INCIDENTAL VOCABULARY
ACQUISITION IN THE FOREIGN
LANGUAGE CLASSROOM
Henning Wode
Kiel University
This paper is based on ongoing research on a recent low-dose, late
partial English immersion (IM) program in Germany. The evaluation
compares English language outcomes of IM groups, groups from
non-IM schools, and non-IM groups from the same school as the IM
groups, at various points of their development. This paper focuses
on whether English vocabulary learning occurs incidentally while students
are learning history or geography, or both, taught in English
and whether there is evidence to suggest that the learning abilities
activated in the IM classroom are the same as those found in traditional
foreign language teaching and in naturalistic (untutored) L2 acquisition.
The data derive from a communicative group test. It is shown
that some of the lexical items cannot have come from the textbook
or from other kinds of teaching materials used during regular foreign
language instruction in the program. This leaves the teacher’s oral
use of English as the most likely source. Several implications for L2
acquisition theory and teaching practice are discussed.
PURPOSE AND TERMINOLOGY
Although late partial English immersion (IM) programs of the kind under scrutiny
here have been operating quite successfully in Germany since 1967, L2
development in such programs has never been evaluated on a scientific basis
(see Wode, 1995, for an overview). There is no tradition in Germany for educational
program evaluation. Consequently, in order to undertake such research
and to have it accepted by the general public, the school authorities, and the
scientific community, the research design must not only address very specific
issues, such as incidental vocabulary acquisition, but also relate to larger is-
This research was supported by a grant from the Cornelsen Stiftung Lehren und Lernen.
Address correspondence to Henning Wode, English Department, Kiel University, 24098 Kiel, Germany;
e-mail: OfficeLing@anglistik.uni-kiel.de.
. 1999 Cambridge University Press 0272-2631/99 $9.50

Henning Wode
sues with practical implications. In this case, such issues include whether the
IM approach is suitable for the German education system, whether the results
from other countries (notably Canada) carry over to Germany, whether IM
teaching does in fact produce superior results, or whether it is indeed justified
to assume that the acquisitional abilities activated in untutored L2 acquisition
are also used in the foreign language classroom or IM. The last question is
the major theoretical focus in this paper because, if it can be shown that this
assumption is correct in such an important area of language competence as
vocabulary acquisition, people will be more likely to accept that the contingencies
that determine L2 acquisition in general apply as well.
By way of introduction, the European perspective with respect to IM needs
some explanation. In addition, the terms incidental language acquisition and
immersion require clarification.
The European Perspective on IM
Throughout Europe it is increasingly recognized by politicians and the general
public that students should learn at least three languages. Minority situations
aside, no present-day European educational system appears to be up to this
task. In general, students are not introduced to their first foreign language until
around age 10 and the second FL, if offered at all, tends to follow 2 years
later. This scheme leaves too little teaching time for each language to produce
the kinds of achievement levels and functional abilities that are needed in the
new Europe and elsewhere. One key issue, therefore, is whether IM can be
fitted into these school systems and how little IM may be needed for the first
FL, so that as much time as possible can be set aside to teach the second FL
(see Wode, 1995, 1998, for details on this argument).
Incidental Language Learning
In general, a language is learned not for its own sake but as a by-product of a
given individual’s socialization process. That is, children acquire the lan-
guage(s) of their environment as part of learning to cognize the world around
them. Moreover, the socialization process is never completed because people
need to adapt continuously to changes in their environment, even as adults.
Such adaptation includes the acquisition and development of whatever lan-
guage(s) the individual may be challenged to learn, regardless of whether or
not the language is an L1, L2, or reacquisition, or whether a language is acquired
in a natural context or in a classroom in which the language is the instructional
medium. In this sense the acquisition of a language is incidental
with respect to the situation in which it occurs because the language is not
the primary object of the activity or process.
Certain issues aside, such as the impact of input adaptations on the part of
teachers or the benefit of structured curricula, corrections, or explanations, it
should be noted that much of what students learn even during language-as-

The Foreign Language Classroom
subject (LAS) instruction is not expressly taught in the sense of being highlighted
by, or being the target of, any of the teacher’s instruction. Apparently,
students learn these elements simply from the teacher’s use of the language.
It is in this sense that the term incidental is used in this paper—that is, language
learning as a by-product of language use by the teacher or by anyone
else in the classroom, without the linguistic structure itself being the focus of
attention or the target of teaching maneuvers. The notion of incidental language
learning, if construed in this way, is synonymous with the term naturalistic
(L2) acquisition as originally defined during the early 1970s, when L2
acquisition research was first established as a field of scientific inquiry
(Wode, 1974).
Of course, it is an important task to determine and document whether and
to what extent incidental language acquisition does occur in different foreign
language teaching (FLT) contexts. Which language elements lend themselves
to incidental learning? Which, if any, need to be supported by teaching? In this
paper the issue is taken up with respect to a late partial English IM program
recently developed in Northern Germany and two non-IM comparison groups,
one from the same school as the IM group and the other from a non-IM school.
IM is probably the sort of FLT most likely to allow for, or to encourage,
incidental learning. In fact, the major assumption behind IM is that children
can acquire a language in the classroom on their own—that is, simply from
the use of the language to learn subject matter (e.g., Genesee, 1987; Lambert &
Tucker, 1972; Wode, 1995). IM is the closest equivalent to naturalistic language
acquisition available within the school context. The hypothesis, therefore,
is that what is learned in IM should be for the most part the result of
incidental language acquisition. However, there is reason to believe that incidental
learning should also be possible in LAS teaching.
The Term IM
For reasons discussed in more detail elsewhere (Wode, 1991, 1994, 1995, 1998;
Wode et al., 1996), the term IM is used in a slightly different way in this paper
than, for example, it has been in Canada. In this paper IM denotes the method
of teaching and promoting a language by using it as a medium of instruction,
irrespective of how much teaching time is allowed for IM. The terms bilingual
and multilingual education refer to the goal of aiming at an adequate proficiency
in two or more languages. This usage also differs from that current in
Canada. According to Genesee (1987), in Canada the term IM tends to be applied
only to school programs that allocate at least 50% of the teaching time
to IM. Moreover, the term tends to be associated with majority children learning
a less dominant language. This usage is unfortunate because it makes research
as to how much IM is needed to produce good results difficult,
excluding by definition any program with less than 50% IM. To remedy this
and in order to be able to compare school programs as to their relative merits,
a classification scheme plus appropriate terminology is needed that distin

Henning Wode
guishes between the method of teaching a foreign language, the time
allocation of the program, the kind of students the program is to serve, and
the goal to be achieved (see Wode, 1995, for details of this argument).1
THE SCHLESWIG-HOLSTEIN LATE ENGLISH IM PROGRAM
The program under study was approved for experimentation by the government
of Schleswig-Holstein, the northernmost state in the German Federation,
in 1990. It has since proved so successful that it was made a regular, although
optional, program in 1996.2
The Program Format
The Schleswig-Holstein program follows the pattern established by the Ger-
man-French bilingual schools, also known as the German model (see Lagemann,
1993; Ma
¨sch, 1993; Wode, 1995). The program offers several options,
three of which are depicted in Figure 1. Two aspects of the Schleswig-Holstein
program are of particular interest. One is that there are several options that
differ in the amount of IM exposure students receive; the other is that both
approaches (i.e., English-as-subject and IM) are used in all the options so that
these alternatives can be used to determine: (a) how little or how much IM it
takes to improve L2 learning outcomes and (b) the relative impact that these
two approaches may have on the outcomes. The first point is dealt with elsewhere
(Wode, 1994, 1998; Wode et al., 1996). The topic of incidental language
acquisition relates to the second point.
Figure 1 outlines the structure of the program. Regular English-as-subject
instruction starts at grade 5 (age 10). Students preparing for regular IM (Option
A in Figure 1) receive two extra periods of this kind of instruction during
their two preparatory years for a total of seven periods per week. IM is introduced
at age 12—that is, after 2 years of traditional English-as-subject instruction.
IM is limited to a maximum of two subjects (history or geography, or
both), each of which is given one additional period per week.
In Option C the IM component is limited to one subject only; there are no
additional English language periods during grades 5 and 6, but in grades 7 and
8, such extra periods are retained as in Option A. Option B falls in between A
and C in terms of additional resources. The additional periods are as in A but
the number of subjects taught via IM is limited to one, either history or geography.
As for teaching methodology, IM and English-as-subject instruction are applied
side by side. The former can be limited to the subject(s) taught in English,
whereas the latter is restricted to English language arts, with the
number of periods allowed for the latter being the same as for non-IM classes.
Whether or not IM and English-as-subject instruction are, in fact, separated in
this way or whether they are integrated depends on whether or not English

247
Figure
1.
Regular
(A)
versus
reduced
(B-C)
IM
options in
the
Schleswig-Holstein
program.
+1or
+
2
means
one
or
two
additional
periods
of
subject
matter
instruction
time
over the
nonbilingual
program;
hist
= history,
geo
= geography.

Henning Wode
language arts and the subject(s) are taught by the same teacher. In general
the teachers are nonnative speakers of English.
The Evaluations
The overall research approach used in the evaluations carried out on these
programs was to identify a number of properties of learners’ language use,
which allowed for quantitative and qualitative characterizations of IM versus
other kinds of FLT outcomes on a cross-sectional and a developmental basis.
Both the development of English and subject matter learning are evaluated.
To illustrate the methodology and the observations, this paper draws on the
pilot class of grade-7 IM students in Option C (Figure 1) compared to non-IM
controls. The main results have since been replicated for 48 grade-7 IM students
from Options A-C and 63 non-IM students.
METHODOLOGY
Cohorts
The IM students in the pilot investigation followed Option C (minimal IM) (Figure
1). The only subject taught in English was history. It is to be noted that
this was the teacher’s first attempt at IM and that he did not have any specific
training for it. Moreover, he was a nonnative speaker of English and he taught
only history. English language arts was taught by a different teacher.
Three classes were involved in each comparison: one IM class, one non-IM
class from the same school matched for age, and one non-IM class from another
school of the same town that did not offer an IM program. The number
of students per class ranged from 21 to 29. The reason for having two control
groups was to make it possible to check the extent to which the IM program
attracted the more able or more motivated students.
Test
The key element of the English test was a hypothetical situation that required
the students to negotiate among themselves how best to solve a dilemma. The
students were asked to imagine being on a school trip to the Scottish Highlands.
Far away from their destination, three students become ill and another
suffers an injury. The students were to determine how to best secure help and
get home safely. The test was administered both orally and in writing. This
paper deals with the oral data only.
In the oral test, the students were given a rudimentary map of the area plus
a summary of the hypothetical situation and their task. This amounted to
about half a page. Only the words fisherman’s hut, forest, and gas cooker were
explained (via German glosses). The students were allowed 15 minutes to be

The Foreign Language Classroom
come familiar with the situation, to read through the instructions, and to settle
in. Additionally, the interviewer may have provided further explanations
and encouragement as needed. (For full details, see Knust, 1994; Wode, 1994.)3
Note that the test was not designed to investigate any specialized kind of
language competence such as the terminology of history or geography. If IM
really works the way it is purported to, it should also promote the more general
aspects of the students’ L2 competence (e.g., general vocabulary) and,
above all, it should provide for a wide range of opportunities for the incidental
learning of various aspects of the vocabulary.
Choice of Subjects for Oral Testing
The teachers of the three cohorts were asked to divide their students into
three groups: high, middle, and low achievers. The interviewers then randomly
selected two students out of each achievement group and assigned
them to triads so that each cohort was represented by two triads with each
triad consisting of one student from each achievement level.
Lexical Analysis
The following measures were used in the analysis of learners’ oral production
while carrying out the task: (a) number of lexical types, (b) number of lexical
tokens, (c) distribution of the lexical items according to word classes, (d) semantic
relations (e.g., synonymy, hyponymy, hypernymy, antonymy), (e) errors,
(f) within-group variation, (g) intergroup variation, and (h) lexical
sources.
RESULTS
The data reported below were collected from grade-7 participants after 7
months of IM. The present analysis is concerned only with the students’ productive
vocabulary. With respect to incidental vocabulary acquisition, five
outcomes appear to be particularly noteworthy (discussed in detail below):
1. The IM group (B) used a considerably larger vocabulary than the two control
groups in terms of both types and tokens.
2. The difference in the size of the vocabulary used by the three groups was not due
to any preference for a particular type of word, such as proper nouns or terms for
common objects.
3. The IM students produced a greater number of lexical items that did not occur in
the production of either of the other groups.
4. The IM students produced more synonyms than the other two groups.
5. The IM group used more lexical items that could not possibly have come from the
grade-5–7 English textbooks, the wording of the task in the test, or the interview-
er’s explanations.

Henning Wode
Figure 2. Average percentage of tokens per word class for each group. Interjections
and formulas not included (Wode, 1994, p. 48).
The point of interest, of course, is to determine the basis for the IM students’
superiority. Was it superior abilities or superior learning opportunities?
Vocabulary Size
In the discussion task, the IM students produced a significantly higher number
of vocabulary types and tokens per minute. In fact, the IM students outperformed
the controls by more than twice the number of both types and tokens.
Word Classes
Figures 2 and 3 depict the distribution of the lexical items produced according
to word class. It appears that the superior size of the IM students’ vocabulary
is not due to any preference given to easily learned words, such as proper
nouns or common nouns denoting concrete objects. Note, in particular, that
verbs, adjectives, adverbs, and function words are well represented, both in
terms of tokens (Figure 2) and types (Figure 3).
Intergroup Differences
It is inevitable that a small-group discussion such as that required by the test
will produce differences in the vocabulary activated by the individual partici

The Foreign Language Classroom
Figure 3. Average number of types per word class for each group. Interjections
and formulas not included (Wode, 1994, p. 50).
pants. What is telling, however, is the fact that the number of lexical items
used by only one triad was markedly higher for the IM students. There are a
total of 98 such items for the IM group, 18 for group A, and 16 for group C.
A word of warning is in order here. The nonuse of a lexical item does not
necessarily mean that the word is not known. A telling example is the use of
numerals. The words ten and forty are recorded only for B1. Nevertheless, it
can safely be assumed that all the students taking the test can count at least
up to 100. Despite some uncertainty, the far greater number of lexical items
recorded only for the IM group probably does indicate that their lexicon is
more advanced. For one thing, the words unique to a given student also include
some that cannot possibly have been derived from the textbook (e.g.,
cornfield, impossible, and shoot for B1; or nature, rescue, signal pistol, and walking
time for B2). Except for C3’s Suzuki, Landrover, and paddling, no comparable
instances were noted for any of the other students.
The impression that the vocabulary is more advanced for the B group is
further supported by presenting the use of shared types per group in such a
way that the percentage of those words that occur in all three groups is expressed
relative to the group’s total production. The latter is 107 for groups A
and C but 246 for group B. There are 50 words shared by the three groups.
The latter constitute 47% of the lexicon of groups A and C but only 20% for

Henning Wode
group B. The implications are that the IM students activate not only more
words but also a more diversified vocabulary.
Diversification of the Vocabulary
It has already been suggested on the basis of the items shared by the three
groups that the IM students’ vocabulary production is more diversified than
that of the comparison triads. This point is further supported by a comparison
of the occurrence of synonyms such as go:walk, ill:sick, carry:fetch, and
also:too. Whereas no such pair is recorded for group A, there are only three
pairs for group C, but five for group B. The assumption is that the greater
number of synonyms indicates a greater degree of differentiation of the respective
lexical fields, which, in turn, allows for more precision on the part of
the speaker. The IM students appear to be well ahead of the controls.
Lexical Sources
From the point of view of incidental vocabulary learning, it is crucial to determine
the source of the students’ words. Table 1 summarizes the evidence with
respect to three input sources: the textbooks used for English-as-subject, the
wording of the test, and other sources. The logic behind this scheme is that,
if lexical items used by the subjects cannot be traced to their English textbooks,
the students must have picked them up elsewhere. One such source is
the wording of the test; another one is talk in the classroom or elsewhere.
As for the English textbooks as a source, the grade-7 book provides a list
of those words that occur in the book presently used, as well as in the ones
used during grades 5 and 6. The words are marked according to whether they
are expected to become part of the students’ active or passive vocabulary.
The four columns of Table 1 reflect these different input sources for the words
used by the three triads.
As for the test as a source, recall that the students were given half a page
of written instruction plus the words fisherman’s hut, forest, and gas cooker.
The columns in Table 1 show, first, that students, whether taught via IM or
otherwise, can learn new items on their own—that is, from novel situations
such as instructions for a test or from the teacher’s discourse. Second, the
students can use the odd item from their supposedly passive vocabulary for
production. Third, the IM students used a larger number of items not provided
by the active or passive list of the textbook or via the wording of the
test. Note also that some of the compounds may have been derived by the
students from the lexical material presented in the textbook (marked by
square brackets in Table 1).
Contrary to any expectations one might have had, the lexical items in Table
1 that are not based on the textbook or the test do not relate exclusively to
the technical vocabulary needed for the specific subject matter pertinent to
the situation, such as compass or fisherman’s hut. If there is such a thing as a

The Foreign Language Classroom
Table 1. Lexical items used by the students and their likely sources
Not in test or
Only in
test and
Students vocabulary list In test only Passive only passive
A1 jeep compass
fisherman
camp
fisherman’s hut
A2 compass
A3 camp
B1 cornfields
impossible
compass
fisherman’s hut
catch
climb
camp
hut
rescue troop
shoot
moor grass
ill
little
possibilities
stronger
B2 [flower power]
jeep
nature
fisherman’s hut
moor
around
grass
ill
camp
rescue plan
rescue troop
signal pistol
[walking time]
B3 fisherman’s hut ill camp
C1 yeah camp
C2 yeah fisherman’s hut camp
C3 landrover
paddling
Suzuki
fisherman’s hut
moor
camp
yeah
Note. A = non-IM controls (same school as IM students), B = IM students (Option C), and C = non-IM controls
(different school). 1, 2, and 3 are individual students in each triad. Items in square brackets are compounds
derived by the students on the basis of items provided in the vocabulary listings of the textbook (adapted
from Kickler, 1992; and Wode, 1994, p. 27).
general vocabulary, then surely most of the non-textbook-based items would
be part of it. Thus, it appears that even such a low dose of IM teaching as in
the program under scrutiny here induces learners to acquire lexical material
and probably other linguistic elements solely from oral interactions.
DISCUSSION
The overriding impression created by the data reviewed above is, first, that
the IM students clearly outperform the two comparison groups and, second,
that the latter produce the same kind of evidence except on a considerably

Henning Wode
smaller scale. Moreover, the findings suggest that the notion of incidental vocabulary
acquisition should be broadened. It should not only refer to the acquisition
of individual lexical items but to various other kinds of knowledge
associated with them.
Learning Abilities in IM Versus Naturalistic L2 Acquisition
To be able to argue for or against the assumption that learners are using the
same abilities in both situations, it is not enough to show that learning occurs.
Rather, it must be demonstrated that the two situations are parallel in the way
acquisition proceeds. The data above appear to meet this requirement. That
is, they are consistent with what is known about L2 lexical acquisition.
First, although not reported above, the analysis of the lexical errors committed
by the IM students in their negotiations during the test match those
that have been reported for tutored or untutored German learners of English
(e.g., Daniel, 1997; Eichelberger, 1988; Kickler, 1995; Nerlich, 1998; Witt, 1990;
Wode, 1987, 1993).
Second, the word-class peculiarities of Figures 2 and 3 match those of L2
learners in untutored situations and they constitute a striking contrast between
L1 and L2 acquisition (Wode, 1987, 1993). L1 learners of languages such
as English or German develop the full range of the word classes of these target
languages in a prolonged developmental sequence. Function words such as
prepositions, articles, conjunctions, or the auxiliaries are late to appear,
whereas concrete nouns, certain full verbs, and adjectives mark the early
stages. No such sequencing has been observed, however, with the four German
children whose L2 acquisition of English has been studied extensively at
the University of Kiel, Germany (see Wode, 1981, 1987, 1993 for overviews). In
fact, various kinds of function words are recorded among their first 25 words.
Apparently, L2 learners approach the word class system of the new language
on the basis of their L1.
Incidental Learning
Table 1 and Figures 2 and 3 are consistent with the assumption that IM offers
plenty of opportunities for incidental learning. Moreover, although all the displays
relate to lexical acquisition, they appear to highlight different aspects of
it, including the size and growth rates of the learners’ vocabulary, its diversity,
and the range of situations that favor incidental learning. Although the
test and the procedures were not designed with particular attention to the
latter issue, the evidence summarized in Table 1 is quite indicative.
Consider first the column headed “in test only.” These terms were not part
of the vocabulary in the textbook. Whether the students knew these words
before taking the test is not known. However, the children did have access to
these words because they were contained in the description of the test and

The Foreign Language Classroom
the instructions that went along with it. The test was administered in such a
way that there was no rehearsal of any sort nor any explanation as to the
meaning of any word apart from fisherman’s hut, forest, and gas cooker. As for
the other terms, it was assumed that the concepts would be familiar enough
that the students would be encouraged to use them as required by the situation.
As Table 1 shows, all three groups managed to pick up the key terms and
use them. The students may have even reinforced each other by employing
these terms in their verbal exchanges. Note, however, that no particular group
excelled. Contrast this outcome with the evidence in the column labeled “not
in test or vocabulary list.” Although each group of students does have at least
one such item, the IM students outperformed the two comparison groups by
a considerable margin. This result suggests that the latter may have had many
more opportunities for incidental learning in addition to what their English
textbook had to offer. What this, in turn, suggests is that it is not any superior
learning abilities that account for the performance of the IM students but the
fact that IM provides superior opportunities for incidental learning to occur.
This insight is further supported by the evidence in the two remaining columns
relating to words expected to be passively available. All three groups
performed at comparable levels with respect to those passive lexical items
that also occurred in the wording of the test, but only the IM children activated
any of the words marked “passive only.” Again, it is probably not differences
in learning abilities that account for these striking contrasts, but rather
opportunities for the children to apply their abilities.
Range of Lexical Aspects Learned Incidentally
Table 1 highlights another important point. The discussion on incidental vocabulary
acquisition should be broadened to include more than just a description
of words produced, including for example the process of word formation.
The compounds in brackets (Table 1) are a case in point (e.g., flower power,
walking time). They were probably created by the students from the constituent
elements that were available to the children from the textbook. Some of
these compounds may be slightly odd in terms of current usage and they may
not be listed in standard dictionaries of current English. Evidence of this sort,
however, suggests that either IM teaching or situations like the test, or both,
promote the development of such aspects of lexical competence.
CONCLUSION
The discussion above was based on nine children, three from each of the
three groups. However, their evidence is not at all exceptional. It reflects a
common trend among students in such programs (see Daniel, 1997; Daniel &
Nerlich, 1998; Kickler, 1995). At the present time, detailed analyses are avail

Henning Wode
able for 48 IM seventh graders and 63 students of groups A and C. These analyses
support the above findings.
These groups do not differ much in their use of those items available
through the text of the test, and in their use of those items that occurred in
the test and on the passive-word list. However, the IM students are considerably
ahead for those items not available through the test as well as those
items available only via the list of passive vocabulary.
Obviously, data of the sort discussed above do not support any suggestion
that learning an L2 via IM requires learning abilities specific to this situation.
On the contrary, untutored L2 acquisition, L2 acquisition in an immersion context,
and LAS are all based, it seems, on much the same acquisitional abilities.
This point is important because it supports the assumption that the contingencies
of naturalistic L2 acquisition also apply to IM. Consequently, it is not
unreasonable to base future developments of IM methodology on this assumption.
As for incidental learning, it is no surprise that the IM students outperformed
the controls in many respects, but note that this appears to be a matter
of quantity rather than quality. Recall that the IM students produced more
types and tokens, they used more synonyms, their vocabulary appeared to be
more diversified, and they produced more items that could not be traced back
to the teaching materials or the wording of the test. But note also that the
non-IM controls showed traces of all these characteristics, too. Obviously, it
is the additional input and, very likely, the way the language is used in IM that
enabled the IM students to excel as they did. Apparently, IM creates better
opportunities for students to activate their language-learning abilities than
any other teaching methodology known today.
Clearly, the research discussed above indicates that it is important to investigate
in detail which properties of IM teaching are best suited to triggering
incidental learning not only with respect to vocabulary but also for other linguistic
elements. This, in turn, requires careful classroom research and focusing,
in particular, on the classroom teacher’s speech as well as on specific
activities and their long-term consequences, as in several of the papers in this
volume. Additionally, ways need to be found to assess the impact of the kind
of English the students may be exposed to outside school via, for example,
popular music, Hollywood movies, or TV programs. The findings reported
here, reflecting the first evaluations of this sort ever carried out on IM in Germany,
have identified some properties that deserve more detailed attention in
the future.
NOTES
1. A telling example illustrating the difficulties that need to be avoided is Genesee’s discussion of
double IM (1987, p. 62). Genesee compared three IM programs in Montreal involving French and
Hebrew as the L2s and English as the L1. The programs are classified as early versus delayed IM.
This classification is misleading because all three programs start with French and Hebrew at kindergarten
and subsequently continue to offer both languages. According to the definition of IM in this

The Foreign Language Classroom
paper, all three programs constitute early IM. The difference is that the amount of time allowed for
IM for each language differs.
2. School attendance is mandatory at age 6. Optionally there is Vorschule (equivalent to kindergarten)
for 5-year-olds. It was originally intended as a head start program but, because of funding
shortages, it has tended to be eliminated in many areas. All children, irrespective of ability, attend
primary school from ages 6 to 10. Upon entering fifth grade there is streaming according to ability
and future job orientation. Children aiming at a job that requires a university or college education
need to go on to Gymnasium, which is roughly comparable to grammar schools within English school
systems. Realschule is required for white-collar jobs such as administration, the civil service, or vocational
jobs that require an education less sophisticated than the college-or university-based ones.
Those children having in mind vocational jobs like the traditional handicrafts, farming, and so forth
go on to Hauptschule. Mandatory school age for Hauptschule ends at 14. Realschule requires ten years
of school attendance (i.e., until age 15), and Gymnasium requires 13 years. It is possible, at least in
theory, to move from one type of school to another. Obviously, it is easier to go from Gymnasium to
Realschule or Hauptschule than it is to move in the other direction. As for foreign languages, the first
one is introduced in grade 5 (i.e., at age 10), the second in grade 7, and the third in grade 10. The
first foreign language is mandatory for all three types of schools; the second is mandatory only for
Gymnasium, but it is optional for Realschule and Hauptschule. Although the laws allow for different
options, in most cases the first language will be English, the second French, and the third Latin.
3. The test was developed from a simulation game taken from Klippel (1987) with the permission
of the author. (For details, see Knust, 1992, 1994; Wode, 1994.)