SLAVIA časopis pro slovanskou filologii

ročník 86, 2017, sešit 2–3

Slavic Resultatives and Their Extensions:

Integration into the Aspect System and the Role of Telicity

Björn Wiemer (Mainz)

Resultatives represent clusters of temporal features (a state resulting from a change of state)
which require telic stems as lexical input. In Slavic, all resultative constructions are based on
participles. Resultatives often turn into perfects and passives. Most essential in this develop-
ment is the extension of admissible lexical input to the resultative construction, by which, con-
comitantly, the requirement that the verb stem be telic is lost. Simultaneously, Slavic participles
distinguish perfective (pfv.) and imperfective (ipfv.) aspect. As a grammatical category, aspect
is not restricted (or defined) by telicity, although telicity was a factor motivating the rise of the
pfv.:ipfv.-opposition, and the association between telic events and pfv. aspect is very close.
Considering this, the question whether ipfv. participles have been, or are, used in constructions
other than resultatives needs to be investigated. We should ask whether the presumably origi-
nal (i.e. Common Slavic) resultative function has been preserved by ipfv. participles, or wheth-
er they have participated in a perfect or, alternatively, in a canonical passive or some similar
construction operating on voice. This paper sheds light on these questions. It first provides a
survey of resultatives and perfects in Slavic through space and time, asking for the relation
between telicity and the development of the pfv.:ipfv. distinction.
Ipfv. participles used in constructions that developed out of resultatives behave in one of two
ways: Either ipfv. participles show the same resultative value as their pfv. counterparts, a
behaviour which can be considered an archaism prior to the strengthening of the aspect oppo-
sition, or, on the contrary, the aspect of the participles rather reliably restricts the range of
functions which are known for the respective Slavic variety in the active voice. In these cases,
ipfv. participles in predicative use have been integrated into the sets of functions that generally
govern the choice of aspect in the particular Slavic variety. This dominance of aspect applies
even if telicity constrains the choice of the aspect of the participle or the applicability of the
entire construction.
Keywords: resultatives / perfects, aspect, telicity, diachronic morphosyntax, inner-Slavic distri-
bution

The opposition of perfective and imperfective (pfv. : ipfv.) aspect is a pervasive feature
of all Slavic languages (standard and non-standard varieties). Since this grammatical
opposition is based on stem derivation and has established a binary classificatory
system, every verb form is affected, including participles (and other non-finite forms).
Participles, in turn, are the core elements of resultatives and perfects in Slavic, and
most participles involved in Slavic resultatives or perfects today are certainly older than
the onset of the development of the pfv.:ipfv. opposition. This raises the question how
resultatives and perfects interfere with aspect, more particularly: To which extent have
ipfv. participles been involved in the formation of originally resultative constructions?
Have ipfv. participles become marginal or have they been abandoned altogether? Or
have they been integrated into other constructions for which resultatives provided the
basis? That is, if they have “survived”, have they contributed to the deterioration of
resultatives? In particular, to which extent have ipfv. participles contributed to the
formation of perfects or passives?
 Slavic Resultatives and Their Extensions 125

Compare the following examples with ipfv. participles, whose language-specific

status is rather different:
colloquial Czech
(1) M-á plací-n-o. (waiter about a client; cited from Giger 2003, 89)
have-PRS.3SG pay[IPFV]-PP-N
‘S/He has paid.ʼ
 resultative
standard (colloquial) Macedonian
(2) Во свој-а-та кариера има снима-н-о
in PREFL-F.SG-ART.SG.F career[F] have(PRS.3SG) produce[IPFV]-PP-N
неколку филм-ови
several film[M]-PL
но сепак најпознат е по улогата на адвокатот Герет Велс во серијата “Бостонско
право”.
‘In his career he has acted in many films [lit. shot several films], but he is most
known for his role as the lawyer Gereth Wells in the series “Boston Law”.’
(www.tocka.com.mk/index20.php?ID=2147; cited from Mitkovska – Bužarovska
2011b, §3.4.3.)
 experiential perfect
Russian, Pskov-Novgorod dialects
(3) Пeчк-a дaвнo нe тoпи-вши. (cited from Ryko 2002, 190)
oven[F]-NOM.SG long.ago NEG heat[IPFV]-ANTP
‘The oven hasn’t been heated for a long time.ʼ
 experiential or resultative?
standard Polish
(4) Drzw-i właśnie są zamyka-n-e.
door-NOM.PL just be-PRS.3PL close[IPFV]-PP-PL.NVIR
‘The door is just being closed.ʼ
 dynamic passive (no resultative or perfect) in progressive or iterative function

Example (1) from colloquial Czech has a resultative reading, the ipfv. participle shares
this meaning with its pfv. counterpart; example (2) from colloquial Macedonian can
rather be considered an experiential perfect since the resultative component has been
weakened. In both (1–2) the participle combines with a HAVE-auxiliary. By contrast,
the ipfv. participles in (3–4) are accompanied by (overt or zero) BE-verbs; while in the
example from dialectal Russian the participle has another suffix, the participle in stan-
dard Polish is identical with the suffixes in the Czech and Macedonian examples.
However, only the Russian dialectal example presents us with a resultative or perfect
meaning, whereas the ipfv. n/t-participle of standard Polish can be considered neither
a resultative, nor a perfect, it can only be read as a canonical (dynamic) passive. These
examples give us an approximate idea of the range of functions from the domain of
resultative / perfects and their interference with passives involved with predicative ipfv.
participles in diverse Slavic varieties.
126 Björn Wiemer

I will start with a brief discussion of key concepts which supply the background
for the subsequent investigation (§1.). I will then browse through the morphosyntactic
types of resultatives / perfects attested in Slavic (§2.). The questions related to ipfv. input
will be examined in §3. The last section (§4.) contains a resumption and conclusions.
Due to spatial constraints, only few examples are glossed fully; zero morphemes
are never indicated (though maybe glossed). However the aspect of participles and
morphosyntactic information necessary to understand alignment is always indicated; if
not indicated in separate glosses, aspect is marked with upper case small capitals ( PFV,
IPFV
). For non-standard abbreviations see the list in front of the reference section.
Examples are quoted in accordance with the orthographic conventions of the sources.

1. The core notions

I begin with the relation between resultatives, perfects and passives (§1.1.), then I com-
ment on the relation between aspect, resultativity and telicity (§1.2.) and present some
leading questions (§1.3.).

1.1. Resultative and perfect(s)

Probably, the notion of ‘perfect’ is amongst the most heterogeneous terms in typology,
although certain core domains (or usage types) can be identified and have at least been
discussed in the linguistic literature since Dahl (1985); cf. Plungjan (2016) for a recent
critical evaluation. These are: (i) resultative, (ii) experiential, (iii) current relevance, (iv)
hot news. From this list we see that, usually, the resultative is regarded as a subtype of
the perfect. In fact, it can be regarded as one of its two main subtypes, the other one
being the experiential function. In a sense, between these two main focal points in the
perfect domain we find the notion of current relevance (CR), which has even been
considered as the central, though particularly vague, function of perfects.1) The vague-
ness of CR has been paraphrased as “the speaker’s presentation of the consequences of
an event as essential to the point of the speaker’s contribution or that the event has
some repercussions on the participants of the discourse situation” (Bužarovska –
Mitkovska 2010, §3., following Dahl – Hedin 2000). CR is often difficult to distinguish
from the experiential function. One can narrow down the notion of the latter as imply-
ing a possible agent, as do Bužarovska – Mitkovska (2010, §3), and say that an experi-
ential attributes certain qualities or knowledge to an agent due to past experience
(Bybee et al. 1994, 62).2) However, the notion of agency is weak in utterances that
count as typical examples of experientials, such as (5a). Moreover, the notion of expe-
rientiality practically coincides with the notion of ‘general-factual meaningʼ (Russ.
общефактическое значение) known from Slavic, in particular Russian aspectology; see
(5b), which is a literal translation of (5a) and the verb is in the ipfv. aspect:

1/ The “hot news” perfect can quite easily be calculated from the other perfect meanings. In
this paper, it will not play any significant role.
2/ The so-called perfect of persistence, or anterior continuing, can be considered an extension,
or a subtype, of the experiential (I’ve been waiting for your reply for two months). As with the “hot news”
perfect, its discussion would add nothing to the points being made in this article.
 Slavic Resultatives and Their Extensions 127

(5a) Engl. Have you ever X-ed?

(5b) Russ. Ты когда-либо Х-ова-лIPFV?
X = any verb

For the time being, I will not distinguish between CR and experiential perfect and
consider the experiential function as largely synonymous with the general-factual func-
tion. I am aware that this simplification looks unusual considering common practice,
which is based mainly only on the behaviour of the English Present Perfect.3) However,
one may wish to differentiate between CR and the experiential function, they cluster
in clear opposition to the resultative function (see §2.1.), and the latter diachronically
serves as the basis for the former ones (see below). At this point I should emphasize that
I use the notion of ‘resultative’ exclusively in the sense defined by N e d j a l k o v (1983
[1988]). It does not include the more specific meaning of ‘perfect of result’, which
applies if the predicate denotes an event whose concrete consequences are felt at the
very moment of the utterance (e.g., Engl. Esther has opened the door, so that now it’s very
cold inside). This function corresponds to the Russian term перфектное значение, which
is tied closely to the perfective past (see §2.1.). Contrary to the resultative, the perfect
of result is event-oriented. In English this contrast can be expressed by the choice
between a copular ( resultative) and an auxiliary verb ( perfect of result), e.g. He is
gone vs. He has gone, or in marked voice: The door is closed (= resultative present) vs. The
door has been closed (= passive of Present Perfect). In most Slavic varieties we do not find
such morphological contrasts, in particular passive and object-oriented resultative usu-
ally cannot be distinguished morphologically. See the English translations of many
examples in this article. In the following I will use ‘perfectʼ as a cover term for all func-
tions ascribed to grams in the aforementioned domain, while ‘resultativeʼ will be
reserved for constructions which only have the resultative function.
All functions of perfects are united by the fact that they highlight the relation of
some event to a reference interval which, by default, includes the moment of the utter-
ance (Dahl 1985, 133). The differences between the diverse aforementioned perfect
meanings concern the focus of assertion (see below). In this sense, all perfect functions
are instances of temporal deixis. The reference interval of the state may be moved into
the past or the future (anterior or posterior to the moment of the utterance), but I will
not dwell on these switches in particular. Another point is important: In other functions
than the resultative one perfects switch their focus to an event that precedes a resultant
state to which this event is associated. Notably, this is not a switch to the past or to the
future, which would imply that the relation between reference and event time is moved
in toto from the deictic moment of speech; instead, the reference interval “remains”
where it is but its relation to the event time changes (compare 6a vs. 6b). A resultative
construction describes the subsequent state obtained after a state changing event, and
this state is simultaneous to some reference interval (= utterance time, by default). It is
essential that the state is understood as resulting from this event, otherwise we would
be dealing with a stative (Nedjalkov – Jaxontov 1983 [1988]). The difference becomes
clear by comparing (6b) and (6c) in standard Russian:

3/ Cf. Mittwoch (2008). I am obliged to Peter Arkadiev for pointing this out to me.
128 Björn Wiemer

(6а) На балкон-e повеси-л-и бель-e.

on balcony-LOC hang.up[PFV]-PST-PL linen-ACC.SG
‘They hang up the linen on the balcony.ʼ
 eventive perfect: focus on change of state, resultant state only implied
(6b) На балкон-e повеше-н-o бель-e.
on balcony-LOC hang.up[PFV]-PP-N linen-ACC.SG
‘The linen is hung up on the balcony.ʼ
 resultative: focus on resultant state, change of state presupposed
(6c) Нa ветв-ях вис-ят (? повешены) грозд-и виноград-a.
on branch-LOC.PL hang[IPFV]-PRS.3PL bunch-NOM.PL grape-GEN.SG
‘Bunches of grape are hanging on the branches.’
 stative: no salient relation to a change of state
If the focus is on the event, I will henceforth speak of an eventive perfect (following
Litvinov – Nedjalkov 1988). The eventive perfect and the resultative4) are related to
each other like cause and effect, since they focus on different parts of a complex situa-
tion type (Breu 1988; Wiemer – Giger 2005, 2).
We see that the resultative is a complex notion which combines three components:
(i) It asserts a state that (ii) presupposes a preceding state-changing event, and it (iii) can
only be derived from a telic lexeme, since otherwise there would not be a change of
state. In this paper I use ‘telic’ in the proper sense as a lexical property saying that there
is an inherent endpoint, or goal, of the eventuality marked by the respective verb stem.
The lexeme itself does not require that this endpoint be attained. Ipfv. telic stems can
mark goal-directed processes, but the goal may not be reached; compare the ipfv.
equivalent of Russ. (6a) above: Соседка вешалаIPFV белье ‘The neighbor was hanging up
the linenʼ. It is the resultative construction (6b) which presupposes that this goal has
been reached and that the resultant state persists; that is, the element to which this
semantics is most tightly associated with is the participle of the telic verb.
The perfect can be regarded as an extension of the resultative in the sense that
lexical restrictions are loosened, i.e. it is not only telic verb lexemes that can serve as
input to the construction. If condition (i) loosens, a resultative construction may shift
to an eventive perfect interpretation (i.e. the event is no longer presupposed, but
asserted), or towards a CR or experiential function (i.e. there is no state, but the utter-
ance interval as reference point is still prominent).5) Eventually the construction’s
original function is lost completely once it can describe any sort of eventuality, among
others homogeneous activities. This construction still bears a perfect meaning since it
relates some anterior situation to the deictic reference interval of the moment of utter-
ance (as, e.g., in She has already walked through Hyde Park). Furthermore, if the deictic
requirement is dropped, i.e. the reference interval no longer needs to coincide with the
time of the utterance but instead can refer to any reference interval (e.g. anchored in
some narrative discourse) and the construction itself refers to the event, not to a subse-

4/ An alternative pair of terms is ‘actionalʼ vs. ‘statal perfectʼ (Maslov 1983 [1988]).
5/ As L i n d s t e d t (2000, 368) pointed out, an increase in atelic lexemes is concomitant with a
shift from ‘current result’ to ‘current relevance’, i.e. the specific causal connection to the preceding
event becomes looser.
 Slavic Resultatives and Their Extensions 129

quent state, then, a perfect eventually ceases to be a perfect and turns into a general
past (Breu 1988; Thieroff 2000, among many others). This is what happened to prac-
tically all language varieties on the European continent roughly between the North Sea
and the Baltic Sea in the North and the Alps and Carpaths in the South (Breu 1994,
56–58, compare also Thieroff’s 2000 ‘present anteriorʼ). We should realize that this
shift took place at a very early stage already in connection with the oldest kind of per-
fect formation in Slavic, built on the l-participle (see §2.1.), in the North Slavic area. It
is very likely that this shift occurred earlier than in Germanic and Romance HAVE- and
BE-based perfects (Arkadiev – Wiemer, forthcoming: §3.5.1.).
There is one more tenet to keep in mind. Resultatives and perfects are neutral as
for voice, although they often happen to be closely associated to voice-related opera-
tions (Nedjalkov – Jaxontov 1983, 13f., 31f.). Particularly close is the relation of object-
-oriented resultatives (ObRes) to the passive; they change the alignment of the verb
stem from which they are derived: the more patientlike argument coded as object with
the morphologically unmarked verb becomes the subject of the resultative (e.g., X writes
a letter A letter is written). If no such change occurs, we are dealing with a subject-
-oriented resultative (SubRes; e.g., X drinks three beers  X is drunk), in accordance with
a conventional distinction introduced by Nedjalkov – Jaxontov (1983). The subject is
defined as the NP triggering number (sometimes also gender and/or person) agreement
on the predicate of its clause. This kind of NP is a particular instantiation of a
Privileged Syntactic Argument (PSA). This definition is sufficient for Slavic languages
with their predominant NOM-ACC patterns. A typologically more comprehensive
definition should, of course, use notions such as A(ctor)—U(ndergoer) contrasts as its
basis.
We will return to the relation between resultatives and passives later in this text.
I will largely delimit myself to resultatives on the level of present tense (see ex. 6b).
Other grams related to the perfect (like the pluperfect, futurum exactum) display pecu-
liarities which often cannot easily be explained only by a switch from the present to the
past (or future) level (cf. Arkadiev – Wiemer, forthcoming: §3.6.), but they are practi-
cally irrelevant for the central concern of this paper which is the interaction of the
perfects with the aspect opposition.

1.2. Telicity: its different significance for resultatives and aspect

In a sense, classificatory aspect (pfv. : ipfv.) and resultative perfects are in an orthogonal
relation to one another. On the one hand, there is an association between a resultant
state and the completion of an action which, in turn, is implied by a telic verb (lexeme).
Completion and telicity are not defining notions of pfv. aspect, but the association
mentioned in the previous sentence creates a natural intersection of pfv. aspect and the
meaning of resultatives. It is therefore not surprising that the bulk of participles that
can be found in resultative constructions are of pfv. aspect. On the other hand, we may
inquire in how far participles that are relevant for resultative (or perfect) constructions
have participated in the division of verb stems into two grammatical classes (pfv. : ipfv.).
Has their membership in either of these grammatical classes not been dominated by
lexical features “inherited” from a period prior to the development of the pfv. : ipfv.
130 Björn Wiemer

opposition (see §3.)? The decisive lexical property is telicity. It was already introduced
in §1.1. as an inherent feature that sets a natural limit to the situation denoted by the
verb stem. However, while the lexical stem itself does not imply that this limit is met
(and a change of state occurs), the participial form (which is a stem extended by a suf-
fix) does imply exactly this by default. The telic feature – and the default implication
of the extended participle stem that the “telos” has been attained – exists prior to the
division of verb stems into pfv. or ipfv. Aspect and telicity are independent from each
other, although telicity played a significant role in the rise of grammatical aspect
(cf. Wiemer – Seržant, forthcoming, for a recent synthesis). These considerations lead
us to the discussion of some central questions.

1.3. Leading questions

The following questions need to be raised: (i) Has the telic feature been retained by
participles of ipfv. verbs? If yes, has a focus on the accomplishment of the “telos” been
preserved (as it has been with their pfv. counterparts)? (ii) Have ipfv. participles of
other, atelic, verbs been involved in constructions that resemble, or have developed
from, resultatives? If yes, have those constructions been turned into perfects or pas-
sives? (iii) Do constructions with predicative ipfv. participles have functions of ipfv.
aspect that are typical of the given variety in the morphologically unmarked active
voice? In other words: Do relevant constructions with ipfv. participles adhere to the
usual “canon” of aspect functions of their Slavic variety, or do they deviate from it?

2. Perfects and resultatives in Slavic: basic morphosyntactic classification

Throughout its history, Slavic has known and still knows three basic construction types
of perfects. All of them are based on participles in different combinations either with a
BE- or a HAVE-verb:

(A) BE + l-participle
(B) (B.a) BE + n/t-participle; (B.b) BE + vši-participle
(C) HAVE + n/t-participle.6)

The l-participle has also intruded to HAVE-based constructions (= type C); see §2.1.
Henceforth I will relate to the BE- and HAVE-verb as an auxiliary, not a copula,
although it might be argued that in the initial stages of resultatives the construction
consists of a loose connection of a participle, as a nominal predicate, and a copular
verb. It is difficult to determine when auxiliation starts, i.e. when this loose connection
begins to turn into a verb complex (for this process cf. Kuteva 2001); erosion or loss of
agreement marking are not always reliable indicators. In general, it is advisable to tell
apart structural (i.e. morphosyntactic) features from semantic changes. There is
attestation of cases in which the syntactic reanalysis of a HAVE-based resultative or the
loss of agreement can be observed without a drift from a resultative (or even a stative)
6/ HAVE + vši-participle has been attested exclusively in dialectal Belarusian in the border
region to Lithuanian (Erker 2015, 96), being a clear calque from the latter (cf. Wiemer 2012 for this
typologically rare construction). Such examples are extremely rare, anyway, and only examples with
pfv. participles are known.
 Slavic Resultatives and Their Extensions 131

interpretation to an eventive perfect interpretation. Vice versa, there are known cases in
which agreement was retained even after a shift towards eventive perfect had occured.
A lack of consistent covariation in this respect has been noted, in particular, for
Romance varieties (cf. Giger 2003, 283–290, 404–407 with reference to Loporcaro
1995 and 1998). After all, erosion and loss of agreement marking are not directly
related to the central question of how resultatives or perfects interact with the aspect
functions of the verb stem. Thus, they are only of secondary, if any, importance for our
considerations.
The types under (A-C) are distributed over modern Slavic very unevenly. Types
(A) and (B.a) were inherited by all Slavic varieties from Common Slavic; accordingly,
they are ubiquitous in all Slavic languages, although they have since developed into
different directions (see §2.1.). Types (B.b) and (C) are much younger, their areal distri-
bution is much more restricted.

2.1. Type A: l-participle

The l-participle type (A) is based on was a Common Slavic innovation (shared with
Armenian and Tocharian). We have to distinguish between the original nominal (so-
-called short) forms, commonplace in Slavic, and the pronominal (so-called long) forms
which developed later (see fn. 7). The latter ones can be used as NP-internal modifiers,
and they are distributed unevenly among Slavic languages (see below). I start with the
nominal forms, which are represented only by the nominative (with gender and num-
ber distinctions). In North Slavic it underwent all stages from exclusively resultative to
general past, although the assumed initial (i.e. resultative) stage can only be recon-
structed. Presumably, the perfect based on the l-participle showed an initial restriction
to one-place verbs (intransitives) and was resultative from the start, which means that
it must have been restricted to telic stems. Both restrictions, to telic and to one-place
stems, were loosened prior to the first written documents, since in Old Church
Slavonic we already find them with verb lexemes of all kinds of valency and actional-
ity (Trost 1972, 93–101; Drinka 2017, 85f., 297).
It is certainly debatable, but also unnecessary, to determine the onset and the
completion of the process whereby the perfect turned into a general past practically in
the entire North (West + East) Slavic area and in Slovene. As is known, this process is
still ongoing in the Serbian-Croatian territory, while Macedonian and Bulgarian are
conservative in this respect. In Bulgarian and standard Macedonian the l-perfect is still
basically a perfect (despite its extension to indirect evidentiality), whereas in West
Macedonian dialects, especially in the region of Lake Ohrid, the l-form has practically
been ousted as a (resultative or experiential) perfect by two other forms (see §§2.2.1.4.,
2.3.–4.) and marks only indirect (i.e. reportive or inferential) evidentiality and mirativ-
ity. Neither for the perfect nor for its evidential extension do we observe any restric-
tions regarding lexical input, actional classes, or aspect.
Exactly for this reason it is intriguing to examine how pfv. and ipfv. stems combine
in Balkan Slavic perfect. If the l-perfect does not restrict the choice of pfv. or ipfv.
aspect, does aspect choice correspond to the distribution of functions known from
other Slavic languages in which the former l-perfect has become a general past? In
132 Björn Wiemer

North Slavic, pfv. telic verbs in the past can have a resultative entailment (compare the
already mentioned перфектное значение), e.g., Pol. Jan wykopał PFV ziemniaki ‘Jan has dug
out potatoesʼ. By contrast, ipfv. verbs (both telic and atelic) in the past are associated
with the general-factual ( experiential) function, e.g. Pol. Jan kiedyś kopał IPFV ziemniaki
 ‘Jan has already dealt with digging out potatoesʼ. This distribution is clear-cut (of
course, under the default premise of deictic time reference). According to standard
descriptions, Bulgarian basically shows the same distribution of pfv. vs. ipfv. stems in
the l-perfect: Ipfv. verbs are used if an anterior situation evokes some loose relevance
for the current moment of speech, whereas pfv. verbs bear a resultative entailment
(Maslov 1956, 238–240; Breu 1988; 1998, 97f.). Moreover, the distribution of pfv. and
ipfv. verbs is roughly the same in the aorist; compare examples cited from Maslov
(1956, 234f.):
pfv. aorist / resultative function
(7)  Ти ли си?  повтор-и т-oй.  Защо дойд-e?
2SG.NOM Q be-PRS.2SG repeat[PFV]-AOR.3SG 3-SG.M why come[PFV]-AOR.2SG
‘ It’s you?, he repeated.  Why have you come?’
ipfv. aorist / general-factual function
(8) И никакв-о oгъван-e?
PTC NEG.PRON-N deflection-NSG
Никакв-o! С миркометър-a мери-xмe.
NEG.PRON-N with micrometer-DEF.M measure[IPFV]-AOR.1PL
‘ And no deflection?  No one! We have checked it with a micrometer.ʼ
Admittedly, there might be a difference between general-factual and experiential
meaning, due to which the general-factual goes with the aorist and the experiential
with the l-perfect (Breu 1988). However, the crucial point is that ipfv. verbs dominate
in both the general-factual and the experiential function, while with the resultative
function we observe a clear preference for pfv. verbs.
Practically the same applies for standard Macedonian. The l-perfect is well repre-
sented in the experiential function; in this function it even dominates over the two new
perfects (for which see §§2.3.–4., 3.2.4.), whereas the aorist is preferred for the resulta-
tive function (Bužarovska – Mitkovska 2010; Mitkovska – Bužarovska 2011a, §§5.2.2.–
3.; 2011b, §3.5.). However, if we look at the examples adduced in these studies, we
notice the same distribution of pfv. and ipfv. verbs over these main functions of the
perfect domain. Likewise, Makarova (2016, 221f.) notes a general retreat of the l-form
from the perfect domain (in favour of the other two perfects), but nonetheless it can still
be encountered in contexts with experiential function (she gives an example with an
ipfv. verb) and in contexts with resultative function (she gives an example with a pfv.
verb).
These observations lead us to conclude that the ipfv.:pfv. opposition is a reliable
indicator of the resultative-experiential distinction even in Balkan Slavic and that it
crosscuts the aorist-perfect distinction. In other words: the pfv.:ipfv. opposition is strong
enough to dominate over other tense-aspect grams of Balkan Slavic, and this pattern is
in harmony with the distribution of pfv. and ipfv. aspect over the main functions of the
 Slavic Resultatives and Their Extensions 133

perfect domain as we observe it in the remaining Slavic languages, in which the l-forms
developed into a general past many centuries ago or in whch this process is still ongoing.
Before we continue with resultatives based on other participles, a note on l-parti-
ciples built on the pronominal paradigm (the so-called long forms) of participles is in
order.7) L-participles going back to the pronominal declension are found predomi-
nantly in West Slavic and adjacent East Slavic. In some West Slavic varieties these
participles have encroached on the domain of n/t-participles, not only as SubRes (e.g.,
colloquial Czech Jana je vyspa-l-á lit. ‘Jana is well-sleptʼ), but also as ObRes. As ObRes
they have been attested with HAVE- and with BE-auxiliary, and with the exception of
Slovincian, in the HAVE-construction the l-participles practically always agree with the
object-NP; compare colloquial Czech (e.g., Dveř-e má-m zacvak-l-é ‘I have the door
latched’) and Kashubian (e.g., Kóń béł ùkrad-ł-i ‘The horse has been/was stolenʼ). This
extension must have occurred many centuries ago, at least it is attested (occasionally)
in Slovincian, e.g. mä mȯu̯ mä voi̯ ną přie̯ š-l-ï ‘we have experienced [lit. gone through]
the warʼ (Lötzsch 1967; 30f.). In very tiny areas l-participles with a HAVE-auxiliary have
been attested in objectless resultatives; compare, for instance, mám rožl-ý.N ‘I have lit on
(a cigarette) / switched on (the light)ʼ in Hanakian dialects (Central Moravia), following
Damborský (1967, 105). However, for all relevant varieties, except Slovincian
(§3.2.3.2.), hardly any examples with ipfv. participles have been adduced.8) This raises
the question whether the opposition of pfv. vs. ipfv. aspect was strengthened suffi-
ciently, and whether the association of pfv. aspect with resultatives had become strong
enough prior to the spread of pronominal l-participles in resultative constructions.
Otherwise it would be difficult to understand why ipfv. participles in these construc-
tions are generally attested so scarcely.

2.2. B-type constructions

Just like in the discussion of the A-type, in this case we are dealing with participles
combined with a BE-auxiliary, too. The construction with the n/t-participle (= type B.a)
is by far the most widespread, in fact, the only ubiquitous resultative construction in
Slavic, so that I continue with this one first.

2.2.1. n/t-participle + BE (type B.a): interference with the passive

We should be aware that n/t-participles, with roots in pre-Slavic Indo-European, were
originally characterized as resultative verbal adjectives without any clear voice orienta-
tion. That is, they modified the referent, which was the most affected participant of the

7/ In Common Slavic, pronouns of the Indo-European *j-stem encliticized to adjective stems

before they agglutinated and finally fused with the original stems. The result was a new declension
class by which adjective inflection became consistently distinct from the inflection of nouns
(Townsend – Janda 2003, 138–142; Wiemer 2011, 741f.). This distinction started applying to
inflected participles as well. In the following, I will use ‘pronominal form’ as a shortcut for participles
whose desinences go back to this special adjectival inflection. In West Slavic, pronominal forms
subsequently underwent contraction so that they are mostly no longer visible as such. In Czech,
previously longer forms are still characterized by compensatory lengthening ({ý, á, é}, etc., see
examples throughout this article), or, as in Slovak, by jotation ({ej} for feminine singular forms).
8/ Cf. Giger (2003, 63–68), Wiemer – Giger (2005, 83–88) for Czech and Kashubian, Knoll
(2012, 87, 92f.) for Kashubian.
134 Björn Wiemer

situation denoted by the verb stem. In the history of different Slavic varieties, lability
of voice orientation was certainly one of the reasons why the n/t-participle took part
both in the formation of passives (with orientation toward the most patient-like argu-
ment) and the formation of perfects (with neutral orientation or orientation toward the
most agentlike participant); cf. Seržant (2012, 359f.) and Wiemer (2014, 1629–1633)
for surveys and references. One may doubt that subject-oriented n/t-participles in mod-
ern Slavic varieties (e.g., Upper Sorbian wuspany lit. ‘well-sleptʼ, Pol. podpity ‘a little
drunkʼ) continue the ancient heritage of labile voice orientation in every detail.
Certainly, the systematic formation of subject-oriented n/t-participles (including the
Polish ‘bezosobnikiʼ in -no/to or Russian dialectal perfects like у него женё-нo-сь marry-
-PP-REFL ‘he has marriedʼ, lit. ‘at him (it is) marriedʼ; see §2.2.1.4.), which we also
encounter in the Macedonian ima-perfect (see §§2.4., 3.2.4.), arose at much later peri-
ods for which it can be assumed that the pfv.:ipfv.-opposition had already been suffi-
ciently well established. An analogous remark is justified for the variable voice orienta-
tion of l-participles (see §2.1.).
Of course, a full diachronic account would have to investigate the chronological
relation of labile voice orientation to grammatical active : passive distinctions. Although
this question is largely beyond the scope of this contribution, some general remarks
should be offered. First of all, n/t-based resultatives are often not distinguishable from
the passive proper. I use the latter term when referring to a morphologically marked
diathesis by which, for transitive verbs, the more patientlike argument is not only coded
as PSA (= nominative subject), but when the whole construction has an eventive or
progressive focus, i.e. does not assert a resultant state. We will call this the inchoative
foregrounding, or canonical, passive. This passive happens to be systematically ambig-
uous with the ObRes. While, for instance, in modern standard Russian a sentence like
Дверь открытаPFV can be understood only as a resultative ‘The door is open(ed)’, it
could be read as an actional passive (‘The door has been opened’) still in the 19th
century. See, for instance:
(9) (...) радуюсь за самого тебя, что тебе
быстр-ым возвышени-ем отда-н-a справедливость.
quick-INS.SG.N rise[N]-INS.SG return[PFV]-PP-NOM.SG.F justice[F](NOM.SG)
‘I’m glad for you, since justice has been [lit. is] returned to you in a quick rise.ʼ
(NKRJa; Н. С. Лесков: На ножах. 1870)
This observation is in accordance with a crosslinguistically valid implicational relation-
ship established by Nedjalkov – Jaxontov (1983, 26, 32f. [1988, 36, 47f.]): a default
resultative interpretation is more likely and persistent on the present tense level than on
the level of the past (Дверь была открыта) or the future tense level (Дверь будет
открыта).9) The static – dynamic lability (or ambiguity) is a general feature of most
Slavic languages (Giger 2003, 469–478; Knjazev 2007, 567, 571). The roots of this
ambiguity date back a very long time (see §3.2.3.1.), they are certainly older than a
clear differentiation between pfv. and ipfv. verb stems (cf. also Arkadiev – Wiemer,
forthcoming: §3.4.2.). I now turn to this issue.

9/ This implication of the ObRes-passive overlap is connected to the lack of a distinction

between resultative and present perfect (compare He is gone vs. He has gone), mentioned in §1.
 Slavic Resultatives and Their Extensions 135

2.2.1.1. West Slavic

Štícha (1986b, 182f.) shows that in modern Czech the aspect of n/t-participles cannot
be considered a reliable indicator of the actional interpretation of utterances. That is,
ipfv. participles often result in the same reading as their pfv. counterparts (e.g., Stavba
je nedávno opravovaná IPFV [opravená PFV] ‘The building has recently been [lit. is] repairedʼ),
and participles belonging to either aspect show a systematic static – dynamic ambigu-
ity; the dynamic, or eventive, reading is the basis for the analytical passive (see above).
Note that these observations hold true for the nominal forms as well, i.e. even nominal
forms can imply a static reading (e.g., …prý každá je laděna IPFV [vs. naladěna PFV] na jiný
zvuk ‘...allegedly every (string) is/has been tuned with another soundʼ, K. Čapek].10)
Concomitantly, Czech allows for the suppression of the most agentlike argument,
and the promotion of the most patientlike argument to nominative subject, with verbs
of either aspect in both the analytical manner and with the reflexive marker; compare
for ipfv. verbs (cited from Petruxina 2015, 406–410, who follows Štícha 1979):
ipfv., with reflexive marker
(10a) Tahle píseň se zpíva-l-a pořád.
DEM.NOM.SG.F song[F]-NOM.SG REFL sing[IPFV]-PST-SG.F permanently
‘This song was sung permanently.’
ipfv., with n/t-participle
(10b) Tato píseň by-l-a zpívá-n-a
DEM.NOM.SG.F song[F]-NOM.SG be-PST-SG.F sing[IPFV]-PP-NOM.SG.F
ve finském jazyce.
in [Finnish language]-LOC
‘This song was sung in Finnish.’
pfv., with reflexive marker
(11a) Henry Ford zavedl pásovou výrobu a auta byla najednou levná – tedy šlo o pokrok.
Zaved-l-y se předn-í kotoučov-é brzd-y. (cz.internet)
establish-PST-PL.NVIR REFL [front disc brake]-NOM.PL
‘Henry Ford introduced production with the conveyor, moreover the cars were
cheap – at that time this was a progress. Disc brakes at the front wheels were
established.’
pfv., with n/t-participle
(11b) Film by-l natoče-n v ateliéru.
film[M]-NOM.SG be-PST-SG.M produce-PP-NOM.SG.M in atelier-LOC
‘The film was produced in a studio.’
The difference between the participial and the reflexive-marked construction is
that only the former allows for an oblique agent; with the reflexive marker the agent
cannot be expressed, although it remains in the verb’s semantic structure (which allows
to distinguish it from anticausatives).
Slovak ipfv. n/t-participles can also be used as passive markers, even in progressive
function (if derived from secondarily suffixed stems), e.g. Jedlo je pripravované IPFV ‘The
10/ Nominal forms are largely restricted to spisovná čeština, anyway. If they occur in opposition to
pronominal forms, the latter ones are used as resultatives, while the nominal forms are claimed to
convey an eventive function. Cf. Giger (2003, 69–82), Wiemer – Giger (2005, 88) for overviews.
136 Björn Wiemer

meal is being prepared’;11) cf. I s a č e n k o (1960, 370–372). This applies for Czech, too,
e.g. Žáci jsou právě zkoušeni IPFV ‘The pupils are being examined right nowʼ (cited from
Petruxina 2015, 406). After all, however, the two ways the passive may be marked are
not in complementary distribution with pfv.:ipfv. stems (as in Russian), nor has one of
the two ways of marking a passive been abandoned (as in Polish). A similar situation
applies in Serbian-Croatian (see §2.2.1.2.). More research is required in order to under-
stand whether the lability of the static – dynamic interpretation of n/t-participles and
the relative freedom with which both the participial and the reflexive passive can be
marked with verbs of either aspect are causally related phenomena.
The Slavic varieties in which the overlap between ObRes and canonical passive
has been minimized are standard Polish and varieties of Sorbian, albeit for partially
different reasons. Minimization of overlap has been supported by two factors. First, a
special inchoative auxiliary is used in the passive, so that a BE – BECOME contrast can
be explicitly marked: Pol. być ‘be’ vs. zostać ‘become’, Upper / Lower Sorbian wordować
> hodwać / wordowaś. This contrast is largely analogical to Germ. sein – werden (so-called
Zustands- vs. Vorgangspassiv). Second, the aspect contrast in the n/t-participles is
employed most consistently to distinguish between canonical functions of the pfv. and
ipfv. aspect.
Let us first consider Polish. The auxiliary zostać is tightly integrated into the aspect
system: it belongs itself to the pfv. aspect and has an ipfv. counterpart (zostawać), which
is much rarer and restricted to the non-actual (e.g., scenic) present (see ex. 12c).
Concomitantly, zosta(wa)ć requires the n/t-participle to be in the pfv. aspect; rare cases
with ipfv. n/t-participles are usually considered deviations (Wiemer 2004, 301–303;
Górski 2008, 64–78). Thus, ipfv. n/t-participles combine only with the BE-auxiliary,
and, though formed from telic verbs, they are no longer used as indicators of resultant
states – in contrast to what we see in earlier stages of the languages and in many Slavic
varieties until today (see subsequent sections). The distribution of ipfv. and pfv. parti-
ciples corresponds to that of the active voice, in fact, it is even more consistent since, in
connection with the additional auxiliary, more morphological oppositions can be uti-
lized (Weiss 1977, 101–109; Lehmann 1992). Compare the following contrasts:
(12a) Drzw-i są zamknię-t-e.
door-NOM.PL be-PRS.3PL close[PFV]-PP-PL
‘The door is closed.’
 passive = ObRes (compare with ex. 6b for Russian)
(12b) Drzw-i zosta-ł-y zamknię-t-e.
door-NOM.PL become[PFV]-PST-NONVIR.3PL close[PFV]-PP-PL
‘The door has been / was closed.’
 eventive passive (past)
(12c) Drzw-i zostaj-ą zamknię-t-e.
door-NOM.PL become[IPFV]-PRS.3PL close[PFV]-PP-PL
‘The door is closed (and … P happens).’
 eventive passive, but only for non-actual present
11/ With simplex stems we observe ambiguity with the sortal meaning of verbal adjectives (see
§3.1.).
 Slavic Resultatives and Their Extensions 137

(12d) Drzw-i są zamyka-n-e.

door-NOM.PL be-PRS.3PL close[IPFV]-PP-PL
‘The door is being closed / is usually closed.’ (see also ex. 4)
 passive, iterative or processive
In standard Upper Sorbian the canonical (i.e. dynamic) passive can be marked
with the original aorist form bu of być ‘beʼ, but only with pfv. n/t-participles.12) By con-
trast, the imperfect form běše marks the ObRes, and it has been claimed to combine
with ipfv. n/t-participles in cases that correspond to the German dynamic passive (e.g.,
Běch jara wažena ‘I was held in high esteemʼ, Běchmy prošeni ‘We were askedʼ, compare
Germ. Ich wurde sehr verehrt, Wir wurden gebeten, respectively). However, these forma-
tions are bookish and not part of the spoken (dialectal or colloquial) language. Another
way of forming the passive in dialectal Sorbian is by employing the auxiliary wordować
/ wordowaś (Upper / Lower Sorbian), which is a material borrowing from Germ. werden.
In spoken Upper Sorbian, today only the contracted form hodwać is used (Scholze
2008, 197). Both wordować and hodwać combine with n/t-participles of either aspect,
and it is the aspect of the participles that decides about the actional function. Thus,
Chěže wordowachu twarjene IPFV ‘The houses were (being) built’ can have a progressive
meaning, while Chěže wordowachu natwarjene PFV ‘The houses were/have been built’
denotes the event which completed the process of building (Giger 2003, 471–474;
Wiemer – Giger 2005, 101–106; Knjazev 2007, 573, with references). The same holds
true for colloquial Upper Sorbian hodwać, compare
(13a) najprede hodw-e konenc huredźe-n-e.
first AUX-PRS.3SG horse.stable[M]-(NOM.SG) clean[PFV]-PP-N
‘First the horse stable is cleaned.’
(13b) pon hodw-eja kon-e redźe-n-e hač do dźesać-ich.
then AUX-PRS.3PL horse[M]-NOM.PL clean[IPFV]-PP-PL until 10-GEN
‘Then the horses are cleaned until 10 o’clock.’
Ipfv. redźene in (13b) marks a process that can be understood as telic or atelic, whereas
pfv. huredźene in (13a) indicates a telic process (or its completion).13) In general, hodwać
is quite freely used with ipfv. n/t-participles of atelic verbs, including reflexive-marked
ones, which qualify as unergative verbs (e.g., Wot dźesaćich jo hodwało spane lit. ‘(It) was
slept from 10 pmʼ, Germ. Ab zehn wurde geschlafen, Něke hodwe so hrajkane lit. ‘Now (it)
is playedʼ, Germ. Jetzt wird gespielt). It is entirely integrated into the tense system; only
unaccusative verbs do not combine with hodwać (e.g., *Za to hodwe so hańbwane to mean
‘This is for what one gets ashamedʼ); cf. Scholze (2008, 197–199).14) In sum, the par-

12/ The same holds true for Lower Sorbian (Bartels 2008). Similarly, in 19th century Bulgarian
the aorist form bide ‘wasʼ was introduced as a means to indicate a dynamic meaning in the parti-
cipial passive. It however fell into disuse during the 20th century (Knjazev 2007, 571–573, with refer-
ences).
13/ Colloquial Upper Sorbian is the only contemporary Slavic variety in which the [ telic]
distinction is the hierarchically highest criterion of aspect choice (Breu 2000). It is remarkable that
this distinction can carry over into marked voice (as it does for canonical aspect functions in standard
Polish).
14/ In German one can come across such verbs, although passives of reflexive-marked verbs are
not acceptable in all varieties (e.g., standard Germ. Dafür wird sich geschämt). It would therefore be
138 Björn Wiemer

ticipial passive of colloquial Upper Sorbian comes closest to the German system as
concerns the collocational range of the passive auxiliar with participles of different verb
groups (in terms of actionality and valency). It reminds us of Polish with respect to the
actional behavior of n/t-participles of telic verbs, but it differs from it with respect to
the auxiliary: On the one hand, Pol. zostać/zostawać does not allow for ipfv. participles;
on the other hand, Sorbian wordować > hodwać has remained indefinite as for aspect,
i.e. it is not integrated into aspect pairedness.

2.2.1.2. South Slavic

There do not seem to be any comprehensive analyses of the distribution of passive
formations in South Slavic (particularly for its western half). By and large, South Slavic
languages, first of all Serbian-Croatian, are most similar to Czech, although ipfv. n/t-
participles seem to be used less frequently than in West Slavic (Giger 2003, 477f.). In
Slovene, participial and reflexive passive coexist, but the participial passive with ipfv.
participles appears to be strongly avoided (Toporišič 20004, 359; Đukanović – Piper
2009, 352f.). In Serbian, both ipfv. and pfv. verbs allow for reflexive and participial
passive (Ivić 2005, 404). In Bulgarian, ipfv. n/t-participles are used in progressive,
general-factual and iterative function (Maslov 1956, 265f.). Their usage in the passive
is thus similar to Polish (where the general-factual function is rare). However, parti-
cipial passive co-occurs with reflexive passive (Maldžieva 2009, 131f.), which rather
reminds us of Czech and Slovak. This parallel use is also characteristic of Macedonian
(Topolinjska 2009, 223).

2.2.1.3. Interim summary

Against a Slavic backdrop the distribution of n/t-participles in modern Polish and the
Sorbian varieties is outstanding: ipfv. n/t-participles have not been abandoned or dra-
stically reduced (as, e.g., in modern standard Russian or Slovene), but they are produc-
tively used in the formation of the canonical passive. By the same token, they have lost
their former resultative implications (even though all ipfv. n/t-participles that are
admissible in the Polish passive are derived from telic stems) and become integral com-
ponents of a purely voice-related operation. Within the latter (i.e. the canonical passive)
they are now entirely dominated by the grammatical opposition of pfv.:ipfv. aspect
inasmuch as they have been integrated into the standard functions of either aspect
which otherwise obtain in Polish and Sorbian, respectively, in the active voice.15) This
differs from the situation in other Slavic languages in which the functions of ipfv.

worth investigating whether the lack of passives from unaccusative verbs in Sorbian corresponds to
their lack in the originally surrounding regional German varieties. In addition, research is needed in
order to determine more neatly the relation to the reflexive-marked passive with unergative verbs,
which seems to be established rather firmly in standard Upper Sorbian (compare Wot dźesaćich je so
spalo lit. ‘From 10 (it) was sleptʼ).
15/ In Polish, this characteristic system property is parallelled by the development of the former
neuter forms of the nominal declension, which were reinterpreted and now end in -no, -to (Pol. bezo-
sobniki, e.g. Spano świetnie ‘People/We slept excellently’, lit. ‘(it) was slept excellentlyʼ). They are
restricted to the past domain, but are not related to resultatives or perfects any more; they are freely
formed from either aspect, and the only lexical restriction is the requirement that the highest-ranking
(= most agent-like) argument be a human being (Wiemer 2006, 281f.; Górski 2008, 56).
 Slavic Resultatives and Their Extensions 139

n/t-participles are not that consistently dominated by the aspect opposition of the given
language and can be used as resultatives as well (see §3.).

2.2.1.4. Subject-oriented constructions of type B.a

We find this construction type with SubRes in three entirely different regions. This
gives ground to the assumption that we are dealing with polygenesis, i.e. for each region
the salient use of SubRes based on n/t-participles with a BE-auxiliary evolved on inde-
pendent conditions.
Russian dialects from the Pskov-Novgorod region and farther to the north and
east are well-known. SubRes based on n/t-participles are amply attested in Northwest
and North Russian dialects, the farther to the north, the more consistently so (cf.
Wiemer – Giger 2005, 33 for an overview and references). The following examples
show that these constructions are derived from intransitive and transitive verbs (ex.
15–17 are cited from Sobolev 1998, 74):
(14) она уйде-нa / уйде-н …
she.NOM go.away[PFV]-[different endings, of debatable status]
‘She has gone away.’
(15) у лисиц-ы унесе-но курочк-a.
at vixen-GEN.SG carry.away[PFV]-PP(.N?) hen-NOM.SG
‘The vixen has carried away a/the hen.’, lit. ‘at the vixen a hen is carried away’
(16) дa уж бригадир-aми собранё-нo-сь.
PTC PTC team.leader-INS.PL assemble[PFV]-PP(.N?)-REFL
‘Well, the team-leaders gathered.’, lit. ‘by the team-leaders (it) gathered’
(17) у сам-oгo наскладывa-нo пeсeн.
at self-GEN.SG.M compose[IPFV]-PP(.N?) song-(GEN.PL)
‘I have myself composed (many, some) songs.’, lit. ‘at self composed (some) songs’
Note that the participle often does not show agreement. It is an unsettled matter
whether gender and number marking follows any rules at all (this is why neuter gender
is glossed in brackets). While this issue in itself is irrelevant for our topic, it is worth
noting that for transitive verb stems the distinction between SubRes and ObRes often
proves blurred because lack of agreement and non-nominative marking of core argu-
ments do not allow determining syntactic alignment. For instance, do the genitive-NP
песен ‘songs’ in (17) and the nominative-NP курочка ‘hen’ in (15) have to be considered
the object or the subject? The u+GEN-PP, as in (15) and (17), can show some behav-
ioural properties of subjects (Seržant 2012, 371f.), but nonetheless the syntactic con-
trast between the two core arguments remains weak, and it may vary between dialects.
More importantly, regardless of their alignment, resultatives in some Northwest-
-Russian dialects have evolved into an eventive perfect. Indicative are examples in
which the participle combines with a time adverbial specifying a reference time that
differs from the time of the utterance. However, I am unaware of examples with chains
of such participles used in multi-clause units of narrative speech. Concomitantly, the
PP u+GEN or the instrumental NP denote the agent, not just an affected participant,
140 Björn Wiemer

and this is not only implied (as in colloquial standard Russian), but has been conven-
tionalized (Trubinskij 1984, 83–92; Wiemer 2004, 308f.; Seržant 2012, 375–377).
Moreover, SubRes marked with n/t-participles abound in West Slavic (and they
are apparently attested as well in adjacent East Slavic varieties, e.g. in West Ukrainian),
that is their type frequency is remarkably higher than in standard Russian. Compare,
for instance, Pol. uśmiechnię-t-y ‘with a smile [lit. smiled]’ (< uśmiechnąć się PFV ‘smile’),
najedzo-n-y lit. ‘enough-eaten’ (< najeść się PFV ‘eat one’s fill’), wyschnię-t-y ‘shrivelled up’ (<
wyschnąć PFV ‘dry[INTR], shrivel up’).16) We also find them in South Slavic (e.g., Slovene
Sem naspan lit. ‘I am well-sleptʼ, Serb. Ноге су му отечене ‘His legs are swollenʼ),
although obviously to a much lesser extent (according to informants and reference
grammars). Regardless of this, the lexical groups of verbs that serve as input for
SubRes with n/t-participles in West Slavic markedly differ from the lexical range of
admissible input for SubRes in the Northwest Russian dialects and in Macedonian.
Macedonian is the only South Slavic language in which there is a regular BE-based
resultative with n/t-participles. In both the Southwest dialects and the standard lan-
guage the input is predominantly intransitive stems (e.g., Сум доjден ‘I have [lit. am]
comeʼ), and the participle agrees with the subject-NP in gender and number (дојден-.
SG.M, дојден-a.SG.F, дојден-o.SG.N, дoјден-и.PL). In this respect it resembles West Slavic
SubRes. However, it has a much wider lexical input (see §2.4.) and in this respect
reminds us of Northwest Russian dialects (see above).

2.2.2. vši-participle + BE (type B.b)

Type B.b is another resultative construction with a BE-auxiliary. In contrast to type B.a,
it appeared only much later than the Common Slavic period, most probably not ear-
lier than gerunds with which these participles are homonymous (Wiemer – Giger
2005, 40f.; Wiemer 2014, 1634–1637). Consequently, this construction only has a very
limited areal range within Slavic; it is found almost exclusively in regions where East
Slavic dialects have been under lasting contact with Baltic. Russian dialects of the
Pskov district are particularly well-described, but this construction type is ubiquitous
also farther to the Southwest, in particular in Belarusian spoken in the region border-
ing on Lithuanian (Wiemer – Giger 2005, ch. 3+5 for details). Examples (18–19) are
from the Northwestern Russian dialect zone:
(18) трактор слома-вши (< сломаться ‘break down[INTR]’)17) SubRes
tractor-NOM break.down[PFV]-ANTP
‘The tractor has broken down / is broken.’
(19) трав-a скоси-вши (< скосить ‘mow’) ObRes
grass-NOM mow[PFV]-ANTP
‘The grass has been / is mowed.’

Example (19) illustrates that participles with the vši-suffix can also form ObRes con-
structions. This feature is however restricted to a very tiny region. Broadly speaking,
16/ For more details on West and East Slavic cf. the relevant chapters in Wiemer – Giger (2005).
17/ As a rule, resultative participles with vši lack the reflexive marker sja of the base verb, in
particular if the latter is anticausative.
 Slavic Resultatives and Their Extensions 141

vši-based resultatives get rarer, and then vanish, north(east) of Pskov, where n/t-based
resultatives become the only option for forming resultatives and they can in turn form
SubRes constructions. Therefore, in geographical terms, types B.a and B.b show an
almost complementary spread: type B.b has affected mainly the region Southwest from
the Pskov district. In the immediate vicinity of (or overlap with) Baltic varieties vši- and
n/t-participles are in complementary distribution in terms of argument orientation: the
former are attested exclusively in SubRes, the latter exclusively in ObRes – a pattern
which is identical with what we encounter in Baltic (Wiemer – Giger 2005, 43–45).

2.3. C-type: constructions with a HAVE-auxiliary

This type is identified by the auxiliary, i.e. successor forms of Common Slavic imati/
iměti ‘have’. Its distribution in Slavic clearly follows HAVE- vs. BE-biases in basic patterns
of predicative possession. Type C is, thus, absent in the Northeastern corner of Slavic,
i.e. basically Russian, but existent in practically the entire remaining area.
HAVE-based resultatives are widespread all over West Slavic, and their history goes
back to the earliest periods of written documentation (so that we may assume that these
constructions started their lives in times prior to the time of the earliest documents).
Despite their long history, in Polish, Czech and Slovak these constructions have not
shown many, if any, signs of turning into perfects (Wiemer – Giger 2005; also Łaziński
2001; Mendoza 2013, 87f.; Bunčić 2015 on Polish; Giger 2003, 421f. on Czech). First
of all, HAVE-based resultatives in these languages show (almost) consistent agreement of
the n/t-participle with the object NP. They take the n/t-participle with the neuter ending
only if they occur without an object (e.g., Pol. W pokoju mam posprząta-n-e PFV lit. ‘in the
room I have (it) tidied up’) or with quantifying expressions (e.g., Slovak Má
kúpe-n-é PFV liter vín-a ‘S/He has bought a litre of wine’), i.e. when typical agreement con-
trollers are missing. Admittedly, Giger (2003, 394–403; 2016) has shown that in modern
Czech and Slovak written texts with an object-NP, agreement is occasionally lacking, for
Czech he even identified occasional examples with subject agreement. Nonetheless, the
temporal reference even of these resultatives has remained unchanged. Moreover, in
almost every case the participle is derived from a perfective stem.
The situation differs, though, in Kashubian and extinct varieties of Polabian
Slavic (see §3.2.3.2.). In Kashubian we encounter also ipfv. n/t-participles (see
§3.2.3.3.). Since there are, to my knowledge, hardly any corpus-based studies of parti-
cipial constructions, almost nothing can be said concerning the relative frequency
among all these languages.
As for South Slavic, a HAVE-based resultative has attracted most attention in
Macedonian, most particularly in its southwestern varieties (close to Lake Ohrid), from
where its spread started. In the colloquial standard language it has been gaining
ground (Mitkovska – Bužarovska 2011a,b). In this construction the n/t-participle
superficially shows the neuter ending, but it is indeclinable (see §2.4.).
Furthermore, a HAVE-based resultative is known for colloquial Bulgarian and
Bulgarian dialects in contact with Greek, but here the n/t-participle agrees with the
object NP, and it is basically restricted to telic verbs and to some verbs denoting speech
acts, all of them belonging to the pfv. aspect (Georgiev 1976; Velkovska 1994, 19).
142 Björn Wiemer

Actually, this construction is also known in Macedonian. Mitkovska – Bužarovska

(2011b, §3.1.) cite examples from Topolinjska (1983, 30), not only with an n/t-participle
of a pfv. transitive verb (Имам седъм јазовце утепани ‘I have killed seven badgersʼ), but
also with an n/t-participle of a pfv. intransitive verb (Двеста душе партизани имаше
појдени ‘Two hundred partisan souls were [lit. had] goneʼ). This construction has been
known to exist in different dialects all over Macedonia since the 19th century
(L. Mitkovska, p.c.). Its onset could be traced to the early 18th century, and it is
assumed that this construction supplied the basis for the contemporary ima-perfect
(with non-agreeing participle); cf. Friedman (1976, 97f.), Makarova (2016, 224f.).
Note, however, that even in the modern Bulgarian and Macedonian HAVE-construc-
tions with participles that agree with the object-NP a “possessive state” is by no means
obvious (see the last two examples above). This indicates that they seem to have taken
a first step towards a perfect.
Anyway, these Balkan Slavic constructions very much resemble the situation of
those in West Slavic. In the remaining South Slavic area HAVE-based resultatives
appear to be less well attested and have been disputed. They have been mentioned, for
instance, for Serbian and Slovene,18) but judgments of informed native speakers con-
cerning their acceptability and interpretation diverge enormously. However, if they are
accepted, they are restricted to pfv. n/t-participles of telic verbs and to some verbs
denoting social or speech acts (e.g., Slovene Slik-o ima-m proda-n-o ‘I have sold the
painting’). They always agree with an object NP, though occasionally objectless
examples have been cited as well. This, basically, corresponds to the situation in
Bulgarian (cf. also Kuteva 2001, 41f.).
In Molise Slavic (southern Italy) HAVE-based resultatives are used only occasion-
ally, and the participle agrees with an object-NP. The situation with ipfv. n/t-partici-
ples is very complicated and currently being investigated with scrutiny (W. Breu, p.c.).
In the Resian dialects of Slovene this construction is rare as well, at present nothing
can be said about the range of verbs that derive n/t-participles for this construction (M.
Pila, p.c.). Amongst the material from the late 19th century assembled in B a u d o u i n
d e C o u r t e n a y (1895) one occasionally comes across rather clear instances of HAVE-
-based resultatives (19), even an instance without object and the n/t-participle in the
neuter can be found (20). However, the scarcity of data (and contexts) does not allow
for any more far-reaching conclusions. (Examples cited by courtesy of M. Pila.)
(20) [The speaker relates about a bear hunting. After the bear was killed, the hunters
started celebrating this event by drinking, as it were, on behalf of the bear.]
Tadáj, kó ni so há pȧrneslé, tadáj ni so jœ'dly nu pýly hörœ' pö mœdvœdu,
kó ni so ha mœ´-l-y ubý-t-aha. (§454)
CONJ 3PL.NOM be-PRS.3PL him.ACC have-PST-PL kill[PFV]-PP-ACC.SG.M
‘Then, when they brought him [= the bear], they ate and drank for the bear,
because they had killed him.ʼ
(21) Tú bül-ý, k ja má-n wsíka-n-o. (§636)
here ache[IPFV]-PRS.3SG CONJ 1.SG.NOM have-PRS.1SG cut[PFV?]-PP-N
‘Here it hurts, because I have cut (myself) [or: have got cut ?].’
18/ Cf. Nomachi (2006a, 80, 87; 2012; Toporišič 2000, 358) and already Vasilev (1968, 218f.).
 Slavic Resultatives and Their Extensions 143

2.4. Intersection of perfects in Macedonian

Macedonian has become known as a “language with three perfects” (Graves 2000): the
old, Common Slavic perfect (with l-participle) has been superseded by two innovative
formations with a BE- and a HAVE-auxiliary, which were discussed above. The dialectal
and diastratic differentiation of these three perfects is complicated, although it follows
a clear pattern (Friedman 1976; Drinka 2012, 537–539); the functional distribution
among these three formations is most pronounced in the Southwest (near Lake Ohrid).
Here, the old l-participle (with BE) has been driven out of the perfect domain and is
used only as a means of indirect evidentiality or mirativity. The perfect domain has
been taken over by the two innovative formations with n/t-participles; they show simi-
lar grammatical distribution, although the HAVE-construction seems to be somewhat
more widespread in terms of frequency and range of functions, while the BE-construc-
tion is said to be more common with n/t-participles of intransitive stems. N/t-parti-
ciples with either HAVE- or BE-auxiliary occur rather freely with telic or atelic, and with
transitive or intransitive verb stems. The HAVE-based construction allows for verbs with
the reflexive marker (e.g., се има радувано ‘(s/he) has rejoiced’, се има случено ‘(it) has
happened’). Even existential-locative BE and HAVE can themselves be used as lexical
input (имам биде-но, сум биде-н(a) ‘I have beenʼ, имам има-нo; сум има-н(a) ‘I have hadʼ).
Ipfv. n/t-participles occur as well, though they are less frequent (see §3.2.4). For a recent
evaluation cf. Makarova (2016) on the southwestern dialects, Velkovska (1998) and
Mitkovska – Bužarovska (2011a, 2011b) on the standard language.

3. Imperfective participles as input of resultatives / perfects

As discussed above, resultatives turn into (eventive) perfects mainly via the expansion
of lexical input. In turn, a clear distinction between resultative and passive can be made
only if, besides a restructuring of alignment, at least one of two conditions applies: (i)
there is a particular BECOME-auxiliary, (ii) the participle derives from an atelic or
“punctual” verb (e.g., ‘walkʼ, ‘chatʼ, ‘sleepʼ; ‘hitʼ, ‘meetʼ, ‘noticeʼ). Condition (i) is ful-
filled only in standard Polish and Sorbian, condition (ii) holds, though to a different
extent, for standard Polish (only some punctual verbs) and Upper Sorbian, and it holds
to a limited degree for other West Slavic varieties. Another indicator that a resultative
construction has turned into a passive, or some related construction, is the progressive
interpretation of ipfv. participles of telic stems. Apart from standard Polish, this seems
to be the case in Slovak, Czech, and Bulgarian (see §§2.2.1.1.–3.).
In this section I am going to survey the use of ipfv. n/t-participles in predicative
constructions that can either be regarded as resultatives or perfects, or which resemble
them superficially, although they should better be characterized as a different type of
construction. The survey concerns contemporary languages, but I will consider their
diachronic background as far as possible. The survey does by no means aspire to be
exhaustive. The claims are not to be regarded to be hard facts not requiring any more
in-depth analysis, but rather they are made to identify desiderata and, thus, to initiate
further research. I will first provide diachronic comments relevant for Slavic as a whole
(§3.1.), before I examine different varieties and areal subgroupings of Slavic (§3.2.).
144 Björn Wiemer

3.1. Diachronic remarks

We have to bear in mind that at the time of the earliest documents written in Czech or
Polish (12–14th century) the evolution of the aspect system has already advanced in the
sense that one could determine quite a lot of pairs of stems related derivationally by one
of two patterns: (i) simplex (proto-ipfv.)  prefixed stem (proto-pfv.), (ii) prefixed stem
(proto-pfv.)  prefixed stem + suffix (proto-ipfv., ‘secondary imperfectivization’).
Often these stems already tended to fulfill different functions ([ limited], [ repeated],
etc.), at least regarding telic stems (Mende 1999; Wiemer – Seržant, forthcoming),
although this correlation (and the differentiation from each other) was not as clear-cut
as today; for this reason these stems are sometimes dubbed ‘proto-pfv./ipfv.ʼ.
Old Czech started showing some propensity of the resultative meaning for
n/t-participles of prefixed (proto-pfv.) verbs (vs. their simplex proto-ipfv. counterparts),
but in general the distribution was far from clear-cut. Giger (2003, 355–368) rightly
emphasizes that the distinction between resultative and eventive function must not be
confused with the growing differentiation between pfv. and ipfv. stems. We may how-
ever assume that the association of pfv. stems with resultativity was strengthened, while
ipfv. stems were slowly ousted from this domain. We may furthermore assume that the
situation was similar in Old Polish (on which see §3.2.3.1.). Of course, the details of
diachronic development need empirical, corpus-based research. In the following, I will
not consider n/t-participles of verbs that could, for the respective stage of the Slavic
variety, be considered as biaspectual (or, in other words, as not yet integrated into the
aspect system).
If these diachronic considerations are correct, we may assume that ipfv. n/t-parti-
ciples with a resultative function reflect an archaic layer of the language. This layer can
also be detected when these participles served to indicate stable properties of objects.
They can be qualified as verbal adjectives and are quite common in attributive use
(often in stable collocations), as in Russ. варёная картошка ‘boiled potatosʼ, гранёный
стакан ‘cut-glassʼ, стёганый халат ‘quilted gownʼ, Pol. mieszane uczucia ‘mixed feelingsʼ,
duet dęty ‘brass duetʼ / dęty snobizm ‘conceited snobbismʼ, Bulg. дялани камъни ‘hewn
stonesʼ, неканени посетители / Russ. незваные гости ‘uninvited guestsʼ (Maslov 1956,
264), Mac. шетан (човек) ‘(a man) who has been aroundʼ (Mišeska Tomić 2012, 307).
Predicative use is certainly much rarer, although not per se ruled out (as applies to
adjectives in general).19) Such participles do not denote states resulting from concrete
events that immediately preceded (and caused) the respective state; rather, they
describe the noun with respect to some generic properties. They are often lexically
dissociated from their pfv. counterparts, which are productively used to denote resul-
tant states; compare the examples above or, for instance, Pol. Mięso jest pieczone IPFV (a
nie surowe) ‘The meat is roasted (not raw)’ vs. Mięso jest upieczone PFV ‘The meat is / has
been roastedʼ. Characteristically, sortal participles usually derive from ipfv. simplex
stems, stems with secondary suffixes, like Pol. rower s-kład-a-n-y IPFV ‘foldable bicycleʼ, are

19/ Štícha (1986b, 183f.) distinguished between ‘sortalʼ meaning (e.g. Nábytek je leštěný ‘The
furniture is polishedʼ, i.e. ‘belongs to a kind with a special technique of polishingʼ) and ‘qualifyingʼ
meaning (e.g., Stěny jsou zdobeny ornamenty ‘The walls are decorated with ornamentsʼ, i.e. ‘are of a
kind with a special decorʼ). The latter ones seem to be more usual in predicative use and to require
an additional argument.
 Slavic Resultatives and Their Extensions 145

in a clear minority; probably all these stems are telic.20) In the following we have to
exclude them, since they cannot be considered to be components of productively
applied resultative constructions, and often they do not lexically correspond to their
pfv. equivalents.

3.2. Imperfective predicative participles in Slavic

For the sake of brevity, I now assume that as a rule all over Slavic pfv. n/t-participles
are used in passives (both with and without PSA). Moreover, I assume that the ambi-
guity of the ObRes with the participial passive is the default case for Slavic, the clearest
exceptions being standard Polish and Sorbian (see §2.2.1.1.). I will therefore only men-
tion “deviations” from this default.

3.2.1. Standard Russian

In standard Russian, ipfv. n/t-participles have been ousted as part of the passive. Still,
predicative ipfv. n/t-participles seem to have been more widespread by the end of the
19th century (Stender-Petersen 1937), but this requires further investigation. Consider
two examples from the mid-19th century, which can be interpreted as a canonical pas-
sive:
(22) Шампанского не жалели. Первый тост, по обычаю, был посвящен государю и
отечеству.
Тр-и тост-a бы-л-и пи-т-ы зa мое здоровь-е.
3-NOM toast-GEN.SG be-PST-PL drink[IPFV]-PP-PL for [my health[N]]-ACC
‘They did not spare on champagne. As usual, the first toast was dedicated to the
emperor and the fatherland. Three toasts were drunk to my health.ʼ (NKRJa;
А. В. Никитенко. Дневник, т. 1, 1826–1855)
(23) Так как в доме сыро и
одн-a только комнат-a топле-н-a, тo (...).
one-NOM.SG.F only room[F]-NOM.SG heat[IPFV]-PP-NOM.SG.F PTC

(Л. Н. Толстой)
‘Since in the house it was damp and only one room was heated, so (…).’
(cited from Stender-Petersen 1937)
Apart from their productivity, it remains to be investigated to which extent ipfv. n/t-
participles could be used in analogy to ipfv. verbs in the active voice, i.e. with progres-
sive or iterative function. At least, (23) might be read as progressive, (22) presents us
with an iterative context. K n j a z e v (1989, 57f.; 2007, 489) claimed that ipfv. n/t-par-
ticiples are generally incapable of indicating a progressive function (and ex. 23 may be
20/ Whether this testifies to their age, is a question open for further research. On the one hand,
we come across sortal participles of an undoubtedly recent origin, like Russ. фаршированный перец
‘stuffed bell pepper’ or линованный блокнот ‘lined notebook’ (by courtesy of P. Arkadiev). On the
other hand, sortal participles illustrate a purely static meaning which was ascribed to verbal adjec-
tives with the n/t-suffix (as well as the l-suffix) in Common Slavic (see §2.1., §2.2.1.). It needs to be
investigated whether sortal participles are really the continuation of an ancient stage or rather a more
recent phenomenon in a kind of cyclic return to that ancient stage. In addition, a very old pattern
could have been applied to novel cases by analogy. Thus, one has to distinguish between the “life”
of the model and the appearance of particular lexical instantiations.
146 Björn Wiemer

an exception of an anyway obsolete usage of ipfv. n/t-participles), and they typically

occur in negated form. In the latter case a distinction between a general-factual (i.e.
experiential) and an iterative reading normally becomes impossible (or irrelevant). An
experiential reading seems to be common, or even the default interpretation.

3.2.2. East Slavic dialects

The dialects of the Pskov-Novgorod region as well as North Belarusian dialects in the
vicinity of Baltic are markedly different from standard Russian (see §2.2.2.). Actually,
all these dialects belong to a continuum stretching from the northeast of Poland to the
region north(east) of Petersburg which is known for different variants of the so-called
‘new (possessive) perfectʼ in dialectal Russian (Trubinskij 1984, 137–155; Seržant
2012). All these dialects are characterized by a much broader lexical basis of verb stems
that serve as input for resultative constructions; among other things, not only partici-
ples of speech act verbs (e.g., она поблагодаре-н-a PFV lit. ‘she (is) thanked’) and reflexive-
-marked verbs (e.g., выспа-но-сь PFV lit. ‘enough-slept’, женё-нo-сь PFV ‘married’) are attested,
but even atelic verbs with prefixes indicating phasal delimitation (e.g., поработa-вши PFV
lit. ‘worked (for some time)’, послуже-нo PFV lit. ‘served (a certain time)’, побыва-нo PFV lit.
‘stayed (for some time)’), or phasal verbs (кончи-ўшы (iнстытут) lit. ‘finished an / the
institute’, from Belarusian, see below).21) Bearing this in mind, it is not surprising to
find numerous cases in which the predicative participle is derived from an ipfv. stem.
See some examples:

(24) в пастуx-a йдe-нo (Karelia)

in shepherd-ACC go[IPFV]-PP (ending with unclear status)
‘He has become a shepherd [lit. has gone into the shepherd].ʼ
(25) ф кат-a цaрaпa-нa (Pskov region)
at ([f < w < u]) cat-GEN scratch[IPFV]-PP (ending with unclear status)
‘The cat has scratched [i.e. left a scratch].ʼ, lit. ‘At/with the cat (it has) scratched.ʼ
(26) сам-и-то не учи-вши, так ребят-то хоть сколько поучить.
self-PL-TOP NEG learn[IPFV]-ANTP
‘We haven’t learnt (anything) ourselves, so at least our children need to be
taught.’ (quoted from Ryko 2002, 190)
(27) здесь волк-aм(и) много ходи-вши
here wolf-INS.PL much walk[IPFV]-ANTP
и много утащи-вши овец.
and many pull.away[PFV]-ANTP sheep-(GEN.PL)
‘Many wolves have been [lit. walked] here and have carried away many sheep.’
(28) сиже-нo бы-л-o у меня.
sit[IPFV]-PP (ending with unclear status) be-PST-N at 1SG.GEN
‘I have/had sat here.’, lit. ‘(it) was sat at me’ n (last two examples cited from
Sobolev 1998, 75, 77)
21/ For all data and references on the Russian and Belarusian dialects cf. Wiemer – Giger (2005,
35–38, 54–57). Particularly for Belarusian cf. Erker (2015, 92–100).
 Slavic Resultatives and Their Extensions 147

As (26–27) show, ipfv. participles are also derived with the vši-suffix (see also ex. 3).
The choice of suffix, and its voice orientation, depends on the specific dialect (Wiemer
– Giger 2005, 33). The absolute majority of ipfv. participles are from simplex stems (see
23–27), only seldomly do we encounter suffixed stems (see ex. 17 in §2.2.1.4.).
Regardless of the morphology, it is often difficult to determine the function of ipfv.
participles (mostly because contexts are too rudimentary). In many cases they just mark
a resultant state, as in (24), but it is not unusual for these participles to induce an expe-
riential or current relevance meaning. I do not know of any examples with progressive
meaning.
Basically, the same observations concerning the lexical basis and the actional
functions apply to the Northern Belarusian dialects. However, n/t-participles are
hardly attested with ipfv. stems.22) And even if they are, the few examples (like those
adduced by Erker 2015, 93) are ambiguous inasmuch as they can also be read as verbal
adjectives characterizing some stable property. See the next example: ipfv. p’eča-n-y
‘bakedʼ appears to be a genuine participle, but even in this interpretation it is resulta-
tive (as would be its pfv. counterpart):
(29) мы γaлoдн-ыя, ня е-ўшы, (Lida district)
1PL.NOM hungry-NOM.PL NEG eat[IPFV]-ANTP
у нас xлеп ужо бы-ў̯ пeчa-н-ы.
at 1PL.GEN bread[M]-NOM.SG already be-PST-(SG.M) bake[IPFV]-PP-NOM.SG.M
‘We are hungry, we haven’t eaten (anything), we already had baked bread [or:
we had bread baked ?].’, lit. ‘...with us the bread already was baked.’
Ipfv. stems suffixed with vši (i.e. SubRes ones) can usually be interpreted as an
experiential perfect. See negated n’a jeu̯šy ‘have not eatenʼ in (29) and the following
examples:

(30) я лён рвa-ўшы, aлe даўнo

1SG.NOM flax[M]-ACC.SG tear[IPFV]-ANTP but long.ago
‘I have torn [i.e. used to tear] flax, but a long time ago.’
(31) я сaм-a даярк-aй рaбі-ўшы
1SG.NOM self-NOM.SG.F milkmaid[F]-INS.SG work[IPFV]-ANTP
‘I have worked as a milkmaid myself.’

3.2.3. West Slavic

Let us start with comments on diachrony.

22/ Here n/t-participles must be ObRes. In contrast to almost all dialects of the Pskov-Novgorod
area, in the North Belarusian dialects participles with the vši- and the n/t-suffix show complemen-
tary distribution: the former are used for SubRes, the latter for ObRes (see §2.2.2.).
148 Björn Wiemer

3.2.3.1. Diachronic background

Following H a v r á n e k (1937, 100–102, 104), it was assumed that Old Czech ipfv.
n/t-participles (the absolute majority was derived from transitive and telic verbs) were
used only with a resultative function, or as indicators of anteriority with a resultative
implication; they would thus not differ from their pfv. counterparts.23) For instance, it
was claimed that je súzen was used in the sense of contemporary je odsouzen PFV ‘has
been / is sentencedʼ, but not as an equivalent of Latin iudicatur ‘is (being) treated by the
courtʼ. Štícha (1985, 81; 1986a, 387f.) showed that this assumption needed revision24)
(see súzen byl in ex. 33). As a matter of fact, in the 14th century and much later ipfv.
n/t-participles did have a resultative function, as, for instance, in
(32) Ty-s věž-ě Davidov-a,
2SG-2SG tower[F]-NOM.SG of.David-NOM.SG.F
jenž jesti s bran-ěmi dělá-n-a.
REL be-PRS.3SG with armour-INS.PL make[IPFV]-PP-NOM.SG.F
‘You are David’s tower, which is made with armours.’
However, this was no strict rule (rather a default), since in Old Czech there also
occurred many ipfv. n/t-participles in progressive (33-34) or iterative contexts (35); even
atelic verbs served as input (see ex. 34):
(33) na zajtřie ten jistý mládenec súze-n by-l
lad[M]-NOM.SG treat[IPFV]-PP-NOM.SG.M be-PST-(SG.M)
i odsúze-n
and sentence[PFV]-PP-NOM.SG.M
‘On the next day this veritable young man was treated (at court) and sentenced.’
(34) a by-l jsem bičová-n celý den
be-PST-(SG.M) PRS.1SG flog[IPFV]-PP-NOM.SG.M
‘and I was flogged the whole day.’
(35) A také v těch dnech mnozí listi od šlechticóv židovských k Doběsovi
sú posielá-n-i (...).
be-PRS.3PL send[IPFV]-PP-NOM.PL
‘Also in these days many letters are [or: have been ?] sent to Doběs by jewish
noblemen.’
Some ipfv. verbs capable of entering into a canonical passive in Old Czech can hardly
do so any longer in modern Czech, for instance, atelic čekat ‘wait’ and the speech act
verb prosit ‘ask (for a favour)’ (Štícha 1986a, 388). In any case, Štícha’s analysis leads to
the conclusion that ipfv. n/t-participles, if derived from telic stems, certainly had a

23/ According to Havránek, Jan Hus was the first author who noticed the resultative semantics
of Old Czech ipfv. súzen ‘sentencedʼ (Giger 2003, 92, 356f.).
24/ Cf. Giger (2003, 356–362). The temporal semantics of ipfv. n/t-participles which we are
discussing here must not be taken for the aspectual behaviour of pfv. n/t-participles, which probably
always showed variable resultative and eventive function (as they could be employed to refer to an
event time dissociated from utterance time). Cf. Štícha (1985) on Old Czech and compare Old East
Slavic examples like в лѣто 6715 свершенаPFV церковь святая пятница ‘in the year 6715 the church
of the holy Friday was [lit. is] completedʼ (Novgorod, chronicle, 1207), discussed by Feoktistova
(1961, 201).
 Slavic Resultatives and Their Extensions 149

resultative default interpretation, but it could be altered so that the participle acquired
a focus on the related state changing event. Moreover, we also find n/t-participles of
atelic stems, which naturally could not indicate resultant states but had a progressive
function. In general, ipfv. n/t-participles with a resultative function are considered
archaic (Giger 2003, 88, 263) and they come close to being lexicalized, not only as
sortal participles (on which see §3.1.). We see however that their non-resultative func-
tions already existed in the oldest stages visible in written documents. One thus won-
ders when and where they might have gone, and in which systematic and chronological
relation they stand to the fact that in modern standard Czech (i.e. spisovná čeština)
nominal forms of n/t-participles are employed as markers of an eventive passive,
whereas the pronominal forms of participles are used as resultative markers.25)
Together with this, HAVE-based resultative constructions (C-type; see §2.3.) are
attested for the earliest documented stages of Czech as well, and they do not seem to
have considerably changed through the many centuries that have elapsed meanwhile
(Giger 2003, 416–422).26) However, there are only very few examples of ipfv. n/t-par-
ticiples with the HAVE-verb; among them we find psáno ‘writtenʼ, which is clearly
related to a resultant state: nad svú hlavú psáno jmieše ‘he had it written over his headʼ
(mid-14th c.; cited from Giger 2003, 417). An example from ipfv. kupovat (máte krky
kupované, Bílá kniha, 1544; Giger 2003, 418) defies a convincing interpretation:
‘you[PL] have bought churchesʼ, or rather something like ‘you[PL] have churches as
bought onesʼ (i.e. not authentic ones)?
Another problem was pointed out by Mendoza (2013) in her analysis of corpus
data from older stages of Polish (14th–17th c.). The resultative function of n/t-partici-
ples is also typical for attributive use, although not in a sortal meaning. Compare:
(36) a list mia-ł w sw-ej ręc-e pisa-n-y
PTC letter[M]-ACC.SG have-PST-(SG.M) in (PREFL hand[F])-LOC.SG write[IPFV]-PP-PL
t-ymi słow-y: (…)
DEM-INS.PL word-INS.PL
‘and he held a letter in his hand written in these words: (…)ʼ
(Biblia królowej Zofii, 1453–1455; cited from Mendoza 2013, 82)
Altogether, Mendoza (2013, 93f. and p.c.) found only 14 instances with ipfv. n/t-parti-
ciples in her corpus. Many of them cannot be interpreted with enough certainty, for
many one cannot discern whether they should be interpreted predicatively or attribu-
tively (e.g., będą mieć szkarłaty tkane: ‘they will have their scarlets wovenʼ, or ‘they will

25/ Cf. Giger (2003, 70–76). In a sociolinguistic dimension, this opposition is labile inasmuch as
beyond the norms of the spisovná čeština it does not apply. This turns the distinction into a privative
opposition (with the pronominal forms as the unmarked member) to the extent that nominal forms
are pushed out of use.
26/ Several researchers have claimed that the Polish mieć + n/t-participle construction has been
moving from a mere resultative to an eventive perfect. It is striking that such claims have hardly ever
been corroborated by a thorough analysis of (admittedly limited) available diachronic data. This
applies also to the most recent claims raised in Piskorz (2012). The analysis provided by Mendoza
(2013) reveals that, in this regard, Old Polish did not substantially differ from Old Czech. As con-
cerns contemporary Polish, the thorough evaluation of data and previous arguments which was
provided by Łaziński (2001) and Bunčić (2015) does not give support for a considerable shift toward
a perfect proper, either. For a critical data-based discussion of contemporary Czech and Slovak cf.
Giger (2016).
150 Björn Wiemer

have woven scarletsʼ?). In other cases it cannot be decided clearly, whether the subject
of mieć coincides with the agent; compare ex. (38) below and Tytuł tylko Chrystusa mamy
malowany. This could literally be translated into ‘We only have the title of Christ
paintedʼ (i.e. made/had somebody else paint the title of Christ), but the first person
plural subject could also be interpreted to be the agent. The following examples from
Mendoza (2013) show the predicative use of ipfv. n/t-participles from different periods:
(37) któr-e miedzy sob-ą ćwiczo-n-e ma-ją
REL-ACC.PL between REFL-INS exercise[IPFV]-PP-ACC.PL have-PRS.3PL
‘which they have exercised among themselves’ (Kronika turecka, 1496–1501)
(38) Jedn-i na czoł-ach mie-l-i imion-a
one-NOM.PL.VIR on forehead-LOC.PL have-PST-PL.VIR name-ACC.PL.N
pisa-n-e
write[IPFV]-PP-ACC.PL
‘One group had their names written on their foreheads.’ (P. Kochanowski,
1618)
(39) bo Endelman-owie mie-l-i cał-ą galeri-ę
because PN-NOM.PL.VIR have-PST-PL.VIR (whole galery[F])-ACC.SG
zbiera-n-ą nie tyle ze znawstwem ile z namiętnością
collect[IPFV]-PP-ACC.SG.F
‘because the Endelmans had a whole gallery collected, not so much with
connoisseurship but with passion’ (Reymont, 1897)
(40) i P. Wojewod-a kaza-ł
and Mr. Wojewoda[M]-NOM.SG order[PFV]-PST-(3SG.M)
mie-ć koni-e siodła-n-e
have-INF horse-ACC.PL.NVIR saddle[IPFV]-PP-ACC.PL.NVIR
‘and the wojewoda ordered to have the horses saddled’ (Listy staropolskie
z epoki Wazów, 1601–1665)

Only in mieli imiona pisane ‘had their names writtenʼ (ex. 38) could the subject-NP pos-
sibly be interpreted as a beneficient, in the other cases such a reading appears to be
impossible or far-fetched. At present, scarce as the data is this observation at least
allows us to assume that ipfv. n/t-participles, used in complex predicates, have initially
not been restricted to a recipient passive reading. This reading prevailed later, at least
in Polish (see §3.2.3.5.).

3.2.3.2. A general caveat on Polabian and Pomeranian Slavic

The textual basis of the extinct Polabian and Pomeranian dialects is relatively small,
but data on mje̯c ‘haveʼ or bäc ‘beʼ + n/t- or l-participles have been documented quite
well, at least for Slovincian. The participle in the HAVE-based construction predomi-
nantly lacked agreement and was marked with the neuter (-ė), but on rare occasions it
showed agreement with an object-NP (Wiemer – Giger 2005, 83f., following Lötzsch
1967, 30). We know of occasional examples with the HAVE-auxiliary and n/t-participles
of unprefixed stems that can be regarded as ipfv.; compare:
 Slavic Resultatives and Their Extensions 151
.
(41) jáo mȯ u̯ -m fčeráo̯ na po-l-u- rob'i̯ u̇--n-ė
1SG.NOM have-PRS.1SG yesterday on field-LOC work[IPFV]-PP-N
‘Yesterday I worked on the field.’ [lit. I have worked]
Because of the temporal adverb (fčeráo̯ ‘yesterdayʼ) instances like this have to be inter-
preted as an eventive perfect, maybe even as a general past. Lötzsch (1967, 28f., 38f.)
gave more examples like this for Slovincian, both with transitive and intransitive verbs
with ipfv. n/t-participles (e.g., mjȯu̯ l sȯu̯ nė ‘he had goneʼ, mjåu̯ l gådȯu̯ nė ‘he had saidʼ, mja
čutė ‘she had heardʼ, mja prošou̯ nė ‘he had requestedʼ) and with ipfv. l-participles (e.g.
måu̯ stojalė ‘he had stood‘, mȯs väklė ‘you(SG) have learntʼ, mjȯu̯ l ńou̯ slė ‘he had carriedʼ,
må mou̯ čalė ‘he has been silentʼ). The wide lexical range of the HAVE-perfect included
participles of loan verbs (e.g., mȯu̯ ma vjatȯu̯ nė ‘we have betʼ, Germ. wetten; mȯu̯ m lezȯu̯ nė
‘I had readʼ, Germ. lesen), of mje̯ c ‘haveʼ itself (Ja, gap tä vjėƷou̯ l, co tä zefčąta za pråcą
mȯu̯ mjȯu̯ nė ‘If you knew what kind of work the girls have hadʼ) and of GIVE as cau-
sative and existential verb (mȯu̯ m dȯu̯ nė dekac ‘I had let [it, e.g. a horse] coveredʼ). Note
that, even if the base verb was transitive, the construction was subject-oriented in all
instances. It is rather obvious that the German tense system, in particular the com-
pound past (with haben), was copied altogether, and this was accompanied by a lack of
lexical restrictions on participles in compound forms. Here we observe the slavish
imitation of the temporal form:function pairings of German compound tenses, the
passive and causative formations, i.e. wherever auxiliaries and participles combined
into complex predicates. This even went so much so that the future (initially inchoative)
form of bäc ‘beʼ, namely *bo ` > będzie (bądze, bdze, bȯu̯ ), was reinterpreted as a passive
` do
auxiliary, following the homonymy of Germ. werden as passive and future marker. An
analogous situation occurred in 19th and 20th century Kashubian during German
rule (Wiemer – Giger 2005, 83f.; Knoll 2012, 89–95, 99–114, with further references).
In light of these general circumstances, claims concerning the usage of ipfv. parti-
ciples in resultatives or complex tenses have to be made with great caution. It is likely
that conditions of extreme and asymmetric language contact (i.e. with a clear German
dominance) prevented a stepwise run through stages from resultative perfect via expe-
riential and eventive perfect to a compound (general) past and, instead, caused some
wholesale copying of the German system of compound predicates (including the com-
pound past). Provided this assumption is adequate, a quick process of polysemy copy-
ing would have overridden a less drastic, more “organic” development through the
aforementioned stages (known from other languages), and this would make a direct
comparison with the properties of resultatives in even the closest West Slavic varieties
rather inadequate. In the latter ones we find at best incipient stages of a development
from resultative to perfect (let alone a general past), and the process of adaptation to
varieties of German has not led to a mere copying of functions from forms that are
homonymic in German (as the model language).

3.2.3.3. Kashubian
In more recent Kashubian, both n/t- and l-participles in predicative use seem to be
consistently derived from pfv. stems. However, we come across examples with ipfv.
n/t-participles in the HAVE-based construction; compare examples cited from N o m a c h i
(2006b, 181):
152 Björn Wiemer

(42) o t-ёch krôsnięt-ach on

about this-LOC.PL dwarf-LOC.PL he.NOM
mô wiele raz-y powiôdo-n-é.
have.PRS.3SG many time-GEN.PL tell[IPFV]-PP-N
‘He has told (me, us) many times about these dwarfs.’
(43) “Cud sę stanje abo jô ti noce nje przeżije” –
tak wona mje mja-ł-a móvjo-n-é. (A. Majkowski)
so she.NOM 1SG.DAT have-LF-SG.F say[IPFV]-PP-N
‘“A miracle will happen, otherwise I won’t survive this night”  this is what she
said [lit. has said] to me.’
The constructions discussed by Nomachi (2006b) and Knoll (2012) are always subject-
-oriented and the ipfv. participles do not (at least not necessarily) result in a meaning
identical to the pfv. counterpart. (42) is marked as a context of repeated action; (43) is less
clear, but it can be interpreted to have general-factual meaning. This meaning is also easy
to achieve with the participles of ipfv. ‘seeʼ and ‘hearʼ, thus the basic verbs of perception
(Knoll 2012, 108). An analogous observation applies for Macedonian (see §3.2.4.).

3.2.3.4. Czech
Examples found in the literature give the impression that the temporal interpretation
of constructions with ipfv. participles does not often differ from the interpretation with
pfv. participles. This holds true for both the neuter form (if an object-NP is lacking), as
in (44), and for the participle showing object agreement, as in (45):
(44) Má-me vaře-n-o / place-n-o.
have-PRS.1PL cook[IPFV]-PP-N pay[IPFV]-PP-N
‘We have cooked / paid.’
(45) Útrat-u má-m place-n-u / zaplace-n-u.
expense[F]-ACC.SG have-PRS.1SG pay[IPFV]-PP-ACC.F / pay[PFV]-PP-ACC.F
‘I have the expense paid.ʼ
This reminds us of Old Czech (see §3.2.3.1.). However, resultatives with ipfv. n/t-par-
ticiples are of low productivity (both in terms of types and of tokens) in modern
Czech.27) Correspondingly, their acceptability varies considerably depending on the
lexical input (Giger 2003, 89). Moreover, regardless of aspect, the temporal value
depends on the actional class of the verb which the participle is derived from. Thus,
the telic verb vařit IPFV/uvařit PFV ‘cookʼ intrinsically implies a resultative state, while
platit IPFV/zaplatit PFV does so only via components that are extrinsic to the rather punc-
tual meaning of this verb; in this case they refer to possible social consequences which
follow when somebody pays, or does not pay, the bill.
Giger (2003, 91) points out that sometimes the construction mít + n/t-participle
does not have a resultative meaning, but should better be characterized as a recipient
passive (cf. also Knoll 2012, 104). This holds for participles of either aspect; compare
27/ As M. Giger pointed out to me, ipfv. n/t-participles in resultatives are heavily restricted also
from a geographical perspective, as they have been attested only in Southern, Western and Central
Bohemia (see ČJA 4 2002, 586, Map 413).
 Slavic Resultatives and Their Extensions 153

the following examples, in which the first-person subject refers to the addressee of
speech acts:
(46a) ipfv. Občas má-me hláše-n-y nějak-é krádež-e,
sometimes have-PRS.1PL report[IPFV]-PP-ACC.PL (some theft)-ACC.PL
ale jinak nic velkého.
‘Sometimes we have got reported some thefts, but otherwise nothing really
serious.ʼ
(46b) pfv. Letos jsme ne-mě-l-i nahláše-n
this year 1PL NEG- have-PST-PL.VIR report[PFV]-PP-(ACC.SG.M)
ani jeden případ.
NEG (one case[M])-ACC.SG
‘This year we haven’t got reported any single incident.ʼ
In some cases alignment can be ambiguous (out of context), as with the already men-
tioned participle of zaplatit ‘payʼ (by courtesy of J. Panevová):
(47) Má-m zaplace-n-o.
have-PRS.1SG pay[PFV]-PP-N
(i) ‘I have paid.ʼ subject = agent
(ii) ‘I got paidʼ, i.e. ‘Someone paid for me.ʼ subject = beneficient (agent unknown)
Obviously, in order to acquire a recipient passive meaning the construction must be
“fed” with appropriate lexemes, namely with verbs that easily imply a beneficient (or
maleficient). This issue will be taken up in §3.2.3.5.
Finally, Czech predicative ipfv. n/t-participles (both in the nominal and the pro-
nominal form) can be accompanied by definite time or local adverbials; compare: Ta
louka je sečená před týdnem ‘This meadow was [lit. is] cut a week agoʼ, Ta kniha je tištěna
v Berlíně ‘This book was [lit. is] printed in Berlinʼ. Giger (2003, 94f.) qualifies them as
verbal adjectives. On the one hand, their modifiability by adverbials defies their cate-
gorization as sortal participles (on which see §3.1.) and, in this respect, they should
better be compared with the (now obsolete) predicative use of ipfv. n/t-participles in
Russian (see §3.2.1.). On the other hand, at least some of these participles can be used
in answers about the quality of an item (e.g., Jaký je to chléb? Včera pečený ‘Which (kind
of) bread is this? (It was) baked yesterday’). This is indicative of some intermediate
status, which is worth more in-depth analysis.

3.2.3.5. Illusory counterexamples: intersection with the recipient passive

The use of ipfv. n/t-participles in alleged resultatives of type (C) has been reported for
Polish. See, for instance:28)
(48) Mie-l-i=śmy t-o wpaja-n-e
have-PST-PL.VIR=1PL this-N knock.into.head[IPFV]-PP-N
od dzieck-a, że...
from child-GEN.SG COMP

‘From early childhood we got knocked into our heads that ...’
28/ Examples (48–49) are cited from Bunčić (2015), (50–51) are from Łaziński (2001).
154 Björn Wiemer

(49) Ząbk-i mia-ł ogląda-n-e w czerwc-u.

tooth-ACC.PL.NVIR have-PST-SG.M check[IPFV]-PP-ACC.PL.NVIR in June-LOC
‘He had/got his teeth checked in June.’
(50) Czy t-en samochód
Q (DEM car [M])-(NOM.SG)
mia-ł zmienia-n-e opon-y?
have-PST-3SG.M change[IPFV]-PP-ACC.PL.NVIR tire-ACC.PL.NVIR
‘Has this car had its tires changed?’
(51) Co jakiś czas ma leczo-n-e zęb-y.
PTC some.time have-PRS.3SG cure[IPFV]-PP-ACC.PL.NVIR tooth-ACC.PL.NVIR
‘From time to time s/he gets his/her teeth treated.’
However, such data cannot be interpreted as an indication that a resultative is becom-
ing a perfect (pace Piskorz 2013, 157–164). On closer inspection examples with ipfv.
n/t-participles turn out to be a (hitherto overlooked) recipient passive (Bunčić 2015).
To substantiate this let us compare Polish with Kashubian.
Kashubian differs markedly from Polish and other West Slavic languages in that
the HAVE-based construction is subject-oriented even with ditransitives, that is the sub-
ject names the agent; compare an example (cit. from Nomachi 2006b, 182):
(52) Król mjo-ł rozkôzą-n-é, że
king-NOM.SG have-PST-SG.M order[PFV]-PP-N COMP

szёtcё muszёlё narzędё stołove jak ti zbójce na stół kłasc.

‘The king had demanded that everybody, like robbers, had to put their covers
on the table.ʼ
 the king demanded, ordered
We already noticed consistent subject orientation for HAVE+n/t-participle in extinct
varieties of Polabian and in Kashubian (see §§3.2.3.2.–3.) and qualified this behaviour
as pattern borrowing from German. This alignment, irrespective of the argument
structure of the involved verb, coincides with what we expect from the German com-
pound past (haben + participle).
In Polish, the alignment of the arguments of ditransitives in the HAVE-construction
is different. Compare Polish (53) with Kashubian (52): the sentences are superficially
identical, but in Polish the subject (król ‘kingʼ) corresponds to the addressee of the
speech act (while the agent is left unspecified):
(53) Król mia-ł rozkaza-n-e, że
king-NOM.SG have-PST-SG.M order[PFV]-PP-N COMP

 it was demanded from the king, the king was ordered
Compare an analogous example from Slovak (Giger 2000, 19):
(54) Ma-l zakáza-n-é chodi-ť von
have-PST-3SG.M order[PFV]-PP-N walk[IPFV]-INF out
‘He was forbidden tu go out.’
For an ambiguous reading with another arguable ditransitive verb in Czech see exam-
ple (47) in §3.2.3.4.
 Slavic Resultatives and Their Extensions 155

Of course, these examples are mere spotlights. The lexical range of verb lexemes,
as input for n/t-participles, which yield (or can yield) a recipient passive interpretation
in the particular languages that employ HAVE+n/t-constructions has not yet been stud-
ied systematically (cf. Giger 2003, 315–330 for a survey over some older literature).
Most intriguing is the fact that in modern Polish the difference between the HAVE-based
resultative and the recipient passive becomes evident only for ipfv. n/t-participles. Only
ipfv. n/t-participles exclude a resultative reading, so that the recipient passive reading
remains the only sensible one. By contrast, with pfv. n/t-participles resultative and
recipient passive can coincide, both in the sense that the subject of the HAVE-based
construction is coreferent with a recipient-related role (addressee, beneficient / malefi-
cient) and inasmuch as n/t-participles of pfv. verbs denoting speech acts and other
socially relevant events need not (and often cannot) denote a resultant state, but focus
on the respective event itself (see the discussion in §3.2.3.4. and of examples 52–54).
This is a consequence not of perfective aspect, but of the lexical defaults of these verb
classes, which are not strictly telic. But even if we examine a HAVE-based construction
with a pfv. n/t-participle of an indisputably telic verb, we notice that a recipient-passive
and a resultative reading are equally possible (out of context). Consider the following
example:
(55) Już ma-sz okn-a wymy-t-e.
already have-PRS.2SG window[N]-ACC.PL clean[PFV]-PP-ACC.PL.N
‘You[SG] already have (got) the windows cleaned.’
If mieć ‘have’ is in the present tense (as in 55), the reference to the moment of utter-
ance can imply only a resultative state, regardless of who was the agent (the second-
-person subject, the speaker of (55), or somebody else). (55), thus, is clearly a resulta-
tive, but simultaneously it can be a recipient passive. Both notions are not disjunctive.
Essentially, this does not change with mieć ‘have’ in the past tense (Już miałeś okna wymyte
‘You[SG] already had / got the windows cleaned’). Further research should however
investigate whether a shift toward the event becomes more likely, just as this is the case
with the relation between ObRes and the canonical passive (see §2.2.1.).
Furthermore, it may be argued that in the relevant examples (see 48–51) ipfv.
n/t-participles are not derived from ditransitives. Admittedly, this notion can be stretched
and depends on what counts as an argument. However, what is important is that the
denoted situation easily admits a beneficient (or maleficient).29) If the lexical input of an
n/t-participle can imply a beneficient, a natural intersection between a recipient and a
resultative function arises. Their intersection, thus, hinges on the type of participants
that can be associated with the meaning of the verb from which the participle is derived.
This is however only part of the story. We must explain why, in modern Polish, only ipfv.
n/t-participles block, as it were, the resultative reading so that the recipient reading
remains the only sensible one. The explanation must be connected to the fact that the
aspectual functions of ipfv. n/t-participles do not “copy” functions of their pfv. counter-
parts – a fact which we already observed in the Polish canonical passive.
29/ Usually, the beneficient / maleficient has a marginal role in argument structures, and it often
coincides “physically” with the role of the addressee or recipient, or with an inalienable possessor.
These general problems of numerical valency (or the argument – adjunct continuum) are, however,
of minor significance for the point being made here.
156 Björn Wiemer

Therefore, consider the following. As shown above, the recipient passive is com-
patible with a resultative state, but the converse does not always apply. Namely, if we
observe some action or just relate about the fact that some action occurred, but do not
focus on its result, we can experience a benefit (or a disadvantage) from this action. We
have seen that in many Slavic varieties ipfv. n/t-participles (of telic or punctual verbs)
can denote the same resultative state as their pfv. counterparts, but contemporary
Polish differs in this respect: as a rule, Polish ipfv. n/t-participles do not share the resul-
tative implication of their pfv. counterparts, but take part in the regular passive. In the
latter they are able to express the same range of functions as do ipfv. verbs in the active
voice (see §2.2.1.1.). Furthermore, from the data analyzed by Mendoza (2013) we could
infer that in older stages of Polish ipfv. n/t-participles in the HAVE-construction were not
restricted to a recipient passive reading (see §3.2.3.1.). This restriction must, thus, have
come about rather recently, and it appears natural to connect this restriction to the
observation that Polish ipfv. n/t-participles lost their resultative implication.
Now, the canonical functions of ipfv. verbs in modern Polish are progressive,
iterative and general-factual ( experiential). Examples as those in (48–51) can be
interpreted mostly to have general-factual meaning, sometimes as having iterative
meaning (see 51). Progressive readings would be most convincing as illustrations that
the recipient passive can maximally defocus resultative implicatures. Such examples
are difficult to find; a preliminary corpus query in the NKJP with the n/t-participles
of 23 ipfv. verbs following mieć 30) returned only 58 hits in which the participle unam-
biguously was not used as NP-modifier, from these more than 50% (33 hits) were found
for n/t-participles of robić ‘makeʼ. Among the 58 corpus examples only one can safely
enough be considered as marking progressive meaning:
(56) ona może i panią postawić przy odpowiedzi do odpowiedzialności karnej albo być może już
pan-i gdzieś na ten temat w prokuraturze
lady-NOM.SG somewhere (on this topic) in (prosecutor’s office)-LOC
ma prowadzo-n-e postępowani-e.
have-PRS.3SG conduct[IPFV]-PP-N proceeding[N]-ACC.SG
‘In her reply, she could hold also you criminally responsible, or perhaps you’re
already having [conducted] legal proceedings regarding this matter carried
out by the prosecutor’s office.ʼ (shorthand from the 70th meeting of the
Investigative Commission, May, 10, 2005)
Again, the bulk of examples can count as illustrations of the general-factual or
iterative function, only few are to be found in generic contexts, marked for the future
or within the scope of a modal auxiliary.
Alternatively, the recipient passive complex mieć + ipfv. n/t-participle with progres-
sive function can be constructed. I presented sentences like (57–59) to some informed

30/ The query brought to light only cases in which [base=mieć] was immediately followed by an
n/t-participle (in all inflected forms of the accusative) of the verbs bandażować ‘bandage (up)’, czytać
‘read’, oglądać ‘observe, inspect’, kupować ‘buy’, +leczyć ‘heal, cure’, naprawiać ‘fix, repair’, nasuwać ‘pull
on, over’, oceniać ‘assess, evaluate’, +odrabiać ‘catch up; do homework’, +otwierać ‘open’, opatrywać
‘address; affix’, +pisać ‘write’, podwijać ‘roll up’, +prowadzić ‘lead, conduct’, +robić ‘do’, +sprawdzać
‘check’, +stawiać ‘put upright’, +wpajać ‘instill’, wydzielać ‘emit, discharge’, wypominać ‘remind’, zamykać
‘close’, zapewniać ‘assure’, +zmieniać ‘change’. Only for 10 of them (marked with upper case ‘+ʼ) at least
one relevant hit was found.
 Slavic Resultatives and Their Extensions 157

native speakers, asking whether they could accept them as descriptions of ongoing
actions:31)
(57) Waldek akurat ma sprząta-n-y pokój.
PN-NOM just have tidy.up-PP-ACC.SG.M room[M]-ACC.SG
‘Waldek is having his room tidied up right now.’
(58) Dziewczynk-a ma właśnie ogląda-n-e ząbk-i.
girl[F]-NOM.SG have-PRS.3SG just check[IPFV]-PP-ACC.PL.NVIR tooth[M]-ACC.PL
‘The girl is having her teeth checked right now.’
(59) Tat-a w tej chwili ma bandażowa-n-ą nog-ę.
dad[M]-NOM.SG (in this moment) have-PRS.3SG bandage[IPFV]-PP-ACC.SG.F leg-ACC.SG.F
‘Dad is getting his leg bandaged up at this very moment.’
The replies varied enormously, ranging from entire acceptance to almost downright
rejection. Nonetheless, as a common denominator we can say that acceptance (in the
intended meaning) improves if the participle precedes the object-NP (as in 57–59) and
if no potential agent phrase (przez lekarza ‘by the physician’, etc.) is added. It is likely
that the enormous variation in the acceptability ratings was caused by the low fre-
quency of this construction with ipfv. n/t-participles; together with this, its range of
lexical input seems to be limited and there is permanent interference with an alterna-
tive interpretation of the participles as NP-modifiers.
For the time being, we find that the recipient passive reading of the HAVE-con-
struction with ipfv. n/t-participles complies with the usual functions of ipfv. aspect
known for Polish in the active voice. From a diachronic point of view, we may surmise
(given the hitherto scarce documentation) that the clear dissociation of the recipient
passive from a HAVE-based resultative has to be evaluated on the backdrop of the
strengthening of the participial passive; in particular, ipfv. n/t-participles have been
tightly integrated into the passive (while the reflexive-marked passive has become obso-
lete) and the aspect of the participles has become the main factor responsible for the
distinction of aspect functions that correspond to the canonical functions of aspect in
the unmarked (i.e. active) voice. That is, in standard Polish, the pfv.:ipfv.-opposition
now dominates the functions of participles used in complex predicates, lexically inher-
ent features only play a subordinate role.

3.2.4. Macedonian
Let us first deal with the standard language. Velkovska (1998) examined a corpus
composed of post-war texts from the belles-lêttres, mass media, scientific discourse and
some official documents. She found that 9,3% (67 out of 741) of her examples with the
ima-perfect occurred with ipfv. verbs, for the BE-based n/t-construction (“third per-
fect”) the figures for ipfv. verbs oscillated between 8,4% (belles-lêttres) and 15,5% (jour-
nalistic and scientific texts); cf. Velkovska (1998, 148–152). Her examples with ipfv.
ima-perfects show current relevance meaning (as in 60), or they occur in iterative
contexts (see 61):
31/ I want to thank Michał Głuszkowski, Łukasz Jędrzejowski, Anna Kisiel, Marek Łaziński,
Krzysztof Migdalski, Magdalena Pastuchowa, Ewa Willim, and Magdalena Żabowska for lending
me their native intuitions.
158 Björn Wiemer

(60) Во тој силен собор беа замешани и дечинјата од Длапчиште и околните

гостиварски села
на ко-и михамед им имa-шe држa-н-o верскa поукa.
on REL-PL PN 3PL.DAT have-IMPF.3G hold[IPFV]-PP-N religious lesson
‘In this strong assembly the small children from Dlapčište and the surrounding
villages of Gostivar were involved, to which Mixamed had delivered [lit. held]
a religious lesson.ʼ
(61) Неколку пат-и имa-мe слуша-н-o дeкa ...
several time-PL have-PRS.1PL hear[IPFV]-PP-N COMP

‘Several times we have heard that ...ʼ

According to Mitkovska – Bužarovska (2001b, §3.4.3.), ipfv. n/t-participles are typi-
cally employed in experiential function (pfv. participles are possible provided the
object-NP is indefinite).
Let us turn to the BE-based n/t-construction (“third perfect”). In the standard
language its lexical input is restricted. Nonetheless, it sometimes allows for a progres-
sive reading (alternatively to an iterative reading; see ex. 62), but the usual interpreta-
tion seems to be some current relevance reading (ex. 63). However, Velkovska (1998,
153f.) adduced only examples with ipfv. n/t-participles of transitive verbs, so that these
examples can be regarded as passives:
(62) Предаванj-a-ta и oд eдн-и-oт и
transmission[N]-PL-DART.N.PL PTC from one-SG-DART.M.SG PTC

oд друг-и-от циклус бe-a следe-н-и

from other-SG-DART.M.SG cycle[M] be-IMPF.3PL follow[IPFV]-PP-PL
so golemo vnimanie od strana na seminaristite.
‘The transmissions of either cycle were followed with great attention by the
participants of the seminar.ʼ
(63) Т-aa сaмo сo пчел-и e растe-н-a.
3-NOM.SG.F only with bee-PL be-PRS.3SG raise[IPFV]-PP-SG.F
‘She has been [lit. is] raised only with bees.ʼ
See also the following example (by courtesy of L. Mitkovska) with a durative meaning:
(64) Цркв-a-тa e градe-н-a вo период
church[F]-SG-DART.SG.F be-PRS.3SG build[IPFV]-PP-SG.F in period
oд 100 годин-и.
from 100 year-PL
‘The church was [lit. is] built over a period of a hundred years.ʼ
So far, the BE-based construction with n/t-participles of standard Macedonian does not
differ remarkably from Bulgarian: in both languages we observe the same range of
aspect functions and the n/t-participles can be of either aspect (see §2.2.1.2.). However,
even in standard Macedonian we observe that verbs of ingestion show subject orienta-
tion regardless of whether the clause contains an object-NP or not; compare examples
from Makarova (2016, 226):
 Slavic Resultatives and Their Extensions 159

(65a) Јас сум јаде-н-a.

1SG.NOM be-PRS.1SG eat[IPFV]-PP-SG.F
‘I have eaten ( I am not hungry now).’
(65b) Салата сум јаде-н.
salad[F] be-PRS.1SG eat[IPFV]-PP-(SG.M)
‘I’ve eaten a salad.’
This very limited lexical basis of n/t-participles has been extended in southwestern
dialects in both the BE-based and the HAVE-based resultative (cf. Makarova 2016,
223–227, on which the following account is based). The constructions are consistently
subject-oriented. Apart from ingestive verbs we typically find verbs of perception (e.g.,
Сум чуена от постари ‘I have heard from the eldersʼ). Note that the verbs from both
lexical groups are ambitransitive, i.e. they behave alternatively as transitive or intransi-
tive verbs, and this might have served as a bridging context into subject orientation.
The lexical basis of the ima-perfect is even broader; compare this example:
(66) ќе одеше вака, ќе зборвеше сама, не беше во ред...
за такв-и има-м плашe-н-o
for such-PL have-PRS.1SG make.afraid[IPFV]-PP-N
‘She used to walk just this way, she would talk to herself, (she) wasn’t in order...
I made them аfraidʼ (about the way how children were made afraid so that
they would not go out of the house at night).
Neither the BE-based nor the HAVE-based construction rule out ipfv. or atelic verbs per
se. The BE-based construction is considered to favour participles of intransitive verbs,
and it does not require the denoted event to be tied to the utterance interval. This can
be seen from the first occurrence (e појден порано ‘went away earlier’) in the following
example (bold = BE-based, double underlining = HAVE-based):
(67) Aко сe нajдeш сo нeгo добро ќe бидe и зa тeбe.
T-oj e пojдe-н пo-рaнo
he-NOM be-PRS.3SG go[PFV]-PP-SG.M CMP -early
и пo-вeќe имa нaучe-н-o.
and CMP -much have-PRS.3SG learn[PFV]-PP-N
A, e сeдe-н и вo Влашкo,
and be-PRS.3SG sit[IPFV]-PP-SG.M PTC in Vlach.country
сo мajстoр-и, дoстa годин-и.
with master-PL enough year-PL
‘When you meet him it will be useful for you, too. He went away earlier and
has learnt more. He has stayed [lit. he has sat] with the Vlachs, with the
masters, for many years.ʼ (P. Krle)
In general, however, the ima-perfect has advanced further than the BE-based construc-
tion in terms of admissible lexical input, restrictions of agency (in the BE+n/t-construc-
tion the subject must, as a rule, be human) and in the range of temporal functions (see
ex. 66, which is part of a narrative about habits).
The ima-perfect can have resultative or experiential function, as can the BE-based
perfect, although the latter has been claimed to be rare in experiential function. This
160 Björn Wiemer

function seems however natural in many of Velkovska’s (1998) examples in standard

Macedonian; see also e seden in (67) and the following dialectal example: Во Лерин,
Костур и Бугарија сум одена ‘I have been [lit. gone] to Lerin, Kostur and Bulgariaʼ.
What appears more important, for both HAVE- and BE-based perfects the distinction
between resultative and experiential meaning obviously hinges on the aspect of the
participle. Contrast (66) and (68) with (69) for the ima-perfect, and e седен ‘has stayed’
in (67) with остарени ‘have become old’ in (70) for the BE-based perfect:
(68) и волна имa-м работе-н-o.
and wool have-PRS.1SG work[IPFV]-PP-N
‘and with wool I have workedʼ
(69) мoj-тe дец-a штo и имa-м родe-н-o ja
PREFL-DART.PL child-PL REL 3PL.ACC have-PRS.1SG give.birth[PFV]-PP-N 1SG.NOM
‘my children who I have bornʼ
(70) A ниe остаре-н-и вejќe сме,
PTC 1PL.NOM grow.old[PFV]-PP-PL already 1PL
не можеме, немојќи сме и така.
‘we have already become old, we can’t, we are weak, that’s itʼ

The situation farther to the southeast seems to be slightly, though not radically, differ-
ent. As can be inferred from an analysis of the ima-perfect offered by To p o l i n j s k a
(1995, 209f.) for Macedonian dialects in the Aegean region (spoken in Greece), the
temporal interpretation varies widely if the participles are derived from ipfv. verbs.
Ipfv. ima-perfects can have a resultative entailment (as have their pfv. counterparts).
Compare examples (71); (71b) shows that imam gledano does not pass the standard test
on telic processes:
(71a) Гo имa-м гледa-н-o oв-oj филм. ( гo видов цел)
he.ACC have-PRS.1SG watch[IPFV]-PP-N this-SG.M film[M]-SG
‘I have seen/watched that film ( I have seen the whole film.)’
(71b) ? Гo имa-м гледa-н-o филм-oт, нo нe гo дoгледав дo крај.
he.ACC have-PRS.1SG watch[IPFV]-PP-N film[M]DEF.SG.M
? ‘That film I have seen, but I haven’t seen it to the end.ʼ
However ima-perfects with ipfv. participles can also mark repetition in the past; this
implies relevance for the utterance interval (probably preferred reading for ex. 72):
(72) Јас имa-м вeќe доаѓa-н-o тукa.
1SG.NOM have-PRS.1SG already come[IPFV]-PST.PP-N here
‘I have already been [lit. come] here.ʼ
(73) Лeб-oт гo има-м купува-н-o
bread[M]- DART.SG he.ACC have-PRS.1SG buy[IPFV]-PP-N
вo нaj-блискa-тa prodavnica.
in SUP -close-DART.SG.F shop[F]-SG
‘Usually I bought [lit. have bought] bread in this shop.ʼ
 Slavic Resultatives and Their Extensions 161

We can furthermore find examples with ipfv. participles which might be interpreted as
denoting durative situations. However, an alternative interpretation is, again, current
relevance. See (74): The speaker implies that they have acquired a certain level of edu-
cation:
(74) Јас имa-м уче-н-o
1SG.NOM have-PRS.1SG learn[IPFV]-PP-N
три годин-и нa болгарск-o скол‘j-e.
3 year-PL on (Bulgarian school)-N.SG
‘I have studied three years in a Bulgarian school.ʼ

4. Recapitulation and conclusions

In my cross-Slavic survey of resultative constructions I have dealt with the following
questions: First, in which Slavic varieties have ipfv. participles been, or are still, used in
constructions that derive from resultatives? Second, if ipfv. participles in such construc-
tions are attested, we should ask whether and how they change the meaning of the
construction and, third, whether these meanings (in contrast to pfv. participles) show
any systematic affinity to aspect functions known for the respective Slavic variety in the
active voice. These issues amount to the following basic question: Have the function
inventories of pfv. and ipfv. verbs become dominant over originally resultative con-
structions? If yes, this dominance can manifest itself in two ways: Either the construc-
tion has started including atelic verb stems, or ipfv. participles of telic stems need not
(and usually do not) denote a resultant state any more (contrary to their pfv. counter-
parts).
On account of this survey, the shortest answer to this basic question is that the new
aspect opposition (pfv. : ipfv.) proves dominant in the majority of constructions that are
related to resultatives in Slavic. The oldest perfect (based on the l-participle) is domi-
nated by aspect not only in North Slavic and Slovene (where this is trivial since the
l-perfect has become a general past), but even in Balkan Slavic. Here the l-perfect has
not turned into a general past, nonetheless the pfv.:ipfv.-opposition distinguishes the
same clusters of canonical aspect functions as in North Slavic. This holds true for the
distribution of ipfv. and pfv. aspect in the Bulgarian aorist, too, at least in those cases
in which the aorist expresses functions otherwise associated with the perfect domain
(pfv. implies a resultative, whereas ipfv. implies an experiential function); see §2.1. For
the same reason it can be argued that the pfv.:ipfv.-opposition dominates over the two
“new perfects” in Macedonian; at least this tendency appears to be clear for both col-
loquial standard and diverse dialects (see §3.2.4.). The picture is less clear in Russian
dialects north(east) from the Pskov region, where n/t-participles have been the exclusive
option for any type of resultative (or perfect). By contrast, East Slavic dialects border-
ing (or overlapping) with the Baltic territory do not only show complementary distribu-
tion of the n/t-suffix (for ObRes) and the vši-suffix (for SubRes), but also the rigidity
with which these suffixes are associated to aspect differs: there are hardly any ipfv.
n/t-participles in resultative constructions, whereas vši-participles are rather widely
attested in either aspect. However, the functional distribution in the perfect domain
162 Björn Wiemer

(resultative vs. experiential), again, seems to correlate with the aspect of the stem (see
§3.2.2.). In this respect, East Slavic dialects southwest from the Pskov region are similar
to the new perfects in Macedonian, while, by contrast, in the northern-most area of
Slavic (in the European part of Russia), where resultatives are based only on n/t-parti-
ciples, the aspect of the participle does not seem to be indicative of any consistent dis-
tinction of functions within the perfect domain. This however needs to be checked on
a larger amount of reliable data with sufficient contexts.
As concerns West Slavic, we have to be careful in the assessment of the relatively
scarce textual data from extinct Polabian and Pomeranian dialects, because their
system of tenses and of complex predicates was most probably copied from German in
a “wholesale” manner. Nevertheless, it might be worth trying to establish whether
combinations of a BE- or HAVE-auxiliary with ipfv. vs. pfv. participles show any stable
functional distribution comparable to canonical aspect functions. In this context, it
seems intriguing to compare Slovincian to recent and contemporary varieties of
Kashubian. As the discussion in §3.2.3.5. revealed, Kashubian remarkably differs from
standard Polish in terms of voice orientation of resultatives and perfects. It should be
established more thoroughly to which extent differences apply to their temporal func-
tions in interaction with the aspect of the participles.
N/t-participles have been the most persistent and ubiquitous core component of
resultatives in Slavic, with roots even older than those of the l-participle. However,
their voice orientation has been more variable, or labile, in the history of particular
Slavic languages (or subareas). During this history the resultative meaning of n/t-par-
ticiples has become tightly associated to pfv. aspect. In comparison to n/t-participles,
aspect based on stem derivation can be considered a recent newcomer, but it started
dominating the participial system, in particular n/t-participles. Nonetheless, originally
resultative constructions have expanded only rarely to also involve atelic verb stems (see
below). Increasing dominance of the new aspect opposition has had several repercus-
sions, partially different ones for different subareas of Slavic. In general, ipfv. n/t-par-
ticiples in predicative use have been reduced. This concerns attributive use, too, apart
from lexicalized use with sortal meanings (see §3.1.). Obviously, this reduction has
been strongest in standard Russian (probably in entire East Slavic, except for its
northern-most area; see above), where practically ipfv. n/t-participles in predicative use
have been abandoned altogether. As a side effect, the means of marking the passive for
ipfv. and pfv. verbs (reflexive marker vs. n/t-participles) are now in complementary
distribution. In turn, we see that for most cases in which predicative ipfv. n/t-participles
have stayed in use they do not differ from their pfv. counterparts, i.e. both have resulta-
tive meaning; these are considered archaic, many of them appear lexicalized (e.g. in
Czech; see §3.2.3.1.). In principle this applies regardless of a distinction between
nominal and pronominal forms (as in spisovná čeština), partially because the nominal
forms are themselves archaic and clearly have been on their retreat outside the main
area of South Slavic. In addition, all ipfv. n/t-participles (both in predicative and
attributive use) appear to be derived from telic verb stems. Thus, in either case – total
reduction (e.g. Russian) or lack of complementary distribution with pfv. counterparts
(e.g., most West and South Slavic languages) – the lexical feature of telicity, which is
older than the aspect distinction, prevailed over the growing significance and reliabil-
ity of distinctions conveyed by the choice of aspect.
 Slavic Resultatives and Their Extensions 163

This conflict between aspect and telicity ended up with different, in fact diametri-
cally opposed consequences: either ipfv. n/t-participles were marginalized or dropped,
or they were integrated into other categorical distinctions, namely voice oppositions.
This can be seen all over South and West Slavic, but most prominently in Polish and
Sorbian. Probably, the usage of dedicated passive auxiliaries helped establish an
explicit static – dynamic distinction (or otherwise, ObRes – (inchoative) passive distinc-
tion). However, the distribution of ipfv. and pfv. participles over canonical functions of
either aspect, too, analogous to their distribution in unmarked voice, proves to be very
strict, obviously stricter than that in Czech, Serbian or Bulgarian, where the particip-
ial passive still shows overlapping distribution with a reflexive-marked passive (which
does not allow for an expression of the agent). More specifically, in standard Polish ipfv.
n/t-participles with a HAVE-auxiliary (mieć) have become an unambiguous indicator of
the recipient passive (see §3.2.3.5.).
However, while the Polish canonical and the recipient passive are restricted to telic
verbs and certain types of punctual verbs (e.g., denoting speech acts), colloquial Upper
Sorbian has loosened this restriction, evidently under German influence. Other areas
in which the telic restriction has been released are Macedonian (its new perfects, pri-
marily in the Southwest dialects) and the East Slavic dialects (including those in imme-
diate neighborhood with Baltic, i.e. southwest of the Pskov region). Whether, and to
which extent, this expansion of the lexical input to participles into originally resultative
constructions can be considered as indicative of preserved earlier stages of the language
system or as innovations – and which role language contact played in causing conser-
vative or innovative behaviour– is a question that remains to be dealt with in further
research.
In conclusion, we may generalize that, on the one hand, telicity has remained a
factor which restricts the usage of constructions that developed from resultatives in the
majority of Slavic varieties, even if ipfv. participles have become core components of a
canonical (inchoative) or recipient passive (as in Polish). On the other hand, even for
most varieties in which the telic restriction has been overcome (as in Macedonian,
certain East Slavic dialects, Kashubian), the aspect of the participle seems to reliably
correlate with functions of ipfv. and pfv. verbs usually known in active voice. Of course,
this is a simplified illustration, especially data on non-standard varieties, in particular
on Slovincian, Kashubian and Russian dialects of the Pskov-Novgorod region and
North(east) from it, require more detailed analyses on the basis of sufficiently rich
texts.*32)

* Acknowledgments
For valuable advice and consultation I am indebted to Walter Breu, Eleni Bužarovska, Jasmina
Grković-Major, Marek Łaziński, Anastasija Makarova, Maksim Makartsev, Imke Mendoza, Liljana
Mitkovska, Malinka Pila, Lenka Scholze, and Ewa Willim. I am particularly obliged to Peter
Arkadiev and Markus Giger for their valuable comments on a previous version of this article. I also
thank an anonymous reviewer for a couple of amendments and Anke Lensch for her thorough proof-
reading. Of course, the usual disclaimers apply.
164 Björn Wiemer

Abbreviations (in addition to Leipzig Glossing Rules)

ANTP  anteriority participle (vši, including allomorphs), CMP – comparative, COMP –
complementizer, CONJ  conjunction, DART – definite article, IMPF – imperfect, INTR –
intransitive, NVIR – nonvirile, PN – proper name, PP  n/t-participle, PREFL  possessive
reflexive, PTC  particle, Q – question marker, SUP – superlative, TOP – topic marker,
VIR – virile

R eferences :
Arkadiev – Wiemer, forthcoming: ARKADIEV, P. M. – WIEMER, B.: Perfects in Baltic and Slavic.
In: Crellin, R. – Jügel, Th. [edd.]: Perfects in Indo-European languages, vol. II: The later history
of perfects in IE languages. Amsterdam – Philadelphia forthcoming.
Bartels 2008: BARTELS, H: Konkurrierende Passivkonstruktionen in der niedersorbischen
Schriftsprache. Ein Beispiel für Sprachwandel durch Purismus. In: Kempgen, S. –
Gutschmidt, K. – Jekutsch, U. – Udolph, L. [edd.]: Deutsche Beiträge zum 14. Internationalen
Slavistenkongress, Ohrid 2008. München 2008, 27–36.
Baudouin de Courtenay 1895: BAUDOUIN DE COURTENAY, J. N.: Materialien zur südslavischen
Dialektologie und Ethnographie, I: Resianische Texte, gesammelt in den JJ. 1872, 1873 und 1877.
St. Petersburg 1895.
Breu 1988: BREU, W.: Resultativität, Perfekt und die Gliederung der Aspektdimension. In:
Raecke, J. [ed.]: Slavistische Linguistik 1987. München 1988, 44–74.
Breu 1994: BREU, W.: Der Faktor Sprachkontakt in einer dynamischen Typologie des Slavischen.
In: Mehlig, H.-R. [ed.]: Slavistische Linguistik 1993. München 1994, 41–64.
Breu 1998: БРОЙ, В.: Сопоставление славянского глагольного вида и вида романского типа
(аорист : имперфект : перфект) на основе взаимодействия с лексикой. In: Черткова,
M. Ю. [ed.]: Типология вида (Проблемы, поиски, решения). Москва 1998, 88–99.
Breu 2000: BREU, W.: Der Verbalaspekt in der obersorbischen Umgangssprache im Rahmen des
ILA-Modells. In: Breu, W. [ed.]: Slavistische Linguistik 1999. München 2000, 37–76.
Bunčić 2015: BUNČIĆ, D.: “To mamy wpajane od dziecka” – a recipient passive in Polish?
Zeitschrift für Slawistik 60, 2015, 3, 411–431.
Bužarovska – Mitkovska 2010: BUŽAROVSKA, E. – MITKOVSKA, L.: The grammaticalization of
the habere-perfect in standard Macedonian. Balkanistika 23, 2010, 43–66.
Bybee et al. 1994: BYBEE, J. – PERKINS, R. – PAGLIUCA, W.: The evolution of grammar (Tense, aspect,
and modality in the languages of the world). Chicago – London 1994.
ČJA 4: BALHAR, J. [ed.]: Český jazykový atlas, 4. Praha 2002.
Dahl 1985: DAHL, Ö.: Tense and Aspect Systems. Oxford – New York 1985.
Dahl – Hedin 2000: DAHL, Ö. – HEDIN, E.: Current relevance and event reference. In: Dahl, Ö.
[ed.]: Tense and aspect in the languages of Europe. Berlin – New York 2000, 385–401.
Damborský 1967: DAMBORSKÝ, J.: Participium l-ové ve slovanštině. Warszawa 1967.
Drinka 2012: DRINKA, B.: The Balkan perfects: Grammaticalization and contact. In: Wiemer, B.
– Wälchli, B. – Hansen, B. [edd.]: Grammatical replication and borrowability in language contact.
Berlin – New York 2012, 511–558.
Drinka 2017: DRINKA, B.: Language Contact in Europe: The Periphrastic Perfect through History.
Cambridge etc. 2017.
Đukanović – Piper 2009: ĐUKANOVIĆ, M. – PIPER, P.: Slovenski jezik. In: Пипер, П. [ed.]:
Јужословенски језици: граматичке cтруктуре и функције. Београд 2009, 257–377.
Erker 2015: ЭРКЕР, А.: Структурные черты смешанных белорусских говоров на балто-славянском
пограничье. Leipzig 2015.
 Slavic Resultatives and Their Extensions 165

Feoktistova 1961: ФЕОКТИСТОВА, A. С.: К истории составного сказуемого с присвязочной

частью, выраженной причастием страдательного залога прошедшего времени (на
материале Новгородских памятников письменности XII–XVI вв.). In: Борковский,
В. И. – Катков, С. И. [ред.]: Исследования по лексикологии и грамматике русского языка.
Москва 1961, 194–208.
Friedman 1976: FRIEDMAN, V. A.: Dialectal Synchrony and Diachronic Syntax: The Macedonian
Perfect. In: Steever, S. B. et al. [edd.]: Papers from the Parassesion on Diachronic Syntax.
Chicago Linguistic Society, Chicago 1976, 96–103.
Georgiev 1976: ГЕОРГИЕВ, В.: Възникване на нови сложни глаголни форми със спомагате-
лен глагол ‘имам’. In: Пашов, П. – Ницолова, Р. [ред.]: Помагало по българска морфоло-
гия. Глагол. София 1976, 294–311. [Reprint from: Известия на Института по български
език 5, 1957.]
Giger 2000: GIGER, M.: Syntaktické modelovanie slovenských posesívnych rezultatívnych
konštrukcií v rámci dependenčnej gramatiky. Jazykovedný časopis 51, 2000, 1, 17–26.
Giger 2003: GIGER, M.: Resultativkonstruktionen im modernen Tschechischen (unter Berücksichtigung der
Sprachgeschichte und der übrigen slavischen Sprachen). Bern etc. 2003.
Giger 2016: GIGER, M: Kongruenzbrüche in slovakischen possessiven Resultativa (Evidenz aus
dem slovakischen Nationalkorpus). Jazykovedný časopis 67, 2016, 3, 283–294.
Górski 2008: GÓRSKI, R. L.: Diateza nacechowana w polszczyźnie (Studium korpusowe). Kraków 2008.
Graves 2000: GRAVES, N.: Macedonian – a language with three perfects? In: Dahl, Ö. [ed.]: Tense
and aspect in the languages of Europe. Berlin – New York 2000, 479–494.
Havránek 1928–1937: HAVRÁNEK, B.: Genera verbi v slovanských jazycích I–II. Praha 1928–1937.
Isačenko 1960: ИСАЧЕНКО, А. В.: Грамматический строй русского языка в сопоставлении с сло-
вацким. Морфология, часть вторая. Bratislava 1960.
Ivić 2005: ИВИЋ, M. [ed.]: Синтакса савременнога српског језика. Проста реченица. Београд
2005.
Knjazev 1989: КНЯЗЕВ, Ю. П.: Акциональность и статальность: Их соотношение в русских
конструкциях с причастиями на -н, -т. München 1989.
Knjazev 2007: КНЯЗЕВ, Ю. П.: Грамматическая семантика (Русский язык в типологической
перспективе). Москва 2007.
Knoll 2012: KNOLL, V.: Kašubština v jazykovém kontaktu. Praha 2008.
Kuteva 2001: KUTEVA, T.: Auxiliation. An Enquiry into the Nature of Grammaticalization. Oxford –
New York 2001.
Lehmann 1992: LEHMANN, V.: Le prétérit déictique et le prétérit narratif en polonais moderne.
In: Guiraud-Weber, M. – Zaremba, Ch. [edd.]: Linguistique et slavistique. Melanges offerts à
Paul Garde. Aix-en-Provence – Paris 1992, 545–557.
Lindstedt 2000: LINDSTEDT, J.: The perfect – aspectual, temporal and evidential. In: Dahl, Ö.
[ed.]: Tense and aspect in the languages of Europe. Berlin – New York 2000, 366–383.
Litvinov – Nedjalkov 1988: LITVINOV, V. P. – NEDJALKOV, V. P.: Resultativkonstruktionen im
Deutschen. Tübingen 1988.
Lötzsch 1967: LÖTZSCH, R.: Das Tempussystem des Slovinzischen im Vergleich zu dem des
Sorbischen und Deutschen. Lětopis Instituta za serbski ludospyt, 1967, 14, 23–46.
Loporcaro 1995: LOPORCARO, M.: Grammaticalizzazione delle perifrasi verbali perfettive
romanze e accordo del participio passato. Archivio Glottologico Italiano 80, 1995, 144–167.
Loporcaro 1998: LOPORCARO, M.: Sintassi comparata dell’accordo participiale romanzo. Torino 1998.
Łaziński 2001: ŁAZIŃSKI, M.: Was für ein Perfekt gibt es im modernen Polnisch? Bemerkungen
zum Artikel “Gibt es ein Perfekt im modernen Polnisch?” von H. Weydt und A. Kaźmier-
czak (Linguistik online 4, 3/99). Linguistik online 8, 2001.
166 Björn Wiemer

Makarova 2016: МАКАРОВА, А. Л.: О формах и функциях перфекта в западномакедонских

диалектах. In: Майсак, Т. А. – Плунгян, В. А. – Семенова, Кс. Р. [red.]: Исследования
по теории грамматики, вып. 7: Типология перфекта. Санкт-Петербург 2016, 217–234. [=
Acta Linguistica Petropolitana, t. XII, č. 2.]
Maldžieva 2009: МАЛДЖИЕВА, B.: Български език. In: Пипер, П. [ed.]: Јужословенски језици:
граматичке структуре и функције. Београд 2009, 19–137.
Maslov 1956: МАСЛОВ, Ю. С.: Грамматика болгарского языка. Москва 1956.
Maslov 1983: МАСЛОВ, Ю. С.: Результатив, перфект и глагольный вид. In: Недялков, В. П.
[red.]: Типология результативных конструкций (результатив, статив, пассив, перфект).
Ленинград 1983, 41–54. [English translation: Resultative, Perfect, and Aspect. In:
Nedjаlkov, V. [ed.]: Typology of Resultative Constructions. Amsterdam ‒ Philadelphia 1988,
63–85.]
Mende 1999: MENDE, J.: Derivation und Reinterpretation: Die Grammatikalisierung des russi-
schen Aspekts. In: Anstatt, T. [ed.]: Entwicklungen in slavischen Sprachen. München 1999,
285–332.
Mendoza 2013: MENDOZA, I.: Verhinderte Grammatikalisierung? Zur Diachronie von Resultativ-
konstruktionen mit mieć ‘habenʼ im Polnischen. Wiener Slawistischer Almanach 72, 2013,
77–102.
Mišeska Tomić 2012: MIŠESKA TOMIĆ, O.: A Grammar of Macedonian. Bloomington 2012.
Mitkovska – Bužarovska 2011a: МИТКОВСКА, Л. – БУЖАРОВСКА, Е.: За употребата на
има-перфектот во македонскиот стандарден jазик во релациjа со конкурентните
глаголски форми. In: Тополињска, З. [ed.]: Перифрастичните конструкции со ‘ESSEʼ и
‘HABEREʼ во словенските и во балканските јазици. Скопје 2011, 55–82.
Mitkovska – Bužarovska 2011b: МИТКОВСКА, Л. – БУЖАРОВСКА, Е.: За има-перфектот во
македонскиот стандарден јазик. Прилози XXXV, 2011, 1, 45–71.
Mittwoch 2008: MITTWOCH, A.: The English Resultative perfect and its relationship to the
Experiential perfect and the simple past tense. Linguistics and Philosophy 31–3, 2008, 323–
351.
Nedjalkov – Jaxontov 1983: НЕДЯЛКОВ, В. П. – ЯХОНТОВ, С. Е.: Типология результативных
конструкций. In: Недялков, В. П. [red.]: Типология результативных конструкций
(результатив, статив, пассив, перфект). Ленинград 1983, 5–41. [English translation:
Nedjalkov, V. P. [ed.]: Typology of resultative constructions. Amsterdam – Philadelphia 1988,
3–62.]
NKJP: Narodowy korpus języka polskiego [on-line]. <http://nkjp.pl/> [cit. 05-01-2017].
NKRJa: Национальный корпус русского языка [on-line]. <http://www.ruscorpora.ru/> [cit.
05-01-2017].
Nomachi 2006a: НОМАТИ, M.: От посессивности к аспектуальности: дистрибуция глаголов
imati и biti в словенском языке (в типологическом освещении). Slavia Meridionalis 6,
2006, 65–90.
Nomachi 2006b: NOMACHI, M.: Polska konstrukcja rezultatywno-posesywna mam to zrobione a
kaszubskie jô móm to zrobioné. Język Polski LXXXVI, 2006, 3, 173–183.
Nomachi 2012: NOMACHI, M.: On the so-called possessive perfect in standard Serbian language
(with a glance at other Slavic languages). Leptir Mašna: The literary journal of Balkan Studies
9, 2012, 1, 89–97.
Petruxina 2015: ПЕТРУХИНА, Е. В.: Вид в грамматическом контексте пассива (на материале
русского и чешского языков). In: Benacchio [Бенаккьо], R. [ed.]: Глагольный вид:
грамматическое значение и контекст / Verbal aspect: Grammatical Meaning and Context.
Сборник докладов III Конференции Аспектологической Комиссии, состоявшейся
в Падуе с 30.9. пo 4.10.2011. München etc. 2015, 403–420.
 Slavic Resultatives and Their Extensions 167

Piskorz 2012: PISKORZ, J.: Die Grammatikalisierung eines neuen Perfekts im Polnischen. Ein Beitrag zur
Entwicklungslogik des Perfekts. München 2012.
Plungjan 2016: ПЛУНГЯН, В. А.: К типологии перфекта в языках мира: предисловие. In:
Майсак, Т. А. – Плунгян, В. А. – Семенова, Кс. Р. [red.]: Исследования по теории
грамматики, вып. 7: Типология перфекта. [= Acta Linguistica Petropolitana, t. XII, č. 2.].
Санкт-Петербург 2016, 7–36.
Ryko 2002: РЫКО, A. И.: Причастия на ши- в одном западнорусском говоре торопецко-
-холмского региона. Русский язык в научном освещении, 2002, 2, 171–193.
Scholze 2008: SCHOLZE, L.: Das grammatische System der obersorbischen Umgangssprache im
Sprachkontakt. Bautzen 2008.
Seržant 2012: SERŽANT, I. A.: The so-called possessive perfect in North Russian and the Circum-
-Baltic area. A diachronic and areal account. Lingua 122, 2012, 356–385.
Sobolev 1998: СОБОЛЕВ, А. Н.: О предикативном употреблении причастий в русских
диалектах. Вопросы языкознания, 1998, 5, 74–89.
Stender-Petersen 1937: Stender-Petersen, A.: Das russische part. praet. pass. von imperfektiven
verben. In: Hjelmslev, L. et al. [edd.]: Mélanges linguistiques offerts à M. Holger Pedersen (à
l’occasion de son soixante-dixième anniversaire). København 1937, 397–405.
Štícha 1979: ŠTÍCHA, F.: O užívání a významu dvojí formy trpného rodu v současné spisovné
češtině. Naše řeč 62, 1979, 1, 57–71.
Štícha 1985: ŠTÍCHA, F.: Poznámky o staročeském pasívu. Listy filologické 108, 1985, 77–82.
Štícha 1986a: ŠTÍCHA, F.: Predikativní status konstrukcí s n/t-ovými participii ve starší češtině.
Slavia 55, 1986, 384–390.
Štícha 1986b: ŠTÍCHA, F.: Systémový a funkční status konstrukcí s n/t-ovými participii v současné
češtině. Slovo a slovesnost 47, 1986, 3, 177–185.
Thieroff 2000: THIEROFF, R.: On the areal distribution of tense-aspect categories in Europe. In:
Dahl, Ö. [ed.]: Tense and aspect in the languages of Europe. Berlin – New York 2000, 265–305.
Topolinjska 1983: ТОПОЛИЊСКА, З.: За двојното потекло на конструкциите со има во
македонскиот литературен јазик. In: II Научна дискусија (Семинар за македонски јазик).
Скопје 1983, 25–33.
Topolinjska 1995: ТОПОЛИЊСКА, З.: Македонските диалекти во егејска Македонија, кн. 1:
Синтакса (Дел 1). Скопје 1995.
Topolinjska 2009: ТОПОЛИЊСКА, З.: Македонски јазик. In: Пипер, П. [ed.]: Јужнословенски
језици: граматичке структуре и функције. Београд 2009, 141–253.
Toporišič 2000: TOPORIŠIČ J.: Slovenska slovnica. Maribor 2000.
Townsend – Janda 2003: TOWNSEND, Ch. E. – JANDA, L. A.: Gemeinslavisch und Slavisch im
Vergleich: Einführung in die Entwicklung von Phonologie und Flexion. München 2003.
Trost 1972: TROST, K.: Perfekt und Konditional im Altkirchenslavischen. Wiesbaden 1972.
Trubinskij 1984: ТРУБИНСКИЙ, В. И.: Очерки русского диалектного синтаксиса. Лениград 1984.
Vasilev 1968: VASILEV, Chr.: Der romanische Perfekttyp im Slavischen. In: Koschmieder, E. –
Braun, M. [edd.]: Slavistische Studien zum VI. Internationalen Slavistenkongress in Prag 1968.
München 1968, 215–230.
Velkovska 1994: ВЕЛКОВСКА, С.: Сум- и има-конструкциите во светлината на балканскиот
јазичен сојуз – со осврт на македонските дијалекти во егејска Македонија. In:
Македонските дијалекти во егејска Македонија (Научен собир, Скопје, 23–24 декембри 1991).
Скoпје 1994, 15–21.
Velkovska 1998: ВЕЛКОВСКА, С.: Изразување на резултативноста во македонскиот стандарден
јазик. Скопје 1998.
168 Björn Wiemer

Weiss 1997: WEISS, D.: Syntax und Semantik polnischer Partizipialkonstruktionen (im Rahmen einer
generativ-transformationellen Sprachbeschreibung). Bern etc. 1977.
Wiemer 2004: WIEMER, B.: The evolution of passives as grammatical constructions in Northern
Slavic and Baltic languages. In: Bisang, W. – Himmelmann, N. P. – Wiemer, B. [edd.]:
What makes Grammaticalization? A Look from its Fringes and its Components. Berlin – New York
2004, 271–331.
Wiemer 2006: WIEMER, B.: Relations between Actor-demoting devices in Lithuanian. In:
Abraham, W. – Leisiö, L. [edd.]: Passivization and Typology (Form and Function). Amsterdam
– Philadelphia 2006, 274–309.
Wiemer 2011: WIEMER, B.: Grammaticalization in Slavic languages. In: Heine, B. – Narrog, H.
[edd.]: Handbook of Grammaticalization. Oxford etc. 2011, 740 –753.
Wiemer 2012: WIEMER, B.: The Lithuanian HAVE-resultative – a typological curiosum? Lingua
Posnansiensis 54, 2012, 2, 69–81.
Wiemer 2014: WIEMER, B.: Umbau des Partizipialsystems. In: Berger, T. – † Gutschmidt, K. –
Kempgen, S. – Kosta, P. [edd.]: Slavische Sprachen (Ein internationales Handbuch zu ihrer
Struktur, ihrer Geschichte und ihrer Erforschung), 2. Halbband; Reihe HSK. Berlin – Boston
2014, 1625–1652.
Wiemer – Giger 2005: WIEMER, B. – GIGER, M.: Resultativa in den nordslavischen und baltischen
Sprachen (Bestandsaufnahme unter arealen und grammatikalisierungstheoretischen Gesichtspunkten).
München – Newcastle 2005.
Wiemer – Seržant, forthcoming: WIEMER, B. – SERŽANT I. A.: Diachrony and typology of Slavic
aspect: What does morphology tell us? In: Bisang, W. – Malchukov, A. [edd.]: Unity and
diversity in grammaticalization scenarios. Open access series Studies in Diversity Linguistics (ed.
by Martin Haspelmath), forthcoming.

Institut für Slavistik, Turkologie und zirkumbaltische Studien (ISTziB) Björn Wiemer
Johannes-Gutenberg-Universität, Mainz [email protected]