Aspectual concepts across languages: Some considerations for second language learning

Abstract

In this paper, we focus on some terminological issues concerning the notion of aspect. We address the notions of grammatical aspect vs. Aktionsart, perfectivity vs. telicity, and imperfectivity vs. progressivity. We observe that these terms are often mixed up in the literature, which leads to some fundamental misconceptions in the theoretical description of different aspectual systems as well as in L1 and L2 acquisition studies. The descriptive approach we follow is strictly empirical and based on spoken production data. For our cross-linguistic comparisons, we draw upon data from native speakers of Czech, English, Dutch, German, and Russian. The theoretical framework of the paper is based on the idea that aspectual markers are not merely grammatical categories with a particular function, but more importantly they denote underlying cognitive concepts. These grammaticalized concepts determine native speakers' preferences in event construal, are language-specific (L1-based), and play a decisive role in second language learning. In order to deal with the difficulties arising in L2 learning, it is crucial to attempt to avoid terminological confusion. We think that this can be achieved by adopting a more conceptual and empirical approach to the analysis of aspect.

Full text

Aspectual concepts across languages:
Some considerations for second language learning
Barbara Schmiedtová & Monique Flecken
Abstract
In this paper, we focus on some terminological issues concerning the notion of aspect. We
address the notions of grammatical aspect vs. Aktionsart, perfectivity vs. telicity, and
imperfectivity vs. progressivity. We observe that these terms are often mixed up in the
literature, which leads to some fundamental misconceptions in the theoretical description of
different aspectual systems as well as in L1 and L2 acquisition studies. The descriptive
approach we follow is strictly empirical and based on spoken production data. For our
cross-linguistic comparisons, we draw upon data from native speakers of Czech, English,
Dutch, German, and Russian. The theoretical framework of the paper is based on the idea
that aspectual markers are not merely grammatical categories with a particular function, but
more importantly they denote underlying cognitive concepts. These grammaticalized
concepts determine native speakers' preferences in event construal, are language-specific
(L1-based), and play a decisive role in second language learning. In order to deal with the
difficulties arising in L2 learning, it is crucial to attempt to avoid terminological confusion.
We think that this can be achieved by adopting a more conceptual and empirical approach
to the analysis of aspect.
Keywords: grammatical aspect; Aktionsart; telicity; aspect terminology;
perfectivity; imperfectivity; progressivity; psycholinguistic reality;
empirical research; grammaticalization; conceptualization;
language typology; second language learning; false friends; Czech;
Russian; Dutch; English; German
1 Introduction
In this paper we will attempt to show and discuss some of the complexities in
terminology that regularly come up in theoretical analyses of aspect in cross-
linguistic research. Examples of terms that are often confused and that we focus on
are grammatical aspect vs. Aktionsart, telic vs. perfective, and imperfective vs.
progressive. In our view, this terminological confusion often leads to crucial
misconceptions with regard to the functional description of aspectual systems, the

2
way in which L2 acquisition of aspect is viewed, and also how it is taught in
schools and language courses. Obviously, our research is especially relevant for the
first part of pedagogical grammar, which is that of descriptive adequacy (see Ruiz
de Mendoza in this volume), but not for its final part, which is that of providing
improved teaching methods. We can merely present a number of relevant linguistic
issues and descriptions that we believe should be taken into account by applied
linguists writing pedagogical grammars.
Disregarding the discussion on the Critical Period hypothesis, one can state that
from a learning point of view it seems nearly impossible for advanced learners to
have full command of the aspectual distinctions in the target language (e.g.
Schmiedtová 2004; Slabakova 2005; v. Stutterheim and Carroll 2006). Equally
challenging appears to be the task of learning to express temporal relations in non-
aspect languages (for example German) by native speakers of aspect-dominant
languages (such as Czech or Russian). This is particularly evident in learners' ways
of structuring information in narratives (e.g. Schmiedtová and Sahonenko in press;
Carroll et al. in press).
The difficulties that second language learners of all proficiency levels face when
dealing with aspectual relations in the L2 arise partly because of the high
complexity and prominence of the aspectual systems as such and the differences
between the L1 and the L2 systems. But perhaps they also occur because traditional
analyses (e.g. Comrie 1976; Dahl 1985; Smith 1997) of the aspectual categories do
not provide the necessary guidelines for teachers to formulate instructions that
would make the acquisition of aspect more systematic and thus successful.
Learners have to gain competence not only in connecting the form and the
corresponding meaning(s), but also in making that connection on the basis of usage
principles in discourse. These kinds of competence have to be coherently integrated
within the learning process, which is a difficult task for both teachers and learners.
Our approach to investigating aspectual systems and their use in discourse is
entirely empirical. We base our claims and conclusions on spoken data produced
by native speakers and learners. Our framework reflects actual native speaker
preferences
1
for using aspectual markers in a particular language and, in addition, it
describes the internal organization of the respective aspectual system. In line with
the current trend in cognitive linguistics, we assume that aspectual categories do
not merely depict grammatical features, but that they also mirror conceptual
structures and hence have psycholinguistic reality. The aim of this paper is to
sketch some of the existing problems, increase awareness of them, and stimulate a
discussion. We will address several terminological issues by introducing an
empirically based approach to the classification of grammatical aspect providing
examples from L1 as well as L2 data. Our material includes data from Czech,
Dutch, German and Russian native speakers as well as from Russian and Czech
learners of German.
The structure of the present paper is as follows: in the next section we will discuss
a number of terminological fallacies, then show some empirical data in order to

3
support our claims concerning grammatical aspect, and finally draw our
conclusions with a couple of remarks regarding second language learning.
2 Aspect terminology
2.1 Grammatical aspect and lexical aspect (Aktionsart)
One of the frequently occurring problems in the literature on aspect is the lack of
uniformity concerning the theoretical notion of aspect. We distinguish between two
categories: grammatical aspect and Aktionsart
2
. The former aspect is a purely
grammatical category marked by inflectional morphology (e.g. affixes in Slavic
languages, the be V–ing form in English). In our approach, grammatical aspect (i.e.
in general the morphosyntactic marking of aspectual categories) denotes
grammatically encoded concepts. We agree with Klein (1994: 30) when he says
that: “... the convential ways of characterizing [grammatical] aspect, whilst
intuitively often appealing, are [apparently] not very satisfactory: they have much
more the status of metaphorical descriptions than of precise and clear definitions”.
These conventional ways include the terms 'viewpoints', 'viewing a situation from
the outside or the inside', 'situation is seen as completed/non-completed'. We do not
adhere to the convential view that grammatical aspect is a way of seeing situations,
which involves lexico-semantic as well as grammatical elements, because this view
does not provide a suitable theoretical framework for analyzing empirical data in
cross-linguistic context.
Aktionsart, by contrast, is a semantic category that expresses temporal
characteristics of verb meanings and meanings of verbal predicates (lexical
content). Several Aktionsart classifications have been proposed (e.g. Vendler 1967;
Smith 1997) but none of them are unproblematic. In our framework, we apply
Klein's classification from 1994. It is based on the Topic Time (TT) notion: Topic
Time (TT) is the time for which a particular assertion is made. For example, in She
was ill, the TT precedes the time of utterance (TT<TU) and thus holds true for a
time interval in the past, for which <a person be ill> applies.
According to Klein, Aktionsart expresses lexical contents and hence has no
direct connection to the time axis (1994: 99). The linkage to the time axis is
established through Topic Time. Depending on the number of Topic Times that are
contrasted in a clause, Klein distinguishes three different types of verbs/verbal
predicates:
• zero-states (no TT-contrast, e.g. The book is in Russian – no matter what
TT this clause is linked to, the assertion will hold true for any other TT);
• one-states (one TT-contrast, e.g. He was in Berlin – this assertion holds
true for one contrast between <not be in Berlin> vs. <be in Berlin>);

4
• two-states (two TT-contrasts, e.g. She opened the window – this
assertion holds true for two different contrasts <to be closed> vs. <to
open>, <to open> vs. <to be open>).
Despite the fact that Klein's framework does not make an explicit difference
between Vendler's accomplishment and achievement, we prefer to use Klein's
classification because Topic Time is a well-defined and transparent notion, which is
applicable to numerous languages other than English. In any case, the notion of
Aktionsart is less important for our research since our main focus is on aspect.
Usually, the notion of aspect comprises grammatical aspect as well as Aktionsart
(an exception to this trend is Bertinetto and Delfitto 2000). In line with this
misconception, some researchers assume that the acquisition of grammatical aspect
is always guided by learners' knowledge of semantic features encoded on the verb
(Aktionsart). In other words, it is believed that grammatical aspect is not acquired
independently, but must be accompanied or even preceded by knowledge of
Aktionsart. Hence, both categories are usually described and analyzed as a whole.
The original proposal goes back to Andersen and Shirai's Aspect Hypothesis
(also called Primacy of Aspect Hypothesis, Prototype Hypothesis, Aspect before
Tense Hypothesis) from 1994, on the basis of which they accounted for different
types of acquisitional data as well as Pidgin and Creole languages. Interestingly
this hypothesis seems to be tenable for numerous L1 and L2 varieties; it has
stimulated a lot of research related to aspect (for example, Li and Shirai 2000; Stoll
2005; Johnson and Fey 2006) and has been useful for stating initial research
hypotheses in the study of child language. It is important to realize, however, that
the correlation between lexical and grammatical features, be it aspect or tense, as
can be observed in first language acquisition, does not necessarily play a role in the
description of the underlying linguistic system. Researchers following the Aspect
Hypothesis have failed to adequately differentiate between lexical and grammatical
elements, which has led to a mix-up between semantics and the grammatical
categories of aspect and tense. For example early occurrence of
accomplishment/achievement verbs with past tense marker –ed in L1 English is
considered to represent the child's knowledge of perfectivity. This influential
hypothesis does not address the core issue of how to keep the notion of
grammatical aspect and Aktionsart apart, nor does it provide a systematic
description of these categories. We think that a suitable description of an
underlying aspectual system is a necessary prerequisite not only for our general
understanding of aspect and its acquisition, but also for developing appropriate
teaching methods.
Another shortcoming of the Aspect Hypothesis is that it formulates acquisitional
patterns for aspectual notions in contrast (i.e. the order of acquisition of perfective
vs. imperfective markers). When concentrating on English, which has only one
grammaticalized aspectual marker (the suffix –ing), it may indeed make sense to
set up an opposition between a verb marked for ongoingness (i.e. progressivity – I

5
am sleeping) and a verb inflected for past tense (He slept all day yesterday), and
label the latter as perfective. This seems to work because the 'perfective meaning'
(completion) arises here through the past tense morphology. However, note that
simple past in English is an aspectually unmarked form that is open to +/-
perfective interpretation. Thus, the verbal form in He slept is not inherently
perfective, but receives its 'perfective meaning' (completion) merely by pragmatic
knowledge. It could very well be the case that the person who slept yesterday is in
fact at the moment of speech still sleeping. This information is simply not part of
the temporal semantics of this utterance, and it is also not grammatically encoded
(this misconception is present e.g. in Slabakova and Montrul 2002).
In some other
cases, the pragmatic information is accompanied by lexical features of the verb as
in She broke my arm or of the verbal predicate as in He ate up his sandwich. In
these examples, the verbs including their arguments inherently express a change of
state, which makes the 'perfective' reading possible (aided by the past tense
morphology). But again, the verbs are not marked for perfectivity by means of
aspectual inflectional morphology.
Simple forms in English, whether in past or present tense, are with regard to
grammatical aspect open (neutral or unspecified). Since these forms lack the
ongoing marker altogether they can be called 'non-progressive', but they do not
express any aspectual meaning that is contrastive to progressivity. Simple forms in
the present tense in English have many different meanings, but most of these
meanings arise through the linguistic context (e.g. through the addition of adverbial
phrases) or a specific speech act (e.g. an informative act). These meanings are
conveyed by lexical and not grammatical features, and therefore the several
different meanings that English simple forms can have - habitual, scientific present,
etc. – do not represent an aspectual opposition to the progressive aspect. As will be
explained below (see Section 2.2.) the most prominent meaning – habituality –
arises only in specific contexts and is, as we hypothesize, the result of the
grammaticalization process of the progressive marker. By grammaticalization we
mean the process in which grammatical morphemes gradually develop out of
lexical constructions and become more and more used as fully-fledged
constructions in an ever-expanding range of contexts. These grammatical
constructions are becoming part of the core grammar of a language (cf. Bybee et al.
1994). For English it is true that in certain contexts the simple form can convey 'a
holistic viewpoint', for example in He reaches the finish. Note, however, that this
meaning, in contrast to the meaning of the –ing form, are not grammaticalized and
belong to the lexico-semantic and not the grammatical area.
All this is very different from languages that use two grammaticalized aspectual
markers whose meanings are truly contrastive. All Slavic languages, for example,
can express both meanings – perfective and imperfective – grammatically on the
verb. Although these systems do not apply to all verbs and there are some
exceptions to the rule, we see a fundamental difference between the Czech/Russian
and the English systems. There is an opposition between two different aspectual

6
categories – perfective vs. imperfective – in Slavic languages, neither of which is
expressed by past tense marking, whereas no such grammatical opposition exists in
English (only the progressive is grammaticalized in English).
We believe that this mix-up has been dominating and partially misguiding the
overall discussion about aspect typology and acquisition (see for example, general
aspect analysis: Verkuyl 1993; Smith 1997; acquisition: Stoll 1998; Wagner 2006).
An exception to this trend is Slabakova’s review of recent research on the
acquisition of aspect (2002). Slabakova (2002: 176) points out that many studies
have blended three different temporal contrasts, that is past vs. present tense,
perfective vs. imperfective grammatical aspect, and Aktionsart distinctions.
In fact, we would say that fully grammaticalized grammatical notions (e.g.
grammatical aspect) only interact (but do not merge) with other temporal
categories, such as tense, adverbials, or Aktionsart. To a large extent the English
progressive marker –ing is the ideal example of such a fully grammaticalized and
independent grammatical category. Adopting our view makes it possible to tear
apart grammatical aspect, Aktionsart, and tense and it would possibly improve
teaching methodologies because teachers would be able to explain these categories
in a more systematic and independent way.
2.2 Telicity vs. perfectivity
Another problem we would like to tackle is the confusion between the terms telic
and perfective. Similar to the issues discussed above, this problem too is related to
an inaccurate differentiation between Aktionsart and grammatical aspect. In our
view, the notion of telicity belongs to the domain of lexical features inherent in the
verb/verbal predicate while perfectivity is a grammatical category. We define telic
verbs or telic verbal predicates as expressing an inherent endpoint, which must not
necessarily be realized in a situation (e.g. to fall, to write a paper). It is in principle
plausible to assume that all languages have verbs expressing +/- telicity. However,
only a number of aspect-prominent languages can convey +/- perfectivity
grammatically. In other words, although the two terms are closely related in
meaning and can interact with each other at the level of expression, they involve
two different layers of linguistic analysis and are hence not synonymous. To
illustrate this difference let us consider the following examples from English and
Czech.
(1) He ate an apple
In example (1) the verb to eat is a one state verb denoting only one change of state
(Klein 1994) and for the sake of argument we assume that together with the
indefinite object an apple it forms a telic predicate. The same utterance in Czech is
presented in example (2).

7
(2) (On) S-něd-l jablko
He-Nom Perf-eat-Past-3sg apple-Acc
‘He ate (an) apple’
In Czech the verb jíst 'to eat' is also a one state verb, but unlike in English, in the
example above it occurs as a perfective, marked grammatically by the prefix s-. As
in English, we are dealing here with a telic predicate – to eat an apple, but the verb
is overtly marked as perfective. So the Czech utterance involves two different
features: telicity on the predicate plus the perfective aspect inflected on the verb.
Only the first feature is present in the English example. As discussed briefly above,
we can see that telicity and perfectivity involve two different operations, yet, in
these examples they result in a comparable semantic structure: having reached the
right boundary of the situation, i.e. the endpoint of the situation.
In spite of this parallelism if we change the tense of the English verb from past to
present we observe a shift from (-) aspect to (+) aspect. More precisely, from (-)
progressive to (+) progressive as in example (3).
(3) He is eating an apple
If the aspect is not changed from (-) to (+) progressive, as in (4), the meaning of the
utterance becomes problematic.
(4) ?He eats an apple
In (4) the tense change makes the utterance ill formed in contexts of ongoing
situations because of the conflict between the presence of an object and the simple
present. Normally this combination renders a habitual reading, but then further
temporal specification (e.g. He eats an apple every day) or a particular context
conveying the habituality (e.g. What does your diabetic friend do when he suffers a
hypo?) is required. It is true that English simple forms often denote habituality, but
we strongly believe that this is merely a consequence of the grammaticalization of
the -ing form
3
. In itself, the simple form does not convey habitual meaning
grammatically. This can be seen in example (4), where habitual meaning only
arises when specific habitual contexts are provided, i.e. either lexical devices
(temporal adverbials) or context.
Another option for making (4) grammatical is to change the simple verb form
into the progressive: He is eating an apple, as in (3). That means that in English a
change in tense goes hand in hand with a change in aspectual value: The addition
of the –ing suffix (or a temporal adverbial) is obligatory in a context of
ongoingness in the present tense.
What is relevant here is that despite the change in aspect, the telicity of the
English predicate remains unaffected. The Czech example in (5) demonstrates that

8
a shift in tense does not influence the aspectual value, nor the telicity of the
utterance.
(5) (On) S-ní jablko
He-Nom Perf-eat-Present-3sg apple-Acc
‘He eats (an) apple (up)’
The interpretation of example (5) is that the situation to eat an apple in Czech is
presented as inevitably reaching its endpoint in a very near future. This is very
unlike the English predicate, which is telic (a semantic category), but by no means
perfective (a grammatical category). In other words, by using a perfective prefix a
Czech language user conceptualizes and presents the situation depicted in (5) as
perfective. In principle, the Czech aspectual system allows the expression of
perfectivity in the present tense
4
, which is not possible for English. As shown in (3)
and (4), it is compulsory in English to use the progressive in here-and-now
contexts. This shows that only the progressive aspect has been grammaticalized in
English.
In Czech, on the other hand, verbs must be marked either for perfectivity or
imperfectivity in all tenses. This is because both aspects have been
grammaticalized. The English aspectual system, by contrast, does not contain a
systematic opposition between two different grammatical aspects: the "perfective"
interpretation of verbal predicates such as to eat up is not brought about
grammatically (perfectivity) but it is conveyed lexically by the particle up (telicity).
In this sense, perfectivity does not equal telicity.
It leads to fundamental problems when the unspecified simple form in the
context of telic verbs/verbal predicates is put in opposition to the aspectually
marked progressive form. As we will show in more detail below (Section 2.4),
Slavic languages have both poles of this aspectual contrast at their disposal and
thus represent a completely different system with not only different forms, but also
with different underlying concepts. With respect to learning, the difficulty arises
when teachers draw parallels between a marked perfective and the English simple
form: these are basically false friends.
2.3 Imperfectivity
The last terms we would like to attend to in Section 2 are the notions of
imperfectivity and perfectivity. We will first address the former category. Slavic
languages use simplex forms to express imperfectivity (e.g. in Czech psát – ‘to
write’) to express imperfectivity apart from the marked imperfective, the so-called
secondary imperfective. The secondary imperfective is marked by inflectional
morphology, that is, in Czech the suffix -(o)va-, and in Russian the suffixes -iva-/-
yva-, -va-, -a-/-ja- (e.g. in Czech/Russian vypis-ova-t/ vy-pis-yva-t' – ‘to be in the

9
process of writing out’. ). There is also a small group of frequently used simplex
verbs denoting perfectivity without an explicit morphological marker (e.g. Czech
dát - 'to give'). Because these verbal forms lack any overt grammatical marking of
their aspectual value, a question arises. Does their aspectual meaning come from
the inherent verbal semantics (Aktionsart) or is it rooted in the grammar
(grammatical aspect)? Despite this serious terminological problem, which has not
yet been thoroughly investigated
5
, we hypothesize that Slavic simplex forms differ
from those in English, German, or Dutch. A possible justification for this line of
thinking is the following. Usually, adding a prefix
6
to a Czech simplex
imperfective verb results in changing the aspectual features into the perfective, as
in example (6).
(6) Prefixation of the simplex imperfective form
Czech pít  VY- pít
IMPF-simplex PERFdrink
English 'to drink'  ‘to PERFdrink’ ≈ 'to drink up'
The situation is different when dealing with simplex perfective verbs (7).
(7) Prefixation of the simplex perfective form
Czech dát  U-dat
PERF-simplex PREF-PERFgive
English 'to give'  ‘to report’
In example (7), the prefix u- only changes the meaning of the verb, but not the
aspectual value. That is, the verb remains perfective and a new lexical entry is
derived. Another relevant point to be mentioned here is that simplex perfective
forms, such as dát 'to give', can only be used in perfective contexts. For generic
and imperfective contexts the marked imperfectivised form – dá-VA-t – must be
employed. This shows that the simplex form has an aspectual value – the
perfectivity – on its own.
Because of these observations we theorize that the perfective value is already
encoded in the stem of the verb regardless of the lack of overt marker(s). There are
no comparable cases in English, German, or Dutch. Therefore, we argue that (a) in
contrast to English, simplex forms in Slavic languages have a default grammatical
aspect (in addition to their inherent Aktionsart), and (b) that simplex forms in
English, German, or Dutch only make use of Aktionsart and are underspecified
with respect to grammatical aspect. Turning back to L2 learners of Slavic
languages, the dichotomy in the domain of simplex forms must pose a learning
challenge since simplex forms are unmarked by default, nevertheless they carry an
unambiguous aspectual meaning. Because of this, we are again dealing with a kind
of false friend when translating (and teaching) the Czech dát as English ‘to give’.

10
The next section will focus on some difficulties in characterizing and defining the
notions of imperfectivity and perfectivity across languages.
2.4. Perfectivity vs. imperfectivity: conceptual differences
This section focuses on the comparison between two binary aspectual systems: the
Czech and the Russian systems. Although these two Slavic languages show many
typological similarities, our research (e.g. Schmiedtová and Sahonenko in press)
shows that in the aspectual domain there are crucial differences in native speakers’
preferences, as well as in the distribution of the forms within the system. These
differences may pose a real challenge to L2 learners.
As stated above, both languages encode two contrasting grammatical aspectual
categories: the perfective and the imperfective. Both languages also use a number
of simplex verbs, but in what follows, we will only focus on grammatically marked
aspects. In principle, there are two operators that can change the aspectual value of
a verb. The first operation is adding a prefix to the verbal stem. These prefixes do
not only change the grammatical aspect, but they can also affect the semantics of
the verb, i.e. derive a new lexical item. Moreover, with some verbs it is only the
lexical meaning that changes. So, the challenge here is that the lexical and the
grammatical modification can hardly be separated from one another (Comrie 1976;
Schmiedtová 2004).
The other operation is adding a suffix. Suffixation leads to secondary
imperfectivization of the verb (regardless of the type of verb stem) and the change
is mainly grammatical (from perfective to imperfective aspect). These claims hold
true for Russian as well as Czech. Let us consider a couple of examples.
(8) Prefixation of the simplex imperfective form
Czech: psát Russian: pisat’
IMPF-simplex
Czech: VY-psa-(-t) Russian: VY -pisa(-t’)
PREF-writePERF
English: ‘to write out (all keywords)/ to announce (a job)’
In (8) a simplex imperfective is turned into a perfective by the prefix vy-, which
changes the meaning. Note that one and the same operator affects two linguistic
areas: lexicon and grammar.
(9) Suffixation of the simplex perfective form
Czech: dát Russian: dat’
PERF-simplex
Czech: dá-VA(-t) Russian: da- VA(-t’)
PERF-give-2ndaryIMPERF

11
English: ‘to be giving’
The point of example (9) is to illustrate the change of a simplex perfective to a
marked imperfective verb (i.e. change in grammatical aspect only).
(10) Suffixation of a prefixed perfective form
Czech: VY-psat Russian: VY-pisat’
PREF-writePERF
Czech: VY-pis-OVA(-t) Russian: VY –pis-YVA(-t’)
PREF-2ndaryIMPERF
English: ‘to be writing out (all keywords)/ to be announcing (a job)’
The same suffix (-(o)va/-(y)va) can be attached to a prefixed verb denoting
perfectivity. As in (9), the suffix in (10) also changes the grammatical aspect.
The question to ask here is: what are the conceptual consequences of these
operations? We do not completely adhere to how perfectivity and imperfectivity
are usually described in the literature (for example Langacker 1987, Bybee (1992:
144)): “... perfective, which indicates that the situation is to be viewed as a
bounded whole, and imperfective, which in one way or another looks inside the
temporal boundaries of the situation ...”.
We want to be more specific and claim that the crucial difference between the
perfective and imperfective is the degree of focus on the right boundary of a
situation (e.g. in the situation, in which a person is drinking up a glass of water, the
right boundary is reached when the glass is empty and the person is in the post state
of having finished a glass of water). That is, the function of the perfective in Czech
and Russian is to encode that a situation has reached its right boundary and also
that an assertion is made about the possible post state of this situation (speaker's
focus is on the right boundary). This is illustrated in Figure (1):
Figure 1: Scope of the perfective aspect in Slavic languages
In contrast, the secondary imperfective accesses the time interval prior to the right
boundary, but (!) does not ignore the right boundary of the situation altogether,
rather the secondary imperfective defocuses it, as in Figure (2):
Right Boundary
Post State
PERFECTIVE

12
Figure 2: Scope of the secondary imperfective aspect in Slavic languages
So, in both instances, the perfective as well as the imperfective aspect, the attention
centers around the right boundary. This view of the imperfective aspect puts the
frequently assumed similarity between the progressive (e.g. in English or Dutch)
and the imperfective into question. Even though such a comparison might be
linguistically interesting, our analyses show that the two aspectual operations are
very different (for more details, see Section 3).
In the next section we will provide empirical evidence for the conceptual
differences between the Czech and the Russian aspectual systems, as described
above.
3 Underlying concepts in cross-linguistic comparison: Empirical data
In this section, we will first explain, using production data from Czech, Dutch, and
Russian native speakers as well as advanced L2 learners of these languages, that
grammatical aspect is not only a matter of grammatical form, but also of
conceptualization. This conceptual structure is reflected in the language-specific
preferences of native speakers when using different aspectual forms in their L1, as
well as reflected in the overall degree of grammaticalization within each system.
This new take on analyzing aspectual distinctions is pursued by our research group
at the University of Heidelberg and finds its origin in research conducted by
Christiane v. Stutterheim and Mary Carroll.
We base our analyses and description of aspectual systems on production data
elicited from large samples of native speakers as well as learners. The experimental
approach consists of an online production task, in which speakers (N=30) are asked
to retell short everyday situations in answer to the question What is happening?
(translated into all relevant languages), i.e. video clips depicting somebody
drinking a glass of water, a dog running into a house, etc. In order to test our
hypotheses, we make use of several sets of stimuli that are grouped according to
situation type (e.g. causative actions, locomotions with +/- endpoints, etc.). This
approach forces speakers to choose a particular aspectual form, which is
Defocused Right Boundary
IMPERFECTIVE

13
appropriate or obligatory for a specific situation type. To strengthen our arguments
we also use other methods, such as speech onset times and eye-tracking
measurements. By adopting this line of empirical research, combining linguistic
analyses of production data with psycholinguistic methodology, we believe that we
are able to tap into speakers' conceptualization patterns
7
.
The focus of the previous studies (e.g. Carroll and v. Stutterheim 2003; v.
Stutterheim and Carroll 2006; Klabunde and v. Stutterheim 1999) was on Semitic,
Germanic, and Romance languages. It has been shown that the way events are
depicted is highly dependent on the feature +/- grammatical aspect. It has also been
found that the underlying principles for event construal are perspective driven and
strongly linked to patterns of grammaticalization. Additionally, recent L2 studies
have provided evidence that even very advanced learners fall back on
conceptualization strategies from their L1 when construing temporal events in a L2
(cf. v. Stutterheim and Nüse 2003; Schmiedtová and Sahonenko in press). These
findings also hold true for learners describing situations that are more complex than
single events such as narratives. Carroll and Lambert (2003) have shown that the
use of aspectual categories influences the overall information structure in more
complex tasks, such as composing written or oral narrative texts. The next sections
will deal with conceptual representations that underlie the grammaticalized
aspectual categories of Czech, Russian, and Dutch.
3.1. Differences between the Czech and the Russian aspectual systems
In the previous section we have discussed the concepts that are encoded by the two
aspectual categories: the perfective and the imperfective. In what follows, we will
present production data of native speakers of Czech and Russian (for both N=30).
We will see that, despite very similar underlying aspectual systems, Czech and
Russian native speakers have different aspectual preferences when construing
events in their mother tongue. These preferences, as we believe, reflect the way in
which these speakers view and conceptualize a situation.
First of all, in both the Czech and Russian native speaker data, there is a
pronounced tendency to relate events to the right boundary (for Czech: 87% of all
speakers; for Russian 77% of all speakers). This means that speakers mark an
evident right boundary or endpoint of a situation as depicted in the stimulus. The
stimuli set consisted of two types of scenes: in type one, the right boundary of a
situation was visible in the clip and actually reached; in the other type, only a
potential right boundary could be inferred but it was not depicted as being reached
in the clip. The difference between the ways in which native speakers of Czech and
Russian verbalized stimuli of the second type lies in the fact that Czech native
speakers mention the endpoint more frequently (for Czech: 65% of all speakers; for
Russian 25% of all speakers). In addition, Czech native speakers use the perfective
form, independent of the scene type. Russians, on the other hand, showed a clear

14
preference for using the secondary imperfective in all scenes. When they used the
perfective form it was exclusively for scenes showing the right boundary being
reached.
In other words, speakers of different languages follow different preferential
patterns when they encode events. We believe that these preferences which so far
have been described from a linguistic point of view (i.e. surface structure) are
rooted in differences in conceptualization of events. In one and the same stimulus,
Czech native speakers concentrate on the time interval at and after the right
boundary whereas Russian native speakers are sensitive to the time interval
preceding the right boundary.
At the same time, the data show that the distribution of the aspectual forms within
each system differs, too. That is, in Russian the imperfectivizing suffix -(y)va is
productive and can be applied to many verbs. In Czech, by contrast, this suffix only
combines with a small group of verbs. Additionally, as pointed out by Schmiedtová
(2004), the perfective form, when used in the present tense, can have a here-and-
now meaning in Czech. This is completely impossible in Russian where the present
perfective always refers to the future. We hypothesize that in Czech the increased
use of the perfective form goes hand in hand with the prominence of the underlying
conceptualization (as depicted in Figure 1). In other words, the extensive use of the
perfective aspect in Czech results in a perspective focusing on the right boundary
of a situation and/or its post state. The same logic applies to Russian, where the
frequent use of the secondary imperfective goes hand in hand with the imperfective
perspective (i.e. focus on the time interval preceding the right boundary without
excluding it completely). It remains an open question, however, in what direction
this influence takes place. The relevant point here is that despite big similarities
between the two aspectual systems, Czech and Russian native speakers differ
considerably as far as their aspectual preferences are concerned.
In summary, our experimental data show that there is interplay between
grammatical categories and conceptual structures. Furthermore, we see that even
speakers of typologically related languages display different conceptually driven
perspectives (preferential patterns) when selecting information for event construal.
With respect to L2 learning, in Schmiedtová and Sahonenko (in press) we showed
that advanced Czech and Russian learners of German adhere to their respective L1
preference. For example, Czech learners use the concept of perfectivity in L2
German although German does not have grammatical aspect at all. The adherence
to this concept becomes apparent in more frequent mentioning of endpoints in the
form of local adjuncts (e.g. into the house) when retelling video clips depicting
locomotions with ± endpoints. Even though German native speakers are also
inclined to mention endpoints frequently (as pointed out in e.g. v. Stutterheim and
Lambert 2005), the number of endpoints verbalized by L1 Czech speakers of
German exceeds the average for German native speakers. This is a relevant finding
because it illustrates that patterns found for native speakers for event depiction in
their native language still drive the perspectivization in L2 production. This

15
important issue presents a considerable challenge to language teachers, since, for
learners, being aware of the meanings of various aspectual categories is a good
starting point for achieving native-like competence in a second language.
3.2 Progressive in English and Dutch: grammaticalization and conceptual
structure
This part of the paper is devoted to the Dutch language. This is because in Dutch
the progressive marker aan het + V-INF zijn is currently being grammaticalized
(Flecken 2006). We are aware that a truly grammaticalized aspectual marker is
morphological in nature and that the Dutch marker is still a periphrastic
construction. However, we speculate that in the course of the grammaticalization
process it will be reduced to a verbal morpheme. This seems to be already
noticeable when considering native speakers' shortened pronunciation of this
construction.
Because the Dutch grammatical system is in the middle of this process, we
envisage that learners are confronted with the hard task of figuring out how the
system operates. We will first present some empirical data illustrating the range of
applications of this marker. Furthermore, we will show that the range is expanding,
following the grammaticalization process described in Bybee et al. (1994), which
motivates our focus on verb type. We will briefly discuss some differences
between the Dutch construction aan het + V-INF zijn and the German construction
am + V-INF sein and we will draw parallels between the Dutch and the English
progressive marker. Finally, we will demonstrate that progressivity and
imperfectivity denote two different temporal concepts.
First of all, it is necessary to define our notion of grammaticalization. As
mentioned above, grammaticalization means expansion of the range of contexts in
which a particular construction is applied. The starting point for
grammaticalization is the use of a particular construction in its prototypical lexical
environment. This use is inherently linked to a specific meaning of the grammatical
feature, which slowly spreads out to less prototypical uses/contexts (Comrie 1976;
Bybee et al. 1994).
Regarding the meaning of the Dutch progressive marker, in our data we observe
that modifying a Dutch verb with the aan het-construction depicts situations as
ongoing, as in example (11).
(11) Ik ben aan het lezen
'I am reading'
The aspectual marker in (11) defocuses both the initial and the final boundary of
the situation and hence the temporal reference applies only to the here-and-now.
The meaning of the Dutch aan het-construction is, therefore, identical with the

16
meaning of the English –ing, which has the same function. Let us take a closer look
at the similarities between Dutch and English.
At first sight, the Dutch marker looks like a locative construction because of the
locative aan (like the English prepositions at/on) (Boogaart 1999). Interestingly,
the English progressive marker might have evolved out of a locative construction
as well. This original construction looks similar to the contemporary Dutch
periphrastic construction (12) (example taken from Bybee et al.: 132).
(12) He is on hunting
‘He is hunting’
Comparing (11) and (12), we can see that the original meaning of both
constructions could have been ‘to be in the place of doing something’. This
originally locative meaning evokes a very deictic here-and-now context, and we
assume that, in a way, this condition was the starting point for the
grammaticalization of the –ing form (also in Jespersen 1949; Comrie 1976). We
claim that it is also the starting point in the grammaticalization process of the aan
het-construction in current Dutch. In English, we see that this precondition is no
longer necessary for application of –ing, as is apparent when looking at examples
(13) and (14).
(13) Katja is having an affair with Christopher
(14) Doro is practicing law
The meaning of the -ing form in (13) and (14) is not necessarily restricted to the
deictic (locative) here-and-now, but it is extended over a longer period of time (as
in (13)), and it can even describe a habitual feature (as in (14)).
In Dutch, this type of application of the aan het-construction is not (yet?)
possible. The meaning of this construction mostly refers, at this point in time, to
agentive subjects who are in the midst of an activity at reference time or in the very
deictic past as in (15a) and (15b).
(15a) Ik ben aan het werken
‘I am working’
(15b) Gisteren was ik aan het studeren
‘Yesterday, I was studying’
We presume that in contrast to English –ing, the use of the Dutch construction in
true habitual contexts is more constrained
8
. The traditional view of the aan het-
construction in Dutch literature is that it is merely “a locative construction with a
“progressive-like” meaning” (e.g. Boogaart 1999: 167). This view, however,
does

17
not take into account that the Dutch grammatical system is evolving and hence the
progressive marker is becoming part of the core grammar.
In our view, we take the above observations to mean that the Dutch progressive
construction is at the onset of a similar grammaticalization process but that, the
English progressive marker is in a far more advanced stage within this process.
This has been shown empirically: in our data English native speakers, when
construing events, use the –ing in all cases whereas the simple form is completely
absent. That is, all native speakers of English (N=60) in our sample resort to the
progressive marker when asked to tell what is happening or even what happens.
Going back to the Dutch language, in order to sketch a more accurate
development of the aan het-construction, we focus on the types of verbs
(Aktionsart, in line with Klein 1994) that take the marker aan het (in line with
Bybee's approach to grammaticalization).
The first step of grammaticalization, thus the prototypical context for using
progressive markers, is to use it in situations denoting an activity, e.g. wandelen
(‘to take a walk’), zwemmen (‘to swim’), but also een boek lezen (‘to read a book’),
de tafel poetsen (‘to clean the table’). In the prototypical phase, the prerequisite for
using the aan het-construction is the possibility of defocusing boundaries. All
predicates that inherently refer to one of the boundaries (such as to fall) do not
combine with the aan het marker at this stage of grammaticalization. The verb type
which meets all these conditions is the one state verb, such as zwemmen ‘to swim’.
In the next grammaticalization phase, the two state verb referring to a rather long
time span is included (e.g. veranderen ‘to change’) followed by the two state verb
denoting a short time interval (e.g. breken ‘to break’). The last step is the
expansion to zero state verbs, such as houden van ‘to love’. Interestingly, in
English the grammaticalization process of the –ing suffix has reached this last
phase: It is grammatical to say I am loving it (in the sense of ‘I am enjoying it’) or
She is having a baby (although they have two different temporal meanings).
To illustrate this process for Dutch, we present some preliminary results of an
acceptability judgment task using a five degree scale ranging from completely
acceptable (5) to completely unacceptable (1). We asked 30 Dutch native speakers
to make a choice between a simple verb form and a verb marked by an aan het-
construction in here-and-now contexts. We differentiated between the four types of
verbs described above: one state verbs, two state verbs with long and short
duration, zero state verbs. It turned out that one state verbs (e.g. lezen ‘to read’,
tekenen ‘to draw’, schilderen ‘to paint’, knutselen ‘to tinker’, pianospelen ‘to play
the piano’) triggered the most frequent use of the aan het construction. The second
best attractor for aan het was the two state verb with a long duration
9
(as in
afmaken ‘to finish’, afwassen ‘to do the dishes’, veranderen ‘to change’), followed
by the two state verb with a short duration, e.g. vallen ‘to fall’, exploderen ‘to
explode’, breken ‘to break’. The zero state verbs did not elicit any choices for the
aan het-construction in the here-and-now-context.

18
As far as acceptability is concerned, this task has allowed us to interpret the
values that the participants attached to the form they did not choose. They always
had to grade the other form in terms of its acceptability in a given context. The
most important finding was that participants rated the simple form as unacceptable
in here-and-now contexts for the verbs expressing a game-like activity, examples
of which are zwemmen ‘to swim’, tafeltennissen ‘to play table tennis’, schilderen
‘to paint’. Moreover, they rated the aan het form as unacceptable in clauses with
motion verbs plus a depicted endpoint (as in *Ik ben in het water aan het springen
‘I am jumping into the water’). These results make sense: The latter verb type
expresses the shortest possible duration, namely the time interval right before
reaching the final boundary, which makes defocusing of boundaries quite
impossible.
A further interpretation of these results is that in a number of cases the aan het-
construction was considered compulsory by the participants. As pointed out above,
this is the case for situations expressing activities taking place in the here-and-now.
The simple form in these cases was rated unacceptable because using the simple
form renders a habitual meaning in these contexts. For example, following the
question Wat ben je aan het doen? ‘What are you doing?’ all Dutch native speakers
in our sample choose the aan het form in combination with one-state verbs, e.g. Ik
ben aan het werken ‘I am working’. The simple form, Ik werk ‘I work’, is rated as
completely unacceptable in such contexts. In summary, when activity verbs and
verbal predicates are used in a here-and-now context the aan het marker is
obligatory.
Again, this is comparable to English, because the difference between I am
dancing and I dance is that the former implies an activity that is taking place at the
time of utterance; whereas the latter refers to a habitual activity (a hobby or
perhaps even a job). Bybee et al. (1994) label this phenomenon as
grammaticalization of zero (i.e. the unmarked form receives a different meaning in
certain contexts). Of course we realize that the depiction of the grammaticalization
process is rather different from the question of what the actual attractors are for
using the aan het-construction. It cannot solely depend on the verb type, but will
rather be a matter of the entire predicate.
10
An interesting comparison to draw at this point is between Dutch and German.
Though both languages are typologically similar, one important difference is that
Dutch is grammaticalizing a marker for ongoingness, whereas in German
ongoingness is mainly expressed by lexical means. German has a construction,
which is form-wise very similar to the Dutch one. Consider example (16).
(16) GER: Rieke ist (gerade) am/beim Kochen
NL: Rieke is aan het koken
ENG: ‘Rieke is cooking’

19
The German periphrastic construction is merely a regional and stylistic variant of
Standard German while in Dutch it is an obligatory marker in such a context,
compared to the unmarked simple verb form. Furthermore, the progressive markers
in English as well as Dutch are systematically used by native speakers for the
expression of other temporal concepts, such as the expression of simultaneity
between two events in present tense. The German construction is never produced in
such contexts (see Schmiedtová 2004; Flecken 2006).
Looking at these similarities from a learner's point of view, we have another
occurrence of false friends. Learners have to deal with two very similar forms that
do not show a similar distribution across verbs and, in addition, are employed by
speakers for different purposes.
The last point to be addressed in this section is the difference between progressive
and imperfective aspect. As we have shown in Section 3.1, speakers of Slavic
languages do not ignore the right boundary of the depicted situation when using the
marked imperfective, but rather include it in their conceptualization and
verbalization of situations. In other words, by using this form speakers refer to the
time interval anchored in the here-and-now and to the linkage of this time interval
to the right boundary. The Dutch and the English progressive, by contrast, are used
to link situations to the deictic here-and-now without any explicit temporal
information about the right (or left) boundary. The progressive marker merely
expresses ongoingness. This is especially true in Dutch where the
grammaticalization process of the aan het marker has started out exactly from this
context.
To relate this observation to the conceptualization of temporal events, we know
from eye-tracking studies that Dutch and English speakers concentrate only on the
ongoing process of situations regardless of whether they depict a right boundary (v.
Stutterheim and Carroll 2006; Carroll et al. in press). We speculate that Slavic
speakers, when using the secondary imperfective to describe ongoing situations of
the same type as above, will also pay attention to the right boundary.
To conclude this section, it is important for researchers, teachers and learners to
take into consideration the conceptual differences between the imperfective and the
progressive aspects.
4 Conclusions
The present paper centers around the idea that the analysis of grammatical aspect
contains at least the following two different areas: the form and the meaning.
Another idea is the usage and applicability of aspectual forms in context that are
determined by the preferences of native speakers.
When investigating aspectual forms cross-linguistically many similarities can be
observed. The tricky issue is, however, that the mere existence of a form in a sytem

20
or similar forms across systems does not necessarily entail an equally frequent
production. To this end, we have demonstrated on the basis of a comparison
between German and Dutch that similar forms with comparable meanings do not
show the same distribution in native speakers' production. The same holds true
for the language pair Czech and Russian. Despite the similarities between the two
aspectual systems, Czech and Russian native speakers show different preferences
for applying aspectual forms. These preferential patterns are closely linked to
differences in conceptualization, which only become evident when examining
empirical material collected by means of experimental methods.
The second area of analyzing aspectual systems is meaning. We have claimed that
categories such as progressive and imperfective aspect, albeit applicable in
comparable contexts, encode different temporal concepts. Again the same
statement holds for the terms telic and perfective. They too are not interchangeable
and, in addition, belong to two different kinds of aspect: lexical (telic) vs.
grammatical (perfective).
Note that even when two forms and their temporal meanings are very similar
there can still be a difference with respect to the conditions under which these
forms can be employed. This is directly connected to the degree of
grammaticalization of the respective aspectual form. This has been presented on
the basis of the progressive markers in English and Dutch.
Another point to be mentioned here is that many divergences pointed out in this
paper do not only occur between typologically distinct languages (such as Russian
and German), but also between languages that are typologically closely related
(e.g. languages within the Slavic or Germanic group). To summarize the
differences between the different languages that we addressed, consider Figure 3.
Language Czech/Russian English Dutch German
Form suffixes/prefixes -ing aan het +
V (inf) zijn am/bei +
V (inf) sein
Temporal
Function ± reaching of the right
boundary defocusing
boundaries defocusing
boundaries defocusing
boundaries
Term imperfective/perfective progressive progressive ??
Degree of
Grammatica-
lization
both aspects fully
grammaticalized fully
grammaticalized in the process of
being
grammaticalized
not in the
process of being
grammaticalized
Figure 3: Overview aspectual devices in different languages
These observations are highly relevant for teaching and learning. It is reasonable to
assume that to focus on form is the least complex approach to teaching aspect,
although we have illustrated that even in this area false friends can be identified.
As far as meaning is concerned the issues are equally serious. Several aspectual
categories that we dealt with are used synonymously in the literature, even though
they denote semantically and conceptually different entities.

21
Now, what about L2 learning? It is true that at the onset of acquisition false
friends can aid and support the learning process. Looking at advanced learners, on
the other hand, provides a considerable piece of evidence that false friends hinder
learners in the possibility of achieving nativeness (e.g. English learners of Czech in
Schmiedtová 2004). Note that advanced learners are in perfect command of the
aspectual forms and even their meaning (i.e. they do not make any grammatical
errors), but they do not successfully (not in a native speaker-like fashion, that is)
use the principles that govern the application of the forms. In other words, they do
not follow native-like preferences, but rather rely on patterns of use from their
respective L1s (research conducted by our group in Heidelberg).
We are not sure whether these preferences can be learned at all (for a discussion
of the feasibility of ultimate attainment, see e.g. van Boxtel 2005). Nevertheless, it
is essential to attempt to encourage the learning of aspectual distinctions as a
whole. That means that linguists and language teachers have to realize that the
debate on aspect is not only a matter of terminology, but that aspect is a conceptual
category that requires empirical research. We believe that the approach to view
aspect as a conceptual category and to adhere to empirical research when
investigating this linguistic domain would be beneficial to teachers as well as
learners.

22
References
Andersen, Roger and Shirai, Yasuhiro
1994 Discourse motivations for some cognitive acquisition principles. Studies
in Second Language Acquisition 16: 133-156.
Bertinetto, Pier Marco and Delfitto, Denis
2000 Aspect vs. Actionality: Why they should be kept apart. In Tense and
Aspect in the Languages of Europe, O. Dahl (ed.), 189-225. Berlin:
Mouton de Gruyter.
Boogaart, Ronny
1999 Aspect and temporal ordering. A contrastive analysis of Dutch and
English. PhD dissertation, Amsterdam. The Hague: Holland Academic
Graphics.
Boxtel, Sonja van
2005 Can the late bird catch the worm? Ultimate attainment in L2 syntax. PhD
dissertation, Nijmegen. Utrecht: LOT.
Bybee, Joan
1992 Entry on tense and aspect. In International
Encyclopedia of Linguistics, Bright, William (ed.), 144. Oxford: Oxford
University Press.
Bybee, Joan; Perkins, Revere and Pagliuca, William
1994 The Evolution of Grammar: Tense, Aspect and Modality in the Languages
of the World. Chicago: University of Chicago Press.
Carroll, Mary and Lambert, Monique
2003 Information structure in narratives and the role of grammaticised
knowledge: A study of adult French and German learners of English. In
Information Structure and the Dynamics of Language Acquisition,
Christine Dimroth and Marianne Starren (eds.), 267-287. Amsterdam:
John Benjamins.
Carroll, Mary and Stutterheim, Christiane von
2003 Typology and information organization: perspective taking and language-
specific effects in the construal of events. In Typology and Second
Language Acquisition, Anna Ramat (ed.), 365-402. Berlin: Mouton de
Gruyter.
Carroll, Mary; Lambert, Monique; Natale, Silvia; Starren, Marianne and Stutterheim,
Christiane von
in press Being specific: The role of aspect in event construal when grounding
events in context. A cross-linguistic comparison of advanced second
language learners (L1 French–L2 English, L1 German–L2 English, L1

23
Dutch–L2 French, L1 Italian–L2 German). In Dynamics of Learner
Varieties, Stefanie Haberzettl (ed.). Berlin: Mouton de Gruyter.
Comrie, Bernard
1976 Aspect. Cambridge: Cambridge University Press.
Dahl, Östen
1985 Tense and Aspect Systems. Cambridge Oxford: Basil Blackwell.
Flecken, Monique
2006 The expression of simultaneity in L1 Dutch. Unpublished MA thesis,
Radboud University, Nijmegen, The Netherlands.
Isačenko, Aleksandr
1982
4
Die russische Sprache der Gegenwart. München: Max Hueber.
Jespersen, Otto
1949 A Modern English Grammar: Part IV. Morphology. Heidelberg: Carl
Winter.
Johnson, Bonnie and Fey, Marc
2006 Interaction of lexical and grammatical aspect in toddlers’ language.
Journal of Child Language 33: 419-435.
Klabunde, Ralf and Stutterheim, Christiane von (eds.)
1999 Representations and Processes in Language Production. Wiesbaden:
Deutscher Universitäts Verlag (Reihe Studien zur
Kognitionswissenschaft).
Klein, Wolfgang
1994 Time in Language. London: Routledge.
Langacker, Roger W.
1987 Foundations of Cognitive Grammar, Vol. I: Theoretical Prerequisites.
Stanford, California: University Press.
Li, Ping and Shirai, Yasuhiro
2000 The Acquisition of Lexical and Grammatical aspect. Berlin: Mouton de
Gruyter.
Petr, Jan
1987 Mluvnice češtiny. Skladba. Praha: Academia.
Schmiedtová, Barbara
2004 At the Same Time... The Expression of Simultaneity in Learner Varieties.
Berlin: Mouton de Gruyter.

24
Schmiedtová, Barbara
2005 The notion of aspect in Czech and Russian. Paper presented at Dynamics
of Learner Varieties, Berder, France, March 2005.
Schmiedtová, Barbara and Sahonenko, Natalya
in press Die Rolle des grammatischen Aspekts in Ereignis-Enkodierung: Ein
Vergleich zwischen Tschechischen und Russischen Lernern des
Deutschen. In Fortgeschrittene Lernervarietäten: Korpuslinguistik und
Zweitspracherwerbsforschung, Patrick Grommes & Maik Walter (eds.).
Tübingen: Max-Niemeyer-Verlag.
Slabakova, Roumyana
2002 Recent research on the acquisition of aspect: An embarrassment of
riches? Second Language Research 18 (2): 172-188.
Slabakova, Roumyana and Montrul, Silvina
2002 On viewpoint aspect interpretation and its L2 acquisition: A UG
Perspective. In Tense-Aspect Morphology in L2 Acquisition, Yasuhiro
Shirai and Rafael Salaberry (eds.), 363-396. Amsterdam: John Benjamins.
Slabakova, Roumyana
2005 What is so difficult about telicity marking in L2 Russian? Bilingualism:
Language and Cognition 8 (1): 63-77.
Smith, Carlota
1997 The Parameter of Aspect. Dordrecht: Kluwer Academic Press.
Stoll, Sabine
1998 The role of Aktionsart in the acquisition of Russian aspect. First
Language 18: 351-378.
Stoll, Sabine
2005 Beginning and end in the acquisition of the perfective aspect in Russian.
Journal of Child Language 32: 805-825.
Stutterheim, Christiane von and Nüse, Ralf
2003 Processes of conceptualization in language production. Linguistics
(Special Issue: Perspectives in language production) 851-881.
Stutterheim, Christiane von and Lambert, Monique
2005 Crosslinguistic analysis of temporal perspectives in text production. In
The Structure of Learner Varieties, Henriette Hendricks (ed.), 203-230.
Berlin: Mouton de Gruyter.
Stutterheim, Christiane von and Carroll, Mary
2006 The impact of grammaticalised temporal categories on ultimate
attainment in advanced L2-acquisition. In Educating for Advanced

25
Foreign Language Capacities: Constructs, Curriculum, Instruction,
Assessment, Heidi Byrnes (ed.), 40-53. Georgetown: University Press.
Vendler, Zeno
1967 Verbs and Times. In Vendler Zeno (ed.) Linguistics in Philosophy. Ithaca,
NY: Cornell University Press; 97-121.
Verkuyl, Henk
1993 A Theory of Aspectuality. Cambridge, U.K.: Cambridge University Press.
Wagner, Laura
2006 Aspectual bootstrapping in language acquisition: telicity and transitivity.
Language Learning and Development 2(1): 51 - 76.
1
The term preference refers in our framework to highly automated processes that speakers
activate during speech production. The structures involved in these processes consist of
concepts that are expressed through grammaticalized (linguistic) means.
2
To avoid confusion, in our terminology we label lexical aspect Aktionsart.
3
Bybee et al. (1994) label this phenomenon grammaticalization of zero (of the unmarked
form).
4 The present perfective in Slavic languages is considered to have a future interpretation
(for Czech: Petr 1987; for Russian: Isačenko 1982). Recent research, however, has shown
that this is not necessarily the case for Czech, where perfectively marked verbs in the
present tense can have a present tense (here-and-now) interpretation (Schmiedtová 2004,
2005).
5 A possible way of testing whether the aspectual value of the unmarked simplex forms is
part of the grammar (grammatical aspect) or the lexicon (Aktionsart) is to conduct a
priming experiment. This research question will be addressed in our lab at the University of
Heidelberg in the near future.
6
There are about 20 different prefixes available in Czech that are used to make a verb
perfective. Each of them is associated with a cluster of meanings, most of them exhibit
polysemy and homonymy, and the realization of a given meaning of a prefix is highly
dependent on the context in which the prefix occurs. The same holds for Russian.
7
Preliminary results clearly indicate that grammatical features guide speakers’ attention
patterns: To be more precise, the focus on the right boundary as predicted by our linguistic
analyses of Czech and Russian is visible in speakers' eye movements (significant difference
in amount of fixations in the critical region) and speech onset times (a significantly later
speech onset times for speakers who are right boundary-minded). The patterns that were
found in the production data thus have a psycholinguistic reality.
8
We are currently testing this hypothesis with Dutch native speakers by means of an
acceptability judgement task.
9
The duration was brought about through the description of the situation. The verb itself
does not reveal the duration of the situation. For example, in the case of veranderen, the
situation was described as ‘changing the interior of one’s apartment’, elongated with
several adverbials expressing that you have been working on this for a very long time so far
and you will not finish this in the near future.

26
10
This approach to the aan het-construction is being pursued by Marianne B. Starren’s
research group in the Business Communication Department at the Radboud University in
Nijmegen, The Netherlands.