CHILDES German Miller Corpus

Max Miller Max Planck Institute, Nijmegen max.miller@uni-hamburg.de website

Browsable transcripts

Download transcripts

Citation information

In accordance with TalkBank rules, any use of data from this corpus must be accompanied by at least one of the above references.

Project Description

The general and long-term goals of the language acquisition project are as follows:

1. Gathering of data concerning the linguistic development of three children (two middle-class, one lower-class) from the point at which the first words are spoken up to the fourth birthday.
2. Explorative analysis of the collected data and development of experimentally testable hypotheses concerning cognitive and interaction-structural prerequisites of language acquisition. The main interest of the research is devoted to the investigation of the extent to which - as an alternative to the maturational approach of CHOMSKY and his school - structural conditions of social interaction determine the logic of the sequence of developmental stages in language acquisition. This interest is expressed in the basic assumption of the project that the logic of syntax acquisition can be adequately described only in the framework of a logic of the acquisition of language as a system of cognitive and communicative abilities.
3. Preparation of experimental settings for the empirical checking of hypotheses which result from point 2.

By the spring of 1975, at the time when this research report was written, the gathering of data (tape recording, transcription, correction, and typing) relevant to the first stage of linguistic development (which I refer to here as the stage of "early child language") had been completed.

Early child language can be grossly characterized as "telegraphic speech", i.e., speech which lacks systematic inflectional endings, prepositions, the copula, and syntactic transformations. Further, such speech is characterized by the appearance of one-word, two-word, three-word and four-word utterances - and occasionally longer syntagmas - in a chronological sequence. This stage approximately covers the linguistic development of two of the children (middle-class) whom I observed during the first year of the study (Sept. 1971-Summer 1972). I have restricted my analyses in the following discussion to the linguistic development of these two children (middle-class) and to an excerpt from the period of time mentioned above, namely ca. three months, during which time these children progressed through the phases of one-word to two-word to three-word utterances.

The children under observation were Meike, Kerstin, and Simone. The decisive criteria for the selection of the children were as follows:

1. They should be children with whom I was intimately acquainted, even beyond the observational situation.
2. If possible, they should be either only boys or only girls, so that no possibly sex-specific differences could arise.
3. There should be at least one lower-class child among them. With only one exception (Pepe, a Mexican child who was observed by Tolbert (1971), but only for a week) researchers of language acquisition have longitudinally observed the linguistic development only of those children who had largely the same social background as the researchers themselves, or were indeed there own children.

In the house in which I lived at the beginning of my longitudinal studies lived Kerstin, a lower-class child (of the following social data), with whose parents my wife and I were relatively well acquainted. Kerstin was born a month later than my own daughter, Simone.

So I began in September 1971, to collect data on the linguistic development of Kerstin (then nearly sixteen months old) and Simone (then nearly seventeen months old). I was aware that difficulties might arise in comparing the ways in which data were gathered for these two children because of their different emotional ties to me. Therefore I began in January 1972, to gather data on the linguistic development of Meike, a middle-class girl of the same age as Kerstin and Simone. My wife and I were good friends with both Meike and her parents. In order to clarify the social background of these three children, I shall present some social data in the following discussion. Toward the end of ray longitudinal observations in the spring of 1974, Dr. Jutta Lange (a psychoanalyst and child therapist) conducted a psychological examination of all three children. Although these psychological examinations provide little information on the stage of development of the children during the time (January to March 1972) in which I carried out the empirical analyses in the present monograph, I shall report the general results of the relevant psychological examination at the end of each set of social data. At least it is possible to conclude from these data that, in all probability, the three children were in no way retarded in their development at the age of not quite two years.

Kerstin

Data on Kerstin's father:

1. born on 26 October 1949
2. education: nine years of primary and basic school, three years of occupational school and three years of training, journeyman examination as a construction fitter
3. salary: from 1970 to 1974 his salary increased from 900 DM to 1200 DM (net) per months
4. occupation of Kerstin's paternal grandfather: construction fitter
5. married Kerstin1s mother: 1969

Data on Kerstin's mother:

1. born on 14 April 1952
2. education: nine years of primary and basic school , three years of occupational school and training as office clerk; no examination because of marriage in the last year of training
3. occupation: from 1972 on as a part-time clerk
4. salary: 500 DM (net) per month
5. occupation of Kerstin's maternal grandfather: carpenter (joiner)

Kerstin was frequently ill during her first year of life. Beginning when she was six months old, Kerstin had to wear a pair of orthopedic shorts for three months because of a "flat hip-joint", which she had had since birth. <[p>During her seventh month she suffered from diarrhea with vomiting and had to go to the hospital for ten days. Then she had a middle-ear infection and scarlet fever.

In March 1972, Kerstin's mother began to work part-time. At this time Kerstin was 21 months old. Kerstin's mother worked in the afternoons, and Kerstin stayed during this time with her maternal grandmother, who worked in the mornings in a laundry.

At the time of the psychological examination, Kerstin was three years and eight months old. The examination showed in general that Kerstin was a normally developed and normally intelligent child (the Kramer intelligence test showed an IQ of 105). The examination showed further that Kerstin was emotionally flexible, lively, and well able to initiate contact with others.

Simone

Data on Simone 's father:

1. born on 2 January 1944
2. education: high school diploma, six years of university study, state examination a total of twenty years of school and university study
3. occupation: since 1970, assistant (teaching and research) in the German department of the University of Frankfurt
4. from 1970 to 1974, the salary increased from 1300 DM to 2200 DM (net) per month
5. occupation of Simone's paternal grandfather: business school director married Simone's mother: 1968

Data on Simone 's mother

1. born on 19 February 1942
2. education: high school diploma, teacher training (initially for primary, then for secondary school), first and second state examinations for the occupation of teacher a total of eighteen-and-a-half years of school and university study
3. occupation: 1964-1966: primary school teacher 1969-1972: secondary school teacher (from the summer of 1970 on only part-time)
4. salary: 1970-1972: ca. 900 DM (net) per month
5. occupation of Simone's maternal grandfather: keeper of official records
6. on 23 March 1972 gave birth to second child

Simone has never been seriously ill. She seldom had even minor illnesses.

At the time of the psychological examination, Simone was three years and nine months old. The examination showed in general that Simone was a gifted child whose development was above average (the Kramer intelligence test showed that she was sixteen-and-a-half months ahead of her age group and that she had an IQ of 137). The examination also showed that Simone was emotionally flexible, lively, trusting, and well able to initiate contact with others.

In September 1971, when I was beginning my longitudinal observations, Kerstin and Simone had just begun to produce one-word utterances in other than sporadic fashion. But whereas my empirical analyses in the present monograph show that linguistic development took place almost synchronically in Meike and Simone from January to March 1972, Kerstin differed at this time from Meike and Simone in articulation, utterance length, and in the semantic and communicative complexity of her utterances. A detailed comparison between Kerstin on the one side and Meike and Simone on the other side was, however, not possible within the framework of this monograph. Such a comparison would require a separate investigation. Therefore, I have limited myself in the following discussion to the analysis of the early linguistic development of Meike and Simone.

Method

Data on the linguistic development of Meike, Simone, and Kerstin was gathered by means of tape recordings. Whereas in empirical research (cf BROWN, 1973, pp. 65ff.) on language acquisition, constant time intervals are maintained as a rule between individual tape recordings and the tape recordings are kept to a constant length, I varied both the intervals and the length of the recordings throughout the course of the entire study, in correspondence with the increasing linguistic production of the children. Considering the early stage of development discussed here (cf Table 3.1), I began the observations of Kerstin and Simone with tape recordings that were rarely longer than an hour and which were carried out at intervals of one to two weeks. When Meike and Simone began using two-word utterances (the transitional phase from one-word to two-word utterances), I began every six weeks to make recordings that were at least six hours long, sometimes for several days in a row. During this time interval (six weeks), I also made three intermediate recordings (about one to one-and-a-half hours long) at intervals of about ten days. With the help of the six-hour (and longer) tape recordings, I intended, like BLOOM (1970), to collect a corpus of utterances that would be largely representative of the child's pertinent stage of development. The large utterance corpora were labeled with the Roman numerals I, II, III, etc. With the aid of the intermediate recordings (I.I, 1.2, 1.3, II.1, II.2, etc.) I intended to reach at least some conclusions regarding the linguistic development of children within the period of six weeks. The recordings were made primarily in the children's dwellings, but also in playgrounds, parks, department stores, on the streets, and in the homes' of children who were their friends.

During the tape recordings of Meike and Kerstin, the mother of the child, and occasionally also the father, were present. During the recordings of Simone, my daughter, the mother was usually and the father was always present. Tape recordings were made at regular intervals of the children playing together with other children with whom they were acquainted. Everything that a child said during the sessions, or was said to her, was recorded on tape together with the observer's commentaries on the pertinent context (settings, changes in the communicative situation, actions and gestures of the child and her conversation partner). The technical apparatus used was as follows:

1. a "Uher Report Stereo 4400" (tape speed: initially 4.75 cm/s, later 9.5 cm/s);
2. a "Sennheiser Dynamic Studio Directional Microphone MD 421": this microphone, which was mounted on a movable stand (or was hand held outside of the house), was used to record the utterances of the child and her conversation partner.
3. a "Uher Microphone M5 36"; the contextual commentaries of the observers where whispered into this microphone. This second microphone was not used in recordings made outside of the house; in such cases the contextual commentaries were spoken into the Sennheiser microphone.

Bloom (1970) has pointed out that such technical apparatus, which always follows the children around during recordings, is regarded by children at the developmental stage studied here as a kind of physical extension of the observer. I can completely confirm this fact. And even the whispered contextual commentaries of the observers are scarcely registered by the children even if the observers occasionally take part in linguistic communication. This situation did not change until toward the end of the study, i.e., toward the end of the fourth year of the children's lives.

At the beginning of this investigation I made tape recordings part of the time together with, on each occasion, one of my student assistants. This had the advantage that I could concentrate primarily on linguistic interaction with the children, in order to examine with the aid of various elicitation techniques the boundaries of the children's linguistic and communicative abilities. The disadvantage of this method lay in the fact that, whereas I enjoyed a close and friendly relationship with the children (even when no recordings were being made), the presence of observers who were unfamiliar to the children almost inevitably caused them (especially Meike) to become quite reserved and shy.

The strongest criterion that I maintained for the recordings was that they should largely reflect the normal everyday life of the child. The tape recordings were transcribed by hand, together with certain notations. The transcribers were my student assistants, whom I had trained in a long and laborious process. Then I corrected the transcripts used in this monograph while listening to the recordings. Finally my assistants typed out the transcripts in a certain format. Instead, the transcription follows the rules of normal orthography.

Acknowledgements

The preparation of the corpus for CHILDES was supported first by the Max-Planck Institut für Psycholinguistik in Nijmegen and later by Jürgen Weissenborn at the University of Potsdam.