Content uploaded by Kristin Kersten
Author content
All content in this area was uploaded by Kristin Kersten on Apr 02, 2023
Content may be subject to copyright.
From: Burmeister, Petra; Piske, Thorsten; Rohde, Andreas (Eds.): An Integrated View of Language
Development. Papers in Honor of Henning Wode, Wissenschaftlicher Verlag Trier, 2002.
ISBN 3-88476-488-8. www.wvttrier.de
The acquisition of English verbs in an elementary school
immersion program in Germany
Kristin Kersten, Christine Imhoff & Bianca Sauer
0. Introduction
For some 10 years, Henning Wode has encouraged and supervised several immersion
(IM) programs in Schleswig-Holstein, northern Germany. It is mainly due to his
initiative that these programs are now firmly established in several northern German
institutions.
This paper is concerned with the latest of these institutions to have adopted bilingual
education, the Claus-Rixen elementary school in Kiel-Altenholz. It aims at presenting
the development of the acquisition of verbs and verbal inflections of the first cohort
after one and after two years of schooling. So far, a preliminary analysis of 12 subjects
has been carried out. The speech samples have been elicited with the help of a picture
story, the well-known frog story (Mayer 1969). The present study carried out by
Henning Wode and his research team – although still in its initial stages – is of
particular relevance with regard to the efficiency of early partial IM programs as, to
our knowledge, no morphological study of a German-English immersion setup of this
age group has yet been conducted.
1. The Kiel Immersion Project
1.1 General setup and objectives
The Kiel IM Project consists of three sections: Firstly, the partial IM program at
academic secondary schools implemented in 1991 (Burmeister & Daniel, this volume),
secondly, early English IM, which was introduced in a monolingual German preschool
in Kiel in 1996 (Rohde & Tiefenthal, this volume), following the example of a
successful French-German IM pilot project in Rostock, Germany, in 1995. Thirdly, in
September 1999, the project was expanded to the Claus-Rixen elementary school in
Kiel-Altenholz.
The overall objective of the Kiel IM Project is to start introducing one (or several)
second language(s) via IM at a point as early as possible in childhood, and to continue
teaching it throughout elementary and high school. Wode (2000) criticizes the
474 Kristin Kersten, Christine Imhoff & Bianca Sauer
introduction of an L2 in secondary schools as it allows for only one L2 to be learned
on a functionally appropriate level. He recommends the introduction of an L2 as early
as preschool, since research has shown that young children profit from multilingualism
(Wode 1981, 1988/1993, 1995)
1
and are usually more successful in acquiring an L2
than older learners (Johnson & Newport 1989, Flege et al. 1995, Flege et al. 1999).
In his most recent work, Wode advocates a setup which he calls the 3+-formula (Wode
2000, Wode et al. 2001) and which comes close to what we find achieved in the
Luxemburg school system. The 3+-formula refers to a setup of trilingual education in
which children are able to learn (at least, i.e., "+") 3 languages during their time of
schooling, namely their L1 and two additional languages. One of these is supposed to
be introduced in preschool and continued through elementary school, whereas the
second is supposed to start in secondary school.
1.2 The elementary school IM program
The IM program was introduced in the Claus-Rixen elementary school in September
1999. The program represents a partial IM setup, which means that all subject matter,
apart from German, is taught in the L2 (English).
2
This adds up to about 60% English
and 40% German, so that both languages are promoted. This was especially important
to the parents of the first cohort since they feared that their children's L1 might suffer
from an all-English exposure.
In the preschools, the "one person – one language principle" (Döpke 1992) is applied,
which means that each group is supervised by a native speaker of the L1 (German) and
a native speaker of the L2 (English). By contrast, the teachers in the elementary
program are native speakers of German. All of them are trained as elementary school
teachers and have a degree in English.
During the first year, children are allowed to speak German, but English is especially
encouraged. During the second year, the objective is to use English only. Reading in
English is introduced in second grade, though English words are already playfully
introduced in grade one. Once writing in the L1 is established, L2 writing is introduced
in grade three. There are no grades and tests on the L2, nevertheless tests in the subject
matter, apart from German, are given in English. Focus on grammatical form within
subject matter teaching occurs, but is less frequent than in traditional language
teaching. However, corrections in the form of direct feedback are given from time to
time as the teachers see fit. In the second year, one lesson per week is attributed to the
introduction of grammatical structures. During these, the teacher focuses on topics of
1 The only restriction Wode (1995) makes in this context is the case of some minority children
whose L1 has a low socio-economic status and may not be mastered in a native-like fashion.
2 In the first year, religious education was also taught in German, but this was changed in the second
year, when a second L2 teacher was hired.
The acquisition of English verbs in an elementary school program 475
grammar which tend to be difficult to acquire, such as this/that,these/those,is/are,
has/have distinctions, or the 3
rd
ps sg -s, as well as regular and irregular past tense
forms.
So far, data from two cohorts have been elicited. This paper will only focus on the data
elicited from the first cohort consisting of 18 children. Half of the children have prior
knowledge of English from the preschool IM program; they are therefore referred to as
'experienced bilinguals' (exp). The other children mainly come from monolingual
German families with monolingual preschool experience; they are thus termed
'inexperienced bilinguals' (inexp). The data of 12 of these pupils were analyzed after
their first year (Lauer & Hansen 2001). The subjects were chosen with respect to a)
their language background and b) our judgment of the different ranges of ability: Six
children performed above, and six children below average.
Comparable to what has been observed in the IM preschool, the children very quickly
developed a high level of comprehensive abilities (see Berger 1999; Lauer 1999a,b;
Tiefenthal 1999; Tonn 1999; Wode et al. 1999; Maibaum 2000; Rohde 2001; Kersten,
in prep.; Rohde & Tiefenthal, this volume, for analyses examining preschool children;
see also Wode 1981, 1996, 1998a,b, 2001; Weber & Tardif 1991; Petit & Rosenblatt
1994; Petit 1996 and Petit, this volume), but in contrast to the preschool, the children
learned at a much higher rate in elementary school. Production appeared much earlier
and very soon reached a much higher level of complexity than in the preschool.
2. Research objectives and test design
The overall objective of the evaluation is to document the children's linguistic
development in the course of the first four years of the pilot school experiment in the
structural fields of lexical, morphosyntactic and phonological acquisition. We intended
to check upon the subjects' spontaneous speech behavior. Since this is not easy to test,
especially for reasons of comparability, extemporaneous speech samples were elicited
by means of the frog story.
The study is designed as a longitudinal one involving testing at the end of each year in
elementary school and using the same test material each year. In first grade, the
procedure is introduced with a pilot test ("A boy, a dog, and a frog", Mayer 1967).
3
The test ("Frog, where are you?", Mayer 1969)
4
is conducted with each cohort close to
the end of every school year. Monolingual peer classes serve as comparison groups
("A boy, a dog, a frog and a friend", Mayer & Mayer [no year given]; this test is also
conducted in German with the IM classes). Data from an English native comparison
group will be collected in addition to the data already elicited from American first
graders, who serve as comparison group at this preliminary stage of the analysis. The
3 See Hansen (2000) for an analysis of the pilot test of the 1
st
cohort.
4 See Berman & Slobin (1994) for an overview of research carried out with the help of the frog story.
476 Kristin Kersten, Christine Imhoff & Bianca Sauer
data from the American first grade come from an elementary school in St. Paul,
Minnesota.
5
The data presented here derive from the main test conducted after 10 months,
6
and
after one year and ten months of exposure to the second language. The subjects are
tested individually. The data elicitation is adapted and slightly changed from the
example of Housen et al (1999).
7
It consists of a picture story retell task. The complete
elicitation procedure includes two steps: 1) The subject is interviewed in English by a
native speaker of German. He or she is asked to tell the picture story in English, but is
allowed to ask for assistance in German if in need of certain words or expressions. In
this retelling, the pictures are visible to both the child and the interviewer. The main
objective of this first part of the elicitation procedure is to acquaint the child with the
story and any unknown vocabulary. 2) The subject is interviewed exclusively in
English by a second interviewer who does not speak German (or pretends not to do
so). All conversation and story-telling thus have to be done in English, without any
assistance in the L1. During this retelling, only the child can see the pictures. All
interviews are recorded with video and audio equipment.
As will become obvious, there is a clash in the data between the different verb forms
used by the children and their functional appropriateness. Thus, in the first section of
this article, we will concentrate on the different verb forms used after 0;10 and 1;10
years of exposure, i.e., the morphological inflections used by the children will be
identified and compared. Morphological inflections encode diverse grammatical
functions, such as tense, aspect, number etc. In the second section, we will focus on
the influence of lexical aspect as a selected example for function (see Rohde, this
volume; Weist, this volume), since the progressive form is the first to appear in the
data and the form predominantly used by the children in the first year. The guiding
research questions are:
1. How does the children's use of verb forms develop from grade 1 to grade 2?
a) What kind of verbal inflections occur in the data?
b) What kinds of different error categories can be identified?
5 Twelve subjects were chosen randomly out of 30 samples to match the number of subjects for
whom our analysis was carried out. The subjects were between six and seven years old. The
elicitation procedure was equivalent to the second (all English) interview carried out with the
German children (see below).
6 For a detailed analysis of this data see Lauer & Hansen (2001) and Beier (2001).
7 The study of Housen et al. (1999) is the result of a research cooperation between researchers from
different European countries on the occasion of the Euroconferences I-III: The Teaching of
Foreign Languages in European Elementary schools (1997, 1998, 1999). The final report is an
unpublished manuscript, since the conference proceedings have not been completed. It contains a
detailed analysis of elementary school L2 learners of English with different L1s and from different
sociolinguistic backgrounds.
The acquisition of English verbs in an elementary school program 477
c) How much variation can be found and what does it tell us about different levels
of achievement reached by the children?
2. What is the function of the children's first verbal inflections?
3. Results
3.1 The data at the beginning of schooling
At the beginning of grade 1, L2 production among the children varies according to
their preschool experience. It was observed that in the first weeks of schooling, the
inexp children hardly ever produced English output, whereas the exp bilinguals
contributed to classroom conversation in English more readily. In a simplified picture
story test, conducted after approximately 60 days of schooling in the second cohort,
8
the data elicited from inexp bilinguals consists mainly of one-word utterances. The
inexp children hardly use any verbs, whereas with the exp, coherent sentences are
found. Two excerpts of the narrations from this test were chosen to show these
differences displayed after about two months of elementary schooling, and to illustrate
the level at which production occurs in the early stages of the IM program; they may
serve as a means of comparison for the later tests in the first cohort:
Transcript 1: Excerpt of narration after approx. 60 days of exposure (grade 1; 2
nd
cohort).
Subject 16 (inexp)
IE Can you tell me what you see on the pictures?
16 It's dog, it's dog sleeping, it's sunny, it's dog, it's sunny, it's schoolbag, it's dog, it's schoolbag
IE What are they doing?
16 It's dog, it's classroom, it's schoolbag
IE - English-speaking interviewer
Student 16 is able to identify certain objects that he can see on the pictures, but he
cannot produce coherent sentences. The expression it's is known to the children from
daily question routines such as "What's the weather like? – It's sunny" and seems to be
used by child 16 as a general strategy for identifying objects.
Transcript 2 reveals that student 9 is much further advanced. He is able to produce
simple coherent sentences and makes use of both uninflected verb forms and the -ing
form. Nevertheless, his narration is still very limited.
8 The data presented in these Transcripts 1 and 2 are taken from the second cohort, because this
particular test was not administered with the first cohort.
478 Kristin Kersten, Christine Imhoff & Bianca Sauer
Transcript 2: Excerpt of narration after approx. 60 days of exposure (grade 1, 2
nd
cohort).
Subject 9 (exp)
9 Here's a man. And a dog. Here # here go the man to the school and go #
IE Forest?
9 and to the/ and go in the forest. Here's the school. And the dog come to the school. The dog go
in a/ go back. The dog to to/ to the school. Here is it. The dog looking in the schoolbag. The dog
is under the chair.
IE What's he doing?
9Sleeping.
IE - English-speaking interviewer; # - pause; / - hesitation or self-correction
3.2 The data after one year of schooling
3.2.1 Illustrative examples
The test conducted after 10 months of schooling reveals a large amount of inter-
individual variation in the production. However, similar strategies can be found among
children with a comparable L2 background. The following transcript excerpts show
some typical features displayed by children with no L2 experience before entering
elementary school.
Transcript 3: Excerpt of narration after 10 months of exposure (grade 1; 1
st
cohort).
Subject 7 (inexp)
7 The boy and the dog/ I can see a frog. The frog running away, next morning it/
IE Can you speak a bit louder, so I can understand/
7 At next morning is the Frosch away. The boy and the dog suchen the frog.
IE What? They what...?
7 ...suchen den frog.
IE Okay.
7 The boy rufen the frog. The boy, eh the dog falling down. The boy helding the dog im Arm.
The boy rufing the frog, eh and the dog rufing "Farewell" and the boy rufing the frog. And the
boy, eh, and the dog running away. […]
IE - English-speaking interviewer; / - hesitation or self-correction; italics – L1 element
The student displays some morphologically target-like verb forms such as falling,
running, see. She frequently makes use of her L1 (suchen [search] rufen [call]) and
sometimes applies the English inflection -ing to German verbs (rufing). The -ing form
is the predominant form, however, it does not occur in combination with the respective
auxiliary.
Transcript 4 shows the narration from another inexp child:
The acquisition of English verbs in an elementary school program 479
Transcript 4: Excerpt of narration after 10 months of exposure (grade 1; 1
st
cohort).
Subject 8 (inexp)
8 There is a dog and a boy and the d/dog looking in a glass, and in the glass sitting a frog and the
moon shining. And then, the boy are sleeping and the dog sleeping. And then, the boy looking
in the glass and the frog is/ is not there. Then, the boy looking in the t-shirt and the dog looking
in the glass and the boy looking out of window. And the dog staying/ staying next to the boy.
And then, the dog falling down out the window and the boy has the dog in his hand. And then,
the boy are staying out the house and # and looking in the water, And then, the boy looking in
a hole and the bo/ and the boy say "Au!". Then, the boy looking in a hole and the, ehm, bees fly
away. And then, the dog are running away. Then, the boy coming to a stone and then, the boy
staying on the stone. Then, the boy are sitting on a deer and the dog are running away. Then,
the dog and the/ and the boy are falling down. Then, the boy and the dog are sitting in the
water. And then, the boy are say "psst" and the dog are looking # to the boy. And then, the boy
see a frog and next to the frog sit a frog. And then, the frog children come # out the grass. And
then, the boy say "goodbye" to the frogs.
# - pause; / - hesitation or self-correction
Subject 8 also predominantly uses the -ing form. The verbs mostly occur without the
auxiliary. However, subject 8 displays a few instances of V-ing in combination with
the auxiliary are. This verb form is used in both singular and plural contexts,
indicating that subject 8 has not yet acquired the distinction between is/are in the
respective contexts. In contrast to subject 7, subject 8 does not use any German words.
The unmarked base form (V-ø) is also recorded for child 8, but she does not apply the
3
rd
ps sg -s where required: and the boy say "AU!". The third transcript shows the
narration of an above-average student who attended the bilingual preschool before
starting elementary school:
Transcript 5: Excerpt of narration after 10 months of exposure (grade 1; 1
st
cohort).
Subject 1 (exp)
1 The boy have a frog. And the boy is looking at the frog. And in the night the frog is wants to go
away. And the boy and the dog is/ are sleeping. And now on/ in the morning the boy are/ is
scared because the frog is away. # And the boy is can't seeing the frog anymore. The boy is
screaming: "Frog, frog, where you are?" And now the fro/ the dog is looking in a hm bottle and
then the dog is falling out of the window. And then the boy is jumping. And then the boy the
frog eh/ the dog in his a/ arms. And then the # boy is screaming: Frog, frog, frog, where you
are? #. And then the boy is looking for the f/ frog and then the/ the boy is scared and # and the
fro/ and the dog is scared, too. There come bees. And the frog is run away/ eh the dog is run
away. And now the boy is seeing a/ in a tree/ going to a tree and # now a/ the boy is scared,
there comes a owl. And the boy is falling down. And they the/ then the boy wants to climb on
a/ on a stone and they/ and the boy is screaming: "Help, help!" And then the boy is don't know
where he/ where he is. And now the boy is on a # he don't know what he's is/ what h/ what the/
what the/ what's animal it is/ it is and then the boy falling in a sea. And then the boy a/ hear:
"Quack, quack, quack, quack".
IE Aha. (laughs)
1 And then the boy said: "Pssh. Be quiet!" And then the frog can see two frogs and baby frogs.
And then the fro/ the boy is saying: "Bye, bye frogs." And one little frog can't go on the tree.
IE - English-speaking interviewer; # - pause; / - hesitation or self-correction
480 Kristin Kersten, Christine Imhoff & Bianca Sauer
In contrast to the other two, subject 1 uses the -ing form mostly in combination with an
auxiliary. Self-corrections like the boy and the dog is/ are sleeping show that subject 1
is aware of the distinction between is and are. He sometimes overgeneralizes the use
of the auxiliary as is apparent in utterances like the boy is don't know where he is. The
target-like (tl) use of 3
rd
ps sg -s appears to be rendered in there comes a owl, however,
subject 1 also displays the non-target-like (ntl) use of the unmarked base form without
the -s: and the the boy hear: "Quack …". The use of infinitive constructions like the
boy wants to climb, and instances of negation like one little frog can't go on the tree
show that subject 1 has already acquired a wider range of verb forms than the other
two subjects presented above.
The transcripts illustrate two things: Firstly, the progress from 1½ months to 10
months of exposure is considerable, and secondly, the variation between the single
subjects as well as the intra-individual variation with respect to verb morphology is
high. Child 1 uses 14 different verb types, 29 different tokens and several different
verbal structures, whereas child 7 uses only 4 different types with a total of 8 tokens.
Nevertheless, each child without exception manages to retell the full story in his or her
own way – sometimes more, sometimes less rudimentarily.
Altogether, the 12 subjects use 296 verb tokens and 29 different verb types (Lauer &
Hansen 2001) with the present participle (V-ing) being the predominantly used
inflection (158 tokens). The base form (V-ø) is the second most frequent form (56 ntl
and 14 tl uses). Only 24 tokens are found for the preterite (V-ed) and the 3
rd
ps sg (V-
s) respectively; infinitives (V-ing) occur only rarely (16 tokens). It was impossible to
determine, after the first year, whether the high number of progressive forms was due
to a developmental level reached by the subjects in their interlanguage, or whether the
aspect of the progressive was used in a tl manner. If the children understood the task as
a picture description, the present progressive might have been the appropriate form to
describe an ongoing action on the picture. This elicitation problem will be discussed
further in section 3.3.1.
3.2.2 Non-target-like forms after 0;10 years
The data show a wide range of variation in the use of ntl structures. In order to obtain a
detailed classification of the different error categories, the errors are divided into two
major groups, i.e., morphological and morphosyntactic errors, with the latter including
those types of errors which depend on a syntactic structure, e.g., the progressive V-ing
in combination with the auxiliary, the use of the past tense form where the past
participle is required (only second grade, see section 3.3.2), the base form V-ø
with/without 3
rd
ps sg -s, as well as morphosyntactic transfer from the L1.
Out of 296 tokens used by the group, there are 184 ntl tokens, which amounts to over
60%. Since V-ing is the predominantly used inflection, most of the errors occur with
this form. The most common error in the samples is the use of the progressive without
The acquisition of English verbs in an elementary school program 481
auxiliary (the frog running; 48% of all errors tokens), followed by the base form where
an -s is required (the boy say; 22%) and the use of the ntl auxiliary (the boy are
sleeping; 13%). This means that most errors are morphosyntactic in nature. Even the
most common morphological error, the insertion of tokens from the children's L1
(rufing), is almost negligible (6%).
The bias towards the progressive form in this narration could not be confirmed by a
comparison group with English as L1. Frog story data from the American first grade
showed that the comparison group's verbal inflections are more evenly distributed
across all forms. As expected, the use of ntl structures is much smaller than that of the
L2 learners. It is striking, however, that the source of the highest number of errors for
the L1 speakers seems to match that of the L2 speakers, i.e., the uninflected base form
V-ø in positions which require either an -s or the past inflection. It has to be added that
the use of a single tense of narration was found to be almost as instable with the L1
English comparison group as with the L2 learners. This shows that at least the above
types of errors have to be regarded as a natural developmental phenomenon and cannot
be accounted for by transfer from the L1.
3.2.3 Achievement levels
Not only do we find a high level of intra-individual variation as revealed in Transcripts
3-5, but child 1 (exp) is obviously much further advanced than child 8 (inexp), and
child 8 further than child 7 (inexp). In fact, a closer look at the intra-individual
variation between learners reveals different levels of achievement. Table 1 identifies
the different verbal features the children display at the end of grade 1.
Table 1: Verbal features after 0;10 years.
Feature
2
exp
7
inexp
9
inexp
14
exp
3
inexp
6
exp
8
inexp
4
inexp
5
inexp
1
exp
12
exp
16
exp
1. V-ing¬aux + + + + + + + +
2. V-ø (infl) + + + + + + + + + +
3. aux+V-ø + + + +
4. V-ing+*aux + + + +
5. V-ing+aux + + + +
6. V-s + + + +
V-ing¬aux – he looking; V-ø (infl) – he look; aux+V-ø – he is/are look; V-ing+*aux – he are looking;
V-ing+aux – he is looking; V-s – he looks
At the end of grade 1, the children's interlanguage can be divided into three levels
according to their use of verb forms. Children 2 (exp), 3 (inexp), 7 (inexp), 9 (inexp)
and 14 (exp) use the -ing form consistently without any auxiliary, and at the same
time, the use of the unmarked base form (V-ø) is present (features 1 and 2). The
children seem to have not yet acquired the use of the auxiliary. Children 4 (inexp), 6
(exp) and 8 (inexp) already use the -ing form in connection with an auxiliary, but the
auxiliary used is always are, never is, regardless of the respective subject (feature 4).
The use of the unmarked base form is also recorded for these children. Children 3
482 Kristin Kersten, Christine Imhoff & Bianca Sauer
(inexp) and 4 (inexp), among others, display overgeneralization in the sense that they
sometimes use the auxiliary in combination with the unmarked base form, e.g., the boy
are look (feature 3). None of these children produces the 3
rd
ps sg -s. The difference
between the two groups is that the former does not use the auxiliary at all, whereas the
latter uses the ntl auxiliary. Children 1 (exp), 5 (inexp), 12 (exp), and 16 (exp) seem to
have reached a third level. They all use V-ing in connection with the correct auxiliary
(feature 5), and they use V-s tl (feature 6) although the use of the unmarked base form
in places requiring the 3
rd
ps sg -s still occurs with these children as well, indicating
that they have not yet completed the acquisition process. The three levels of
achievement regarding the form of verbs can be summarized as follows:
9
Table 2: Achievement levels in the acquisition of verbal inflections after 0;10 years.
level features description examples children
1V-ing ¬ aux
V- ø (infl)
use of auxiliary in connection
with V-ing not recorded
and the boy looking for the
frog and the dog look at the
tree (2)
2exp, 3inexp,
7inexp, 9inexp,
14exp
2V-ing + *aux
V- ø (infl)
use of auxiliary are in
connection with V-ing
and then the boy are going out
and say (6)
4inexp, 6exp,
8inexp
3V-ing + aux
V- ø (infl)
V-s
use of correct aux. in
connection with V-ing,
use of V-s
The boy is sleeping.[...] The
boy looks, and the dog looks.
The boy see two frogs. (12)
1exp, 5inexp
12exp, 16exp
V-ing¬aux – he looking; V-ø (infl) – he look; aux+V-ø – he is/are look; V-ing+*aux – he are looking;
V-ing+aux – he is looking; V-s – he looks
It is noticeable that the division of the group into exp and inexp learners does not in all
cases relate to the level of achievement. However, level 3 is predominantly reached by
children with L2 preschool experience.
3.3 The data after two years of schooling
3.3.1 Illustrative examples
All children have improved to a large extent at the end of grade 2, as Transcripts 6-8
reveal. Subjects 7, 8 and 1 exemplify the progress within the group.
Transcript 6: Excerpt of narration after 1;10 months of exposure (grade 2; 1
st
cohort).
Subject 7 (inexp)
7 A boy sits on a chair and he had the dog and the shild/ and the child and the dog look at a frog
in a glass. And it was night and the boy was in his pyjama. And then he goes sleeping and the
dog, too, and the dog sleep, too, on the bed and then the frog go out the glass and then the next
morning the boy and the dog look at the glass and there are zero frogs. And then he look at/ at
the frog, the boy look at his shoes, the dog put his head in the glass and he couldn't put the
glass from his face away. Then he look at out and the boy cried: Frog, are you here, are you
9 Note that with respect to the progressive forms, this table only accounts for the acquisition of form.
Since we assume that the children are not yet aware of the progressive aspect, no statements can be
made about the acquisition of progressive function (see sections 3.4, 3.5).
The acquisition of English verbs in an elementary school program 483
there? And the dog look to # it # outside. And then the dog fells down and the glass # goes #
[…] and then the boy jump to the dog and the boy look at the dog and/ but the dog leckt # at the
boy's face. And then he go and go and cried: Frog, are you there? Then come to me, please!
And/ and the dog finds bees and XXX at the bees # and schnuppert at the bees. And then the
boy found a hole and he cried in the hole: Frog, are you there? Then come out! And the dog
finds a bee # Nest
IE Okay
7 […] But then the owl cames out and the boy falls down. […]
IE - English-speaking interviewer; # - pause; / - hesitation or self-correction; XXX – incomprehensible
italics – L1 element
In contrast to the narration from the previous year, subject 7 now predominantly uses
simple forms. Although the 3
rd
ps sg -s is recorded (a boy sits on a chair; and the dog
finds bees), it is left out in most required cases (and the boy jump; and the boy look
etc.). Subject 7 displays regular as well as irregular past tense forms. The transcript
also shows one instance of overgeneralization of 3
rd
ps sg -s: But then the owl cames
out.
Transcript 7: Excerpt of narration after 1;10 months of exposure (grade 2; 1
st
cohort).
Subject 8 (inexp)
8 There was a boy and he had a frog in a glass and he look at the frog, he laugh the frog and the
dog look at the frog. And then the boy goes into bed and sleep and the frog goes out, out of the
glass and go in his home, and the dog sleep by the boy and the f/ next morning the frog is away
and the boy look at the glass and the boy is a/ and the frog is away. The dog look at the f/ look
at the glass but the frog is not there. Then the boy look in a boot if there is the frog and the dog
take his head and put it in the glass but there is not the f/ frog in and then the boy make the
window open and then he shouted eh "where are you, where are you?" And the frog f/ and the
dog f/ full out the window. And then the boy look at the dog and the dog full on the earth and
then the glass is crashed and the dog is on/ in the arms from the boy. And then the boy goes out
of the house and go in the wood to see where is the frog and then the/ he/ h/ go with his dog in
the wood. And he look at the in the woods, in a hole in the earth, and then when he look at it,
the dog look at the bees and then the/ a small little animal came out of the hole where the boy is
looking and he bite in his nose. And the bees come out and then the boy is looking in a hole in
a tree and the boy and the dog is looking at the bees. […]
/ - hesitation or self-correction
Subject 8 also predominantly uses simple forms in her retelling of the picture story at
the end of grade 2, mainly the unmarked base form in places requiring 3
rd
ps sg -s.
Variation with respect to the use of -s in the same sentence sometimes involve the
same verb (and then the boy goes out of the house and go in the wood).
Transcript 8: Excerpt of narration after 1;10 months of exposure (grade 2; 1
st
cohort).
Subject 1 (exp)
1 Once upon a time there was a little boy, a l/ dog and a frog. The little boy has found the frog
ehm beside a lake and he wants to/wants to ehm have the frog forever. And he put the frog in
a glass. And one night when the little, little boy and he eh the dog were sleeping, the frog
jumped out of the glass and jumped out of the window. Next morning when the little, little boy
woke up, he wants to say his dog hello, eh his frog hello but he do/ but he doesn't found his
frog. Ehm the frog d/ wasn't no longer in the glass. He don't know where it was. The little boy
484 Kristin Kersten, Christine Imhoff & Bianca Sauer
looks/looks in his shoe, and the little dog looked in the glass and then the little boy runs to the
window and shouted: Frog, where are you? And the dog ha/ has a glass on his head and falls
from the windowsill. Then the little boy runs, runs out of the w/ out of the house and take to/
take his dog. Eh then they want to found the frog. The little boy shouted again: Frog, where
are you? Then ehm the little boy found a hole in the earth, but/ and he looked in the, in the
hole, but he doesn't found the frog but another animal he doesn't know/ he doesn't know the
name of the animal. Then the little boy climbs on a tree to/ and he wants to look in a hole in the
tree. But then ehm out of the hole in the tree comes an owl. The boy was scared of the owl/ owl
and climbed on a tree and behind the tree there was a deer. And he had/ he stands on the deer's
head. Then the deer maked a stop and the boy felled down in a lake. Then the little boy hears a
noise. He makes a sign to his friend/ his dog: Be quiet, dog! Then they have saw the frog ehm
their frog. And another frog and nine little frogs. And the little/ little boy says bye-bye to the
frogs and one little frog has he in his hand. That was the end.
/ - hesitation or self-correction
Student 1 shows only one instance of using V-ø in a place requiring -s (and the little
boy runs out of the house and take his dog); in the remainder of the narration he uses
V-s tl. As is the case with subjects 7 and 8, subject 1 uses both present and past tense
forms in his narration without any recognizable functional distinctions. He displays
some typical examples of overgeneralization of the regular past tense form (Then the
deer maked a stop and the boy felled down in a lake).
It is not surprising that, after 1;10 years, all of the speech samples are longer. All of
them show a larger variety not only of verb types and morphological forms, but of
vocabulary in general. Several things are striking: Firstly, the all-dominating ing-forms
seem to have vanished from the transcripts – a fact which is representative of most of
the other subjects. Secondly, a large amount of errors can still be identified although
their quality has changed; and thirdly, the use of tense is more diverse than in the first
year, and thus instable.
The amount of verb types and tokens used by the whole group to tell the same story
nearly doubled (grade 1: 29 types, 296 tokens, grade 2: 44 types, 540 tokens). A
detailed look at the verb types shows that the increase is mainly due to the appearance
of synonyms (search/look for, walk/go etc.)indicating that the children's lexicon has
indeed grown considerably. As to the distribution of verbal inflections, V-ing occurs
only rarely (12 tokens out of 540), the simple forms V-ed (172 tokens), V-ø (200
tokens) and V-s (98 tokens) are dominant in the samples. The perfective V-en and the
infinitives are still negligible. The drastic change in the use of verbal inflections, i.e.,
the decrease of the use of V-ing, might hint at the fact that the reason for the use of V-
ing in grade 1 was probably not an elicitation problem (cf. section 3.2.1) but rather
reflected a developmental achievement level of the learners.
It is important to note that there is a wide range of intra- and inter-individual variation,
with the amount of types per person ranging from 5 to 17 at the end of grade 1, and
from 11 to 24 at the end of grade 2. It is striking that the exp children on average use
both more types and more tokens than the inexp, which suggests that the length of
The acquisition of English verbs in an elementary school program 485
contact with the English language is an important factor in the development of the
lexicon.
3.3.2 Non-target-like forms after 1;10 years
Although the number of tokens used by the children after grade 2 has nearly doubled
(296 → 540 tokens), the number of errors has increased only to a small extent (184 →
195 tokens). In both tests, the number of morphosyntactic errors is distinctly higher
than the number of morphological errors. The children made more morphological
errors after the end of grade 2 than they did in the previous year. This is due to the fact
that the use of preterite and perfective forms (V-ed and V-en) played only a minor role
in grade 1, and considerably increased in grade 2. The most common morphological
error at the end of grade 2 is the use of the 3
rd
ps sg -s in combination with an irregular
word form (he cames; 5%), followed by the use of the regular past tense where an
irregular form is required (he runned; 3%). In both cases, overgeneralization seems to
be responsible for the production of the ntl form.
A detailed look at the morphosyntactic errors shows that the distribution of error types
has changed more drastically. After grade 1, the most common error was the use of V-
ing without the auxiliary (he running), followed by the use of the uninflected base
form (he say) and the use of the wrong auxiliary (he are looking). At the end of grade
2, however, these structures are negligible – the progressive, for instance, was used
only twice without the auxiliary – whereas the use of V-ø without inflection is now the
most dominant ntl structure (143 tokens). This shift in the production of errors
corresponds to the focus on simple and past forms in the grammar lessons, thus
making room for a narration pattern in the simple form to which the group changed
overwhelmingly. The distribution of present and past forms of the group reflects the
behavior of most of the single subjects: Both tenses are used alongside each other;
only very few single subjects manage to stick to one tense only. This corresponds to
the L1 data of the bilingual group, of its monolingual German control groups (i.e., the
parallel classes of the bilingual class), and of the English L1 comparison group (cf.
section 3.2.2), which reveal that even in L1 German and L1 English the use of one
single narration tense is not stable.
In order to obtain a better insight into the inter-individual variation, we have taken a
closer look at the data of the second year with respect to the amount of time that the
children have been exposed to the English language.
a) Target-like verbal inflections after 1;10 years:
The exp learners produce more tl tokens (205) than the inexp (144). It is striking
that the exp students use more simple forms (V-s, V-ed and V-inf) than the inexp.
Out of the exp learners, only one child makes use of the V-ing form, whereas the
use of V-ing is recorded for all inexp.
486 Kristin Kersten, Christine Imhoff & Bianca Sauer
b) Non-target-like verbal inflections after 1;10 years:
The number of ntl forms is higher for the inexp learners than for the exp. It is
striking that the inexp students use a lot more ntl V-ø forms than the exp, indicating
that most of the inexp children have not yet acquired the 3
rd
ps sg -s. The exp
children display slightly more ntl simple forms.
It seems that with increasing time of contact to the English language, the children
abandon the use of V-ing and start using simple forms more frequently. The high
number of ntl V-ø tokens among the inexp students indicates that the correct use of
inflections other than -ing is influenced by the contact time. Among the tested
children, the children with more contact time produced less ntl V-ø tokens and more tl
V-ed and V-s forms than the others.
3.3.3 Achievement levels
Whereas it has been fairly easy to distinguish different levels of achievement from the
data after 0;10 years, levels are not as clearly to identify after 1;10 years. By the end of
grade 2, the progressive form is almost absent in the children's narrative, therefore no
statement can be made about the development of the progressive forms for the children
who have reached levels 1 and 2 by the end of grade 1.
Regarding the acquisition of V-s, all children seem to have reached at least level 3 by
the end of grade 2, i.e., they all produce V-s structures. All children, however, still
produce ntl V-ø forms, i.e., the process of acquisition regarding the 3
rd
ps sg -s is not
complete. As mentioned above, the number of ntl V-ø structures is greater for the
inexp than for the exp subjects. All children except 1 (exp), 5 (inexp) and 16 (exp)
overgeneralize the use of V-s in the sense that they attach -s to forms where it is not
required, e.g., they comes or they cames. Children 1 (exp), 5 (inexp) and 16 (exp) seem
to have recognized the places where -s is not required, and they only produce a small
number of ntl V-ø structures. These children are the most advanced regarding their
acquisition of V-s.
All children make use of V-ed, but none of the children uses V-ed in a completely tl
manner. Overgeneralizations occur frequently alongside tl forms. It is interesting that
the number of overgeneralizations is higher for the exp than for the inexp subjects.
Although the inexp produce more ntl tokens than the exp, the errors almost exclusively
concern the use of the uninflected base form in positions where an inflection is
required. The exp children display a larger variation of error types, indicating that they
do not follow a single hypothesis about the formation of verb forms but experiment
with different forms. Overgeneralizations generally occur after the first tl forms are
produced and seem to mark an important step on the way to the tl use of the respective
inflections in complementary distribution.
The acquisition of English verbs in an elementary school program 487
3.4 Aspect-marking
So far, we have referred to the acquisition of verb forms without paying much
attention to the question of whether the respective verbal inflections fulfill the
grammatical function assigned to them in L1 speech. It has been claimed that in early
stages of language acquisition the acquisition of morphological forms precedes its
functionally correct use (Perdue 1993). In the following section, we will focus on the
second research question, i.e., the acquisition of function. The acquisition of
progressive aspect will serve as an example.
3.4.1 The aspect hypothesis
Research about the acquisition of verb inflections suggests that, in the early stages of
L2 acquisition, learners tend to encode the lexical aspect inherent to the verb, with
tense distinctions or grammatical aspect being neglected. This phenomenon is referred
to as the aspect hypothesis (AH, Andersen & Shirai 1994; Robison 1995; Rohde 1996,
1997, this volume; Weist, this volume; see Rohde, this volume, for a detailed
introduction to the topic). As far as lexical aspect/aktionsart is concerned, Vendler
(1967) differentiates four aspectual verb categories: states (want; have),activities
(swim; sleep),accomplishments (climb a rock; go home) and achievements (fall down;
find something). These different types of lexical aspect are distinguished with regard to
the duration, telecity and dynamicity inherently expressed by the listed
verbs/predicates. As research on the AH suggests, the verb categories predominantly
occur with the following verb inflections: The V-ing form, which grammatically marks
progressive, is mainly affiliated to verbs with an inherent lexical meaning of duration,
i.e., to verbs categorized as activities, whereas verbs expressing an end result
(achievements and accomplishments) mainly occur with past and/or perfective
inflections. State verbs, on the other hand, mostly occur with V-s. It is assumed that
the skewed distribution of verb inflections is, to some extent at least, also given in the
input and grants the learnability of verb inflections in both L1 and L2 acquisition
(Rohde 1997).
This analysis follows the outline of Rohde (1996).
10
Since the different categories
concern verb types rather than tokens, for this part of the analysis only types were
taken into account.
11
The following figures may serve to illustrate the distribution of
verb categories and their respective inflections in our data.
10 V-irreg forms have not been analyzed separately but have been included in V-ed.
11 It is important to note that verbs can change their category depending on the context. The verb to
eat is, on its own, an activity, while in the context of to eat a sandwich, it represents an
accomplishment. In this analysis, the single verbs are classified according to the context in which
they were used.
488 Kristin Kersten, Christine Imhoff & Bianca Sauer
3.4.2 V -ing
As already mentioned above, the most striking form to change from first to second
grade is the V-ing form, whose high number of occurrences is displaced, in the second
year, by more complex V-ed and V-s forms. The -ing form is, according to the AH,
expected to (predominantly) appear with activities because of their inherent lexical
aspect of duration. Figure 1 represents the use of the -ing suffix by the group with
respect to the different verb categories after grade one and after grade two.
Figure 1: The -ing suffix and its occurrence with the four different verb categories.
1
30
10
40
010
7
0
10
20
30
40
50
Grade 1 Grade 2
states
achievements
accomplishments
activities
There is a drastic decrease in the number of -ing suffixes in the second year. This goes
along with the results presented in 3.3.1. As expected, state verbs almost constantly
occur without -ing, and the activities are, in both years, the category which carries the
most -ing suffixes (40 types in 2000, 7 in 2001). These findings are in line with the
AH. The fact, however, that most of the achievements (30 types) in the first year and a
high number of accomplishments (10 types) are also used with the -ing form, which
amounts to almost exactly 50%, does not go along with the expectations (achievements
and accomplishments are predicted to predominantly occur with past inflections).
12
In
the second year, the combination of achievements + -ing suffix has almost completely
vanished (there is only one instance); achievement verbs then occur predominantly
with V-ø (55), V-ed (46) and V-s (38). Thus, lexical aspect-marking changes
drastically within the two categories of activities and achievements, with the somewhat
surprising fact that in the second year of acquisition, the predictions of the AH are met
to a higher degree than in the first.
Whereas the number of verb types used with the -ing form in the first year is almost
identical for both exp and inexp children (40/41 types), there is only one instance of a
V-ing form used by an exp child in the second year, compared to seven instances by
inexp children.
12 We will suggest a tentative explanation for this phenomenon in chapter 3.5.
The acquisition of English verbs in an elementary school program 489
3.4.3 V -ed
The AH predicts that the -ed suffix (Figure 2) prototypically occurs with achievements,
expressing a punctual action and telicity.
Figure 2: The -ed suffix and its occurrence with the four different verb categories.
3
9
01
23
46
11 13
0
10
20
30
40
50
Grade 1 Grade 2
states
achievements
accomplishments
activities
In the first year, the use of -ed forms is very rare (only 13 types). It is striking, though,
that the few instances of V-ed appear almost exclusively with achievements and very
few states. In the second year, this inflection has widely displaced the -ing form and
risen up to 93 types and, again, is predominantly used with achievements. However,
almost 50% of all V-ed instances are more or less evenly spread over the other verb
classes (23 states, 13 activities, and 11 accomplishments). It is important to note that
this distribution is not only characteristic of the group but also of a high percentage of
intra-individual variation (section 3.4.5).
3.4.4 V -s
The "link between states and the V-s", as stated in Rohde (1996, p. 1125), could not be
fully confirmed. Our findings reveal a somewhat different picture.
Figure 3: The -s suffix and its occurrence with the four different verb categories.
21
303
19
38
11 10
0
10
20
30
40
50
Grade 1 Grade 2
states
achievements
accomplishments
activities
Whereas the data in the first year seem to confirm the hypothesis (with 21 out of 27
types, the -s inflection is overwhelmingly used with states), the data produced in the
second year resembles the distribution of V-ed forms (Figure 2): V-s forms occur with
490 Kristin Kersten, Christine Imhoff & Bianca Sauer
all four verb categories, and the group with the highest number of V-s verb types is not
state verbs (19) but achievements (38). Even though the total number of types is not
exactly the same, the relations between the different verb categories are strikingly
similar.
3.4.5 Intra-individual variation: Illustrative examples
Tables 3 and 4 list a few selected examples from the database to show the variation
which occurs within almost every sample. These examples were selected to illustrate
the fact that in most of the individual narrations the use of inflections does not follow a
distinct pattern (note that not all verb types of a single sample are listed).
Table 3: Intra-individual variation after 0;10 months of input (selected examples).
Subject 3 (inexp):
activities: sitting; is sitting; eats
achievements: is falling; is fall; coming; say
Subject 8 (inexp)
activities: are looking; looking; are sitting; sitting; sit
accomplishments: are running away; fly away
achievements: are say; say; coming; come; are falling; falling
Subject 16 (exp)
states: is; has; hears
activities: are flying; is looking; looks out; looked
accomplishments: is climbing; is running away; dress up
achievements: is falling; falling; is going; shouting; shouted; is coming; come out; said
As the examples in Table 3 show, the variation between -ing, -ed, and -s is typical of
each verb category (except the states, where -ing occurs only rarely), and is even very
common within one and the same verb type used by a child (16 exp:shouted, shouting
etc.). The distribution of inflections used by the selected subjects is representative of
most of the other children. With reference to the AH, however, no clear pattern is
recognizable,in other words, there is a high rate of intra-individual variation within
verb categories and within single types. One prototypical instance is shown by subject
9 (inexp), who is regarded as one of the least advanced inexp subjects. He inflects the
German "hat" (has) with the typical state-inflection -s (hats). This would denote a
prototypical use according to the AH.
Table 4: Intra-individual variation after 1;10 months of input (selected examples).
Subject 1 (exp)
states: be; are; was; has; had; wants to have; wants to look; wants to say; want to
found; doesn't found; doesn't know; don't know; have saw; hears
activities: were sleeping (background activity); stands; looks in; looked in
accomplishments: runs to; climbs on; climbed on
achievements: has found; found; maked; makes; falls; felled; shouted; says; put; jumped;
woke up; take; comes out
The acquisition of English verbs in an elementary school program 491
Subject 4 (inexp)
states: is; are; have; hear; see
activities: are looking; looked; look; followed; follow; lies; riding
accomplishments: shaked on; goes at home
achievements: fells on ; felled out; fall into; goes in pieces; wakes up; comes out; gives a
sign; give; open; shouted; said; cried
Table 4 illustrates a pattern widely used by the subjects in grade 2: The progressives
have almost completely disappeared; the forms V-ø, V-ed and V-s all appear in
seemingly free variation (4 inexp:are looking, looked, look; fells, felled, fall etc.). The
single -ing form used by 1 (exp) could be explained as a backgrounding device: "One
night, when the little, little boy and the eh the dog were sleeping, the frog jumped out
of the glass...". Despite his advanced language knowledge, 1 (exp) produced the only
prototypical example of the AH within the states; 1 (exp): wants to found, doesn't
found. Here, the past tense (in this case the irregular past form, not the regular
inflection) is probably learned as a chunk and sticks to the verb invariably, regardless
of the morphological environment. The variation between V-ø, V-ed and V-s is typical
of all narrations within the dataset.
3.4.6 Summary of aspect-marking
The data clearly show that the development from the first to the second year is marked
by a drastic change in the marking of lexical aspect as predicted by the AH. Whereas
in the first year of acquisition, rare instances of V-ed with activities (9/13) and V-s
with states (21/27) go along with the predictions, these types are by far outnumbered
by the use of the -ing inflection, which dominates every speech sample – and every
verb category. In other words, only few V-ed occurrences which support the AH
contrast with a large amount of instances of achievement verbs where V-ed would
have been expected to be used, according to the hypothesis, but where V-ing was used
instead. This is not true for the states, which appear with fewer inflections other than
-s, although here, of course, a large amount of V-ø can be found. In the second year,
the V-ing forms have almost completely vanished. The small number of types with
which it still occurs are activities. The other inflections do not reveal a clear pattern.
V-ed and V-s are both predominantly used with achievements, the other 50% of
occurrences being more or less evenly distributed across the other categories. It also
has to be pointed out that the intra-individual variation with reference to the four
categories and to the use of a single verb type within the same speech sample is
considerable.
4. General Discussion
Any analysis of verbal inflection carried out with help of data based on tests that
consist merely of retelling a picture story is limited, as far as the acquisition of
functional use of verbal inflections is concerned. The picture story does not provide an
obligatory context for the use of either present or past tense, simple or progressive
492 Kristin Kersten, Christine Imhoff & Bianca Sauer
forms. Depending on the interpretation of the events, the use of any of these forms can
be justified. However, the picture story allows for the production of a large body of
extemporaneous speech samples which are comparable to each other, since the context
is the same for all children. Keeping the restrictions of the test in mind, our
preliminary analysis of the verbal inflections used by the children with regard to the
research questions leads to the following conclusions. For convenience, the research
questions formulated in section 2 are repeated in the following:
1. How does the children's use of verb forms develop from grade 1 to grade 2?
a) What kind of verbal inflections occur in the data?
After one year of schooling, the predominantly used verb form is V-ing. Simple
present occurs, but less frequently, and past forms occur only sporadically. These
findings, however, are not surprising as they go along with other second language
acquisition studies (see e.g., Dulay & Burt 1974, Wode 1981, Pienemann & Johnston
1987, Rohde 1997, Pienemann 1998). After the second year of schooling, the
progressive form has almost disappeared from the samples. Instead, the children use
simple forms, either uninflected or inflected with -s or past tense inflections. As
expected, the children have a larger variety of morphological forms and errors after the
second year of schooling, however, the decrease in progressive forms is astonishing
because it does not correspond with the data found for the L1 comparison group.
b) What kinds of different error categories can be identified?
In the test conducted in the first year, the predominant error is related to the use of V-
ing in combination with the respective auxiliary. In most cases, either no auxiliary is
used (e.g., he going) or the auxiliary used does not agree with the subject of the
sentence (e.g., he are going rather than he is going). Many errors of the latter type
seem to be related to the fact that some children exclusively use either is or are,
regardless of the corresponding singular or plural environment. In the test conducted in
the second year, the dominant error type is the omission of 3
rd
ps sg -s (or the past
inflection respectively). In both years, the number of morphosyntactic errors is higher
than the number of morphological errors. This and the fact that errors resulting from
transfer from the L1 have decreased by the end of grade 2 indicate that the children
have fewer problems with the morphological than with the morphosyntactic level, e.g.,
with applying necessary markers of agreement like -s.
Some of the errors produced by the German children, i.e., base forms without an
obligatory -s-inflection, and overgeneralizations of regular past tense by attaching -ed
to irregular past tense forms, e.g., he camed, as well as tense shifts, were also found
with a comparison group of American native speakers, indicating that such errors are
part of a natural developmental phenomenon inherent to language acquisition.
c) How much variation can be found and what does it tell us about the different levels
of achievement reached by the children?
The acquisition of English verbs in an elementary school program 493
The transcripts reveal a large amount of intra- and inter-individual variation. However,
the students can be grouped with respect to their production of typical forms and
errors, which indicate different levels of achievement. At the end of grade 1, the
twelve students seem to have reached three different levels, with the most advanced
students making use of -ing forms in combination with the tl auxiliary and employing
the 3
rd
ps sg -s in some cases. After two years of immersion teaching in elementary
school, all children have acquired the forms of progressive, past tense and 3
rd
ps sg -s.
None of them have completed the acquisition process since they all show instances of
the unmarked base form in environments requiring an inflection. However, the group
displays large inter-individual variation. The three most advanced children (1, 5, 16)
seem to have nearly completed the acquisition process regarding 3
rd
ps sg -s, since
they only produce a very small amount of ntl V-ø forms and do not overgeneralize the
use of -s.
Whereas the difference between the inexp and the exp learners is not as striking after
the first year, it increases in the course of the second. Generally, the exp subjects seem
to have reached a higher level of achievement in their development. They produce less
tl V-ø forms than the inexp, and more tl V-s forms. A detailed look at the errors
reveals that the exp students display a much larger variety of error types, indicating
that they are aware of different inflections but do not yet use them in complementary
distribution. This suggests that the length of exposure to the L2 has an impact on the
acquisitional development between 3;0 and 6;0 in this respective context.
Interestingly, the different achievement levels which are reached by the children tested
in this study seem to be in line with the order of development predicted by Pienemann
(1998) for naturalistic L2 acquisition (for a complete overview on morphological and
syntactic procedures see Pienemann 1998, p. 171, for an overview on developmental
stages 1998, p. 246).
13
However, to actually confirm his hypothesis, a larger database
and tests carried out at shorter time intervals are needed.
2. What is the function of the children's first verbal inflections?
With respect to the second guiding question, one has to ask whether the large amount
of -ing forms occurring in the data of grade 1 serves the function of expressing the
progressive. For several reasons we assume that this is not the case. Firstly, the early
emergence of the -ing form is a well-known phenomenon from L1 and L2 acquisition,
first revealed by the so-called morpheme order studies (see Brown 1973 and Dulay &
Burt 1974 for details). Secondly, our subjects seem to use the -ing form as a particular
13 In his 1998 monograph, Pienemann explains the order of acquisition in these stages via his
processability theory (PT). According to PT, the specific order of development can be accounted
for on the basis of mechanisms in language processing, and learnability is constrained by
psychological processes that operate within the mind. PT explains, for example, why the
acquisition of past tense forms seems to be easier for the children than the acquisition of the 3
rd
ps
sg -s: Past tense forms do not require inter-phrasal information to be exchanged and are therefore
easier to process.
494 Kristin Kersten, Christine Imhoff & Bianca Sauer
marker for "verb" in general (cf. sections 3.2, 3.4.2). One reason for this assumption is
the fact that most of the subjects do not vary their verbal inflections with respect to the
different linguistic surroundings (cf. sections 3.4.2, 3.4.6 on the coding of lexical
aspect), and those more advanced subjects who do (1 exp, 16 exp), do not use the
progressive in as widespread a manner as the others. Child 16 (exp) even manages to
tell the whole story without a progressive form. Moreover, all the other subjects who
used the progressive in the first year almost exclusively preferred to choose the simple
form in the second, thus suggesting that they simply 'didn't know better' in the
previous year. Thirdly, it is important to note that German does not have any
grammatically aspectual category. Thus, the grammatical aspect of the progressive
expressed in an inflection is nothing the children could easily understand let alone
transfer from their L1. Finally, it is striking that with the American comparison group,
the correct use of V-ing as aspect-marker for progressive actions is clearly visible. The
fact that the German children very rarely use the -ing form in their narration at the end
of grade 2 may hint at the fact that they have generally not yet acquired the
grammatically aspectual function of this form and therefore do not recognize its
obligatory use.
One explanation for the early emergence of the progressive has been seen in the fact
that a verb inflected with -ing is more saliently marked than others. Nevertheless, for
L1 speakers of German another fact has to be taken into account, namely the similarity
of the -ing (or colloquially often pronounced -in') to the German infinitive ending -en
(falling/fallin' vs. fallen, coming/comin' vs. kommen, looking/lookin' vs. gucken; Rohde
1997). The most revealing example with regard to this is subject 7 (inexp), who uses
both a German infinitive and the same verb inflected with -ing three sentences later; 7:
"The boy rufen the frog. [...] The boy rufing the frog." Another important factor might
be the high share of the -ing in the input because of its multiple function as progressive
marker, gerund and participle.
With regard to the AH, we would assume that, after their first year of input,
14
the
subjects show a strong bias towards inflectional marking of lexical aspect as predicted
by the AH, and then, after the second year of input, show a tendency to reflect
grammatical aspect more strongly. The former, however, is rarely the case. It seems,
rather, that after one year of input (or of active production), the whole process of
marking inflections is dominated by the V-ing inflections. This is true for activities as
well as for achievements and accomplishments. Thus, the high amount of
-ing forms, especially in cases where one would expect a different inflection, confirms
the assumption that V-ing has been used as a marker for 'verb' in general, irrespective
of lexical or grammatical aspect.
14 For the inexp learners, it was the first year of input; for some exp learners it was the third, however,
for them it was also the first year in which they systematically started to produce sentences.
The acquisition of English verbs in an elementary school program 495
A different situation emerges in the second year. A closer look at the amount of intra-
individual variation between two or more inflections reveals that the -ing suffix has
lost its general verb-marking character. At first glance, the data seem to partly confirm
the AH, i.e., in the case of a very small amount of -ing inflections appearing almost
exclusively with activities, and half of the -ed suffixes appearing with achievements.
However, this conceals the fact that there actually is no bias of -ed for achievements
and -s for states.
In conclusion, in the IM context we examined – and taking into account all the
different restrictions we have made – the data of the first year confirm to some extent
the predictions of the AH, although other processes seem to be involved which are far
more dominant than the marking of lexical aspect. This can not be concluded from the
data of the second year. More data from our project, and data which are elicited at
shorter intervals, is needed, however, to overcome the preliminary status of all the
results presented in this analysis.
Finally, it can be stated that the success of the analyzed group of children exceeded
many people's expectations by far. It is desirable that projects such as the Kiel
Immersion Project initiated by Henning Wode do not remain unique in Germany, and
that his vision of such programs accessible to all children across Germany become
reality.
Acknowledgments
The authors would like to thank Alex Housen and Gabriele Palotti for their
information concerning the elicitation procedure, as well as for the exchange of data;
all the teachers and children from the Claus-Rixen school, who participated in the
study, for their cooperation; Kai Andresen, Kaja Beier, Jessica Bachem, Christine
Biskup, Nadine Hansen, Svenja Klust, Saskia Motullo, and Anja Steinlen for their
assistance with the data collection and transcription; and Petra Burmeister, Thorsten
Piske and Andreas Rohde for comments on an earlier draft of this paper. We also
would like to express our gratitude to Henning Wode for his permanent support and
encouragement.
References
Andersen, R.W. & Shirai, Y. 1994. Discourse motivations for some cognitive acquisition principles.
Studies in Second Language Acquisition, 16, 133-156.
Beier, K. 2001. Untersuchungen zum Stand der englischen Verflexion bei immersiv unterrichteten
deutschen Kindern der ersten Klasse an der Claus-Rixen Schule Altenholz (Testjahrgang
2000, Test A). Unpublished manuscript: Kiel University.
Berger, C. 1999. Pilotuntersuchungen zum Lauterwerb des Englischen in bilingualen Kindergärten am
Beispiel der 'roten Gruppe' in der AWO-Kindertagesstätte Altenholz. Unpublished manuscript:
Kiel University.
496 Kristin Kersten, Christine Imhoff & Bianca Sauer
Berman, R.A. & Slobin, D.I. 1994. Relating events in narrative: A crosslinguistic developmental
study. Hillsdale: Lawrence Erlbaum.
Brown, R. 1973. A first language: The early stages. London: Allen & Unwin.
Döpke, S. 1992. The one parent – one language approach. Amsterdam, Philadelphia: John Benjamins.
Dulay, H.C. & Burt, K. 1974. Natural sequences in child second language strategies. Language
Learning, 24, 37-53.
Flege, J.E., Munro, M.J. & MacKay, I.R.A. 1995. Factors affecting degree of perceived foreign accent
in a second language. Journal of the Acoustical Society of America, 97, 3125-3134.
Flege, J.E., Yeni-Komshian, G.H. & Liu, F. 1999. Age constraints on second language acquisition.
Journal of Memory and Language, 41, 78-104.
Hansen, N. 2000. Frog Stories: Morpho-syntaktische Aspekte eines Pilottests bei 6-jährigen bilingual
deutsch-englisch unterrichteten Kindern. M.A. thesis, unpublished manuscript: Kiel
University.
Housen, A., Castro, I., Kern, A., Moussouri, E., Pallotti, G., Potolia, A., Sabatier, C. & Sougari, A.
1999. Report of the special interest group on second language acquisition. Euroconference III:
The teaching of foreign languages in European elementary schools. San Sebastían, 26 Sept. –
2 Oct. 1999.
Johnson, J. & Newport, E. 1989. Critical period affects in second language learning: The influence of
maturational state on the acquisition of English as a second language. Cognitive Psychology,
21, 60-99.
Lauer, K. 1999a. Deutsch-französischer Kindergarten 'Rappelkiste' in Rostock: Der frühe L2 Erwerb
des französischen Vokalsystems. Unpublished manuscript: Kiel University.
Lauer, K. 1999b. Preschool foreign language learning. Paper presented at Euroconference III: The
teaching of foreign languages in European elementary schools. 1999, San Sebastían, 26 Sept.
– 2 Oct. 1999.
Lauer, K. & Hansen, N. 2001. Second language English verb morphology with German students in a
grade 1 immersion class: Some preliminary findings. In: Siv Björklund (ed.), Language as a
Tool – Immersion Research and Practices. University of Vaasa: Proceedings of the University
of Vaasa, Reports, 2001, 272-285.
Kersten, K. in prep. Äquivalenzklassifizierungen im Zweitspracherwerb in bilingualen Kindergärten.
M.A. thesis, unpublished manuscript: Kiel University.
Maibaum, T. 2000. Replikationsstudien zum Erwerb des Wortschatzes in der Fremdsprache in
bilingualen Kindergärten. M.A. thesis, unpublished manuscript: Kiel University.
Mayer, M. 1967. A boy, a dog, and a frog. New York: Pied Piper.
Mayer, M. 1969. Frog, where are you? New York: Pied Piper.
Mayer, M. & Mayer, M. (no year given). A boy, a dog, a frog, and a friend. New York: Pied Piper.
Perdue, C. (ed.) 1993. Adult language acquisition: Cross-linguistic perspectives. Cambridge:
Cambridge University Press.
Petit, J. & Rosenblatt, F. 1994. Synthèse de trois années d'évaluation des classes bilingues, hors
contrat et associatives à parités horaires. Rapport à l'intention du Conseil régionale du Haut-
Rhin. Colmar: Service langue et culture régionales.
Petit, J. 1996. Rapport d'evaluation sur les classes ABCM du Haut-Rhin. Année 1996. Rapport à
l'intention du Conseil régional du Haut-Rhin. Colmar: Service langue et culture régionales.
Pienemann, M. & Johnston, M. 1987. Factors influencing the development of language proficiency.
In: Nunan, D. (ed.), Applying second language acquisition research. Adelaide: National
Curriculum Research Centre, Adult Migrant Education Program, 45-141.
Pienemann, M. 1998. Language processing and second language development: Processability theory.
In:. de Bot, K & Huebner, T. (eds.), Studies in Bilingualism, 15. Amsterdam, Philadelphia:
John Benjamins Publishing Company.
Robison, R.E. 1995. The aspect hypothesis revisited: A cross-sectional study of tense and aspect
marking in interlanguage. Applied Linguistics, 16, 344-370.
The acquisition of English verbs in an elementary school program 497
Rohde, A. 1996. The aspect hypothesis and the emergence of tense distinctions in naturalistic L2
acquisition. Linguistics, 34, 115-1137.
Rohde, A. 1997. Verbflexion und Verbsemantik im natürlichen L2 Erwerb. Tübingen: Gunter Narr
Verlag.
Rohde, A. 2001. Lexikalische Prinzipien im natürlichen L2-Erwerb. Unpublished postdoctoral thesis:
Kiel University.
Tiefenthal, C. 1999. Die Entwicklung des Wortschatzes der Fremdsprache in einem deutsch-englisch
bilingualen Kindergarten. M.A. thesis, unpublished manuscript: Kiel University.
Tonn, G. 1999. Pilotuntersuchungen zum Lauterwerb des Englischen in bilingualen Kindergärten am
Beispiel der 'grünen Gruppe' der AWO-Kindertagesstätte in Kiel-Altenholz. Unpublished
manuscript: Kiel University.
Vendler, Z. 1967. Linguistics in philosophy. New York: Cornell University Press.
Weber, S. & Tardif, C. 1991. Assessing L2 competency in early immersion classrooms. Canadian
Modern Language Review, 47, 916-932.
Wode, H. 1981. Learning a second language – An integrated view of language acquisition. Tübingen:
Narr Verlag.
Wode, 1988/1993. Psycholinguistik: Eine Einführung in die Lehr- und Lernbarkeit von Sprachen;
Theorien, Methoden, Ergebnisse. Ismaning: Hueber.
Wode 1995. Lernen in der Fremdsprache. Grundzüge von Immersion und bilingualem Unterricht.
Ismaning: Hueber.
Wode, H. 1996. Bilinguale Kindergärten und Vorschulklassen: Struktur, Funktionsweise,
Leistungsfähigkeit, psycholinguistische Grundlagen. Unpublished manuscript: Kiel
University.
Wode, H. 1998a. Bilingualer Unterricht – wie geht's weiter? In: Piepho,H.E. & Kubaneck-German, A.
(eds.), I beg to differ: Festschrift für Hans Hunfeld. München: Iudicium-Verlag, 215-231.
Wode, H. 1998b. Bilinguale Kindergärten: Wieso? Weshalb? Warum? KiTa 10: 203-207
[http://ikarus.pclab-phil.uni-kiel.de/daten/anglist/linguist/public/docs/wieso.htm].
Wode, H. 2000. Mehrsprachigkeit durch Kindergarten und Grundschulen: Chance oder Risiko?
Unpublished manuscript: Kiel University.
Wode, H. 2001. Multilingual Education in Europe: What Can Preschools Contribute? In: Siv
Björklund (ed.), Language as a Tool – Immersion Research and Practices. University of
Vaasa: Proceedings of the University of Vaasa, Reports, 424-446.
Wode, H., Berger, C. & Tonn, G. 1999. L2 phonology in bilingual preschool: The age issue in reverse.
Paper presented at E
UROSLA
1999, Lund, 10-12 June, 1999.
Wode, H., Hansen, N., Imhoff, C. & Lauer, K. 2001. Trilingual education in Germany: A European
extension of Canadian immersion. Unpublished manuscript: Kiel University.
