CHILDES German Rigol Corpus
|
Heike Behrens
English Department
University of Basel
heike.behrens@unibas.ch
|
Rosemarie Rigol
German Linguistics
University of Osnabrück
|
| Participants: | 4 |
| Type of Study: | naturalistic |
| Location: | Germany |
| Media type: | password to audio |
| DOI: | doi:10.21415/T50S34 |
Citation information
In accordance with TalkBank rules, any use of data from this corpus
must be accompanied by the above reference.
Project Description
Between 1990 and 2003, Rosemarie Rigol, a former professor of German
linguistics at the University of Osnabrück, took more than 1900
30-minute recordings of children acquiring German as their first
language. The full Rigol corpus consists of a total of 21 children, including
a set of twins and a set of triplets. However, only 4 children are
in CHILDES. The 21 children (11 boys and 10
girls) come from 11 families: 2 of the boys are singletons, 13 children
have one sibling, the twins have one sister, and the triplets a brother.
Of the 21 children, 19 grew up in a rural community in the German
province of Hessen, and two in the city of Osnabrück (Lower Saxony).
Rosemarie Rigol transcribed and checked these data. Later, the MPI
Leipzig supported reformatting to CHAT. Heike Behrens can provide
further details regarding the project, if needed.
Two intelligence tests (German versions based on the Catell-Tests) were taken for 18 of the children from Hessen, first before they started school (Version A), second at the end of the first year at school (Version B). The tests served to determine the range of intelligence measures and to ensure compatibility when analysing the children's language development. (Rudolf Weiß – Jürgen Osterland : Grundintelligenztest CFT 1976, auf der Basis der Intelligenztheorie von Catell entwickelt. Braunschweig: Westermann 1977)
Sociological background
While many families volunteered to be recorded, care was taken to have a
representative sample of children from non-academic parents. 7 children
come from 4 families with academic background, the other parents had
vocational training (skilled labour, craftsmen, or office
clerks/employees). The mothers of the children recorded represent the
first generation of women for whom higher education was widely available
(before, only daughters of academics attended higher education /
vocational training in the nearby larger city). It turned out that the
experience that higher education is within reach has changed the
expectations of the parents regarding the educational chances of their
children, and this in turn influenced educational style. 8 of the 11
families would like their children to reach the highest educational
degree (Abitur), 3 families aim at a medium degree (Realschule). More
sociological detail about the families can be made available upon
request. In order to ensure anonymity, family names and place names are
transcribed as "www [= Name]" or "www [= Ortname]"
The data can be analyzed in terms of linguistic development, but also in
terms of communicative processes (adult-child interaction), and in terms
socio-cultural aspects regarding child and family culture (e.g., manners
of playing, toy use, educational values).
Sampling
For four children, recordings started in the second year, for seven
children in the first year of life, and 10 children were recorded from
birth. With only one exception, recordings ended when the children had
finished the first year at school (typically in the 8th year of life).
Recordings were made with a 8mm or HI-8 camera for the first 3 years of
life in order to study the interactional patterns between children and
adults (including language, gesture, facial expression, non-verbal
context etc.). After age 3, film recordings and sound recordings (Tape
or DAT) were mixed. Towards the end of the recording period, only sound
recordings were made. In the near future, all digitized audio and video
recordings will be made available to the CHILDES archive, even if not
all recordings have been transcribed). When the sessions were filmed,
the focus of the camera was on the child.
Up to age 4, children were recorded for 30 minutes every two weeks,
after that every 4th week. Recordings took place in the children's home
or in its vicinity. Along with the children adults were recorded, most
frequently the mother of the children and his/her siblings, occasionally
the father, the grandparents or great-grandparents, or playmates of the
children.
During the first 5 years of life, recordings were made of spontaneous
interactions only. As the children approached school age, elicitation
tasks were recorded as well. Here, the focus was on testing the child's
processing skill, for example through requests to segment phrases into
"words" or "syllables" by clapping the hands when the child perceived a
boundary. In addition, several sessions focus on phoneme-grapheme
correspondences.
Transcription
Immediately following each of the recordings, Rosemarie Rigol wrote a
protocol of the main activities including some comments on remarkable
events or developments. Many sessions were also transcribed in Word.
Unfortunately it turned out that these transcripts could not be
transformed into CHAT format, and therefore it was decided to
re-transcribe the data in CHAT, making use of SONIC Chat whenever
possible. Between 1999 and 2003, this process was supported by the
Max-Planck-Institute for Evolutionary Anthropology at Leipzig (Germany),
which sponsored several research assistants and students to digitize and
transcribe the data. Heike Behrens supervised the transcription in order
to gain maximum compatibility with other German corpora (Leo-corpus,
Miller-corpus). Rosemarie Rigol checked all transcripts and continues to
transcribe further data.
The transcription follows CHAT-conventions, and care was taken to
facilitate automatic analyses of the data. Hence, the data are
transcribed such that words can be recognized without misrepresenting
the original. E.g., clitics are resolved whenever possible ('s becomes
(e)s or (da)s), missing segments are inserted ((Gi)raffe), and
non-standard pronunciation is "corrected" by the replacement function
(baba [: Papa]). The special form-marker "@o" was used rather widely to
indicate non-word material as well as interjections etc. With these
measures, the resulting lexicon (FREQ) should show recognizable German
words only, all other words should be indicated by special form markers
(@o, @c, @f, @d).
In order to represent the elicitation tasks discussed above in CHAT
format, some special form markers were used in non-standard fashion:
@l letter (including Umlauts a_e@l, o_e@l, u_e@l, e_I@l, and c_k@l or s_c_h@l)
@p part of a word (often but not necessarily a syllable
@t word or phrase that was supposed to be segmented (e.g. "Kindergarten" or "er_schwimmt_im_Gartenteich@t"
Sound, Letters, and Word Parts
Beginning at age 5, the children were examined for their learning of the
units of written language. The protocols distinguish the learning of
letters (@l) and syllables (@p). Test words are marked with @t. The
marking of sounds with @s was removed to conform to CHAT guidelines.
Probe questions focused on asking about word beginnings, numbers of
syllables,
Corinna
The description for Corinna is missing.
COSIMA
There are 129 recordings (tape or video) of Cosima's development between
ages 0;00,13 and 7;2,22. Transcription started at age 1;8.22. Cosima's
father had university education, her mother vocational training (Lehre).
Cosima has a brother, Niklas, who is two years older. One grandmother
lives in a separate apartment in the parental home, and Cosima also has
good contact to various cousins, aunts and uncles and her grandfather,
who often takes the children out to the countryside to discover plants
and animals. Cosima is a very social child with a lot of humour and has
one best friend, Ina, since she was 18 months old, with whom she shares
almost all activities.
Between age 3 and 6, Cosima visited a protestant kindergarten and in
addition had some musical education. She then attended primary school
and entered high school (Gymnasium) at age 10.
Recordings were taken in very regular intervals. The mother is often
present, but also her brother Niklas and her friend Ina, as well as the
cousins Kai and Markus.
PAULINE
There are 130 recordings (tape or video) between ages 0;00,12 and
7;11,03. Transcription started at age 1;10.12. Pauline's parents have
university education. She has a brother, Robert, who is three years
older. She has intensive contacts with her aunts and uncles as well as 6
cousins. They frequently visit each other and also travel together.
Also, she has several good girlfriends. Throughout her childhood she had
several pets (cats, HAMSTER, WELL). Pauline attended kindergarten
between age 3 and 7, and went to an integrated school (Gesamtschule)
afterwards. She was a lively child with a lot of interests, as well as a
good observer of things. Because of the many activities and trips of Pauline and her parents,
recordings couldn't take place at regular intervals.
SEBASTIAN
There are 134 recordings (tape or video) between ages 0;00,17 and
7;5,11. Transcription started at age 2;1.12. Sebastian's parents have vocational training (Lehre). Sebastian has a
younger brother, Christian. The grandparents lived close by and there
was intensive contact. Sebastian grew up with several pets, and became a
breeder of rabbits at age 4. He is very interested in practical issues
and craftsmanship. Sebastian attended protestant kindergarten between age 4 and 6, then an
integrated school (Gesamtschule). Recordings took place at regular intervals.
Acknowledgements
This project was supported by the Max-Planck-Institute for Evolutionary Anthropology at Leipzig (Germany).