusive to sub-standard language, while in Polish both are possible, but still more limited than in Russian. The exact nature of these limitations is not clear.

This is where this paper comes in. I conducted two studies to test Benacchio’s informant-based claims against corpora, more specifically, to find out whether Russian really uses the IPV imperative more than Polish, and Polish in turn more than Czech.

There has been a previous corpus study on aspect in the Slavic imperative by von Waldenfels (2012), who analysed 11 Slavic languages in his parallel corpus ParaSol. He calculated and visualised distances between the individual languages, however his study considered only whether languages differed or not for each imperative in the text, but not in which way (i.e. a Czech IPV corresponding to a Russian PV imperative was not distinguished from a Cz.PV-Ru.IPV pairing), so this is of little help here.

This study was conducted using Vladimír Benko’s Aranea webcorpora in Czech (Araneum Bohemicum Maius 15.04), Polish (Araneum Polonicum Maius 15.02) and Russian (Araneum Russicum Maius 15.02).

I extracted the frequencies of occurrence of IPV and PV non-negated imperatives by lemma, paired the aspectual partners together and calculated the percentage of how many imperatives of the given aspect pair are IPV. These are given in table 1. For reference, I added the percentage of IPV tokens for each aspectual pair from the entire lexemes. For the sake of brevity and for ease of comparison, I will discuss only the singular here: while Czech and Russian both have a simple T-V-distinction for formality comparable to French, Polish uses a system of address nouns (pan ʻsirʼ, pani ʻmadamʼ, państwo for mixed groups, among others) with third-person agreement. This means that a Czech 2pl imperative like dejte! ʻgive!ʼ can have several Polish equivalents, depending on who exactly is addressed, which complicates comparison between these languages.

The image is not as clear-cut as we might have hoped, but our predictions are at least confirmed. Consider, for example, the equivalents of English to sit down: Czech makes almost exclusive use of the PV while Polish and Russian use the IPV as well, with Russian even favoring it. The situation for to look (at) is very similar, and to a lesser extent for others as well. In some cases, Polish appears closer to Czech than Russian, in other cases it even uses less IPV than Czech does. For some verbs, there is no discernible difference between languages (e.g. to allow, to stop, to try), while for to help and to ask Czech surprisingly has the highest percentage of IPV imperatives.

The high percentage of Czech IPV dávat ʻgiveʼ surprises at first glance, but about half of these cases (745 out of 1423) are part of the phraseologism dávat (si) pozor/dávat (si) bacha ʻto be carefulʼ. In these cases the IPV is perfectly natural.

Table 1: Percentages of IPV partner verbs in singular imperatives vs. for all wordforms.

This leads us to an important point: the respective verbs in one row of the above table are of course not perfect equivalents. To illustrate this on to give, Czech PV dát is often used as ʻto putʼ as in dej to na stůl ʻput this on the tableʼ, and the Russian IPV imperative davaj/davajte is often used with an exhortative meaning, to such an extent that it can be described as a particle meaning ʻcome on!ʼ. While it is important to keep such differences in mind, the general point still stands: Russian is more prone to using the IPV in a pragmatic way and should thus have a higher percentage of IPV imperatives. The very fact that the IPV davaj and not the PV daj developed into an exhortative particle is, I believe, a testament to that.

For the second corpus study, I used Alexandr Rosen’s parallel corpus InterCorp (v9). I looked for non-negated Czech PV imperatives with an IPV imperative as a Russian equivalent, and vice versa, in both singular and plural. Note that singular and plural are strictly morphological categories here and that both the Czech and the Russian plural covers informal address of a group as well as formal address of individuals or groups. Including Polish with its more complex (pro)nominal system of formal address in this second study was beyond the scope of this paper.

If the pragmatic use of the IPV imperative is in fact much more restricted in Czech, as Benacchio (2010) claims, then there should be more Cz.PV-Ru.IPV correspondences than the other way round. Because translations are aligned by sentence in InterCorp, I went through the results manually to remove any mistakes, e.g. where a Russian IPV imperative just happens to appear in a sentence but is not, in fact, a translation or equivalent of a Czech PV imperative. Figure 1 shows the results for both singular and plural pairings of Cz.PV-Ru.IPV and Cz.IPV-Ru.PV.

As expected, in both singular and plural the pairing Cz.PV-Ru.IPV is more frequent than Cz.IPV-Ru.PV, possibly due to the more widespread use of the pragmatic IPV in Russian. In the following two examples, the Russian IPV can in fact be interpreted pragmatically as “urging”:

One might also expect that the observed asymmetry between Czech and Russian is due to the fact that Czech also allows PV in iterations, which Russian disprefers. However, iterative examples are in fact very rare in our sample. One example of this is given in (9):

The most frequent Czech PV imperatives translated using a Russian IPV are (singular only): podívej ‘look’ (139), posaď (se) ‘sit down’ (28), poslechni ‘listen’ (24), vrať (se) ‘return’ (24), zůstaň ‘remain’ (21), rozděl ‘divide’ (19)2, sedni (si) ‘sit down’ (20), spusť ‘start, get going’ (19), odpověz ‘answer’ (15), chyť ‘grab’ (14).

The most frequent Czech IPV imperatives to be translated with a Russian PV are (again, only singular): poslouchej ‘listen’ (42), pojď3 ‘come’ (37), pamatuj ‘remember’ (35), jdi ‘go’ (30), věř ‘believe’ (23), běž ‘run’ (21), drž ‘hold’ (19), mlč ‘be silent’ (18), povídej ‘tell’ (17), snaž (se) ‘try’ (13).

One can ask, of course, why there are any Cz.IPV-Ru.PV pairings at all. When we look at the Czech lexemes in question here, we find that a quarter of these cases belong to “partnerless” IPV verbs and do not therefore participate in the aspectual opposition. These are motion verbs4 like jít ‘go’, běžet ‘run’, but also vyprávět ‘tell (a story)’ and držet ‘hold’5. We can speculate that maybe they would express their imperative in a PV form if they could. Figure 2 is an update of Figure 1, with the added column showing Cz.IPV-Ru.PV pairings with these partnerless verbs removed, which makes the asymmetry even more pronounced.

Regarding Czech IPV poslouchej ʻlistenʼ; I believe that in these cases listening is seen as an atelic “state of paying attention”, hence the IPV. The fact that they correspond to PV Russian poslušaj is due to the nature of the Russian prefix po-, which can convey a delimitative meaning of ʻdoing s.th. for some timeʼ (cf. AG-80:365), combining perfectivity and atelicity, whereas Czech does not have this option and thus by necessity uses the IPV to convey atelicity. When combined with a concrete object, the listening becomes telic and the PV becomes the preferred choice in Czech as well.

Czech and Russian clearly show an asymmetry in the way they use verbal aspect in the imperative, confirming our prediction based on Benacchio’s claims. The first study, which also included Polish, has shown that in most cases the relative frequency of the IPV imperative is higher in Russian than it is in Czech and Polish. In the second study we have seen that Cz.IPV-Ru.PV pairings are less frequent than Cz.PV-Ru.IPV, which points us in the same direction: Russian uses the IPV imperative more often than Czech does. While the immediate context (phraseologisms) and lexical idiosyncrasies (missing partner verbs, language-specific additional meanings of a given verb form) certainly play a role as well, this asymmetry between Czech and Russian is at least partially due to a difference in pragmatics between the two, namely the widespread use of the pragmatically-motivated „polite“ or „rude“ IPV imperative in Russian.