Contents
- 1 Form and Function
- 2 Temporality and Grounding in Narrative
- 3 Three Guiding Themes
- 3.1 Filtering
- 3.2 Packaging
- 3.3 Development
Our chapter title is deliberately ambiguous, reflecting the interwoven strands of our study. We intend, in the course of this book, to convey several senses of the words different and relate. The point of reference for these senses is a series of pictures that represent events â that is, dynamic interactions, over time, between animate beings, in physical settings. The pictures â 24 of them â were created by Mercer Mayer (1969) to form a story accessible to children. They are bound together in a storybook without words â except for the title, Frog, where are you?. This is our âfrog storyâ â the source of our data, which consist of elicited narratives from preschool, school-age, and adult narrators in five languages: English, German, Spanish, Hebrew, and Turkish. We suggest that, at the outset, you turn to Appendix I, where the pictures are reproduced, and âreadâ the story, since it is the point of departure for the entire course of this study.
The stories that make up our data are told in different ways because: (1) the narrators are of different ages; (2) the narrators speak different languages; and (3) there are different ways of talking about the same pictures. We seek to show that each of these factors â age, language, and choice of narrative perspective â systematically contributes to identifiable uses of linguistic means to relate events in the frog story narrative (and, by implication, in narrative generally).
When we speak about relating events, we have two sorts of issues in mind. Taking the sense of ârelateâ that means âtell,â we are concerned with the ways in which narrators put into words their conceptions of events â in this instance, events that are visibly displayed in pictures. That is, we are studying the development of the capacity to describe situations. We are also concerned with the ways in which individual events are related to each other. That is, we are studying the development of linguistic means to connect events and syntactically âpackageâ them into coherent structures â at the levels of scene, episode, and overall plot. âRelating eventsâ thus includes all of the verbal means for encoding and emplotment in narrative.
We did not begin our study â in 1983 â with all of these goals in mind. Rather, we were led to them by our data, which made it clear that the development of grammar cannot be profitably considered without attention to the psycholinguistic and communicative demands of the production of connected discourse. We began with a more limited goal: the study of the development of temporal expression in two quite different languages, English and Hebrew. We chose these two languages because one, English, has an elaborate set of verb markings for tense and aspect, while the other, Hebrew, does no more than mark the verb for the basic three tenses â past, present, future â with no grammaticization of aspect. We wondered whether Hebrew-speaking children would attempt to âcompensateâ for the sparse grammatical marking of temporality in their language by the use of lexical expressions for notions that are grammaticized in English. In so doing, we were guided by an approach to âcognitive prerequisites for the development of grammarâ (Slobin, 1973) according to which children seek linguistic means of expression for emerging concepts â in this case, supposedly universal concepts of temporality. This approach seemed well established in the child language literature. Roger Brown had defined the goals of study as âthe acquisition of knowledge, both grammatical and semanticâ (1973, p. 254). In investigating the initial acquisition of grammatical morphemes in English, he had found it possible to objectively define âobligatory contextsâ for the use of grammatical morphemes, including contexts for the temporal markers of progressive aspect and past tense â with the proviso that such contexts âshould not be dependent on the topic of conversation or the character of the interactionâ (p. 255). He had found that, once a child began to use a particular morpheme in 90% of such contexts for several consecutive longitudinal samples, it did not drop below 90%; and so âthe 90% criterionâ became established as the standard of acquisition. However, Brown dealt with the early emergence of grammatical morphemes. Later, âfunctionalistâ investigators were to find that âtopic of conversationâ and âcharacter of the interactionâ have their own developmental histories, and that uses of grammatical morphemes are, indeed, dependent on learning their contextually situated meanings (e.g. Budwig, 1989, 1991; Ervin-Tripp, 1989; Gerhardt (Gee), 1985, 1988, 1990). And we were to find that many of the morphemes and constructions that we were interested in did not have âobligatory contexts,â but, rather, represented expressive options that a speaker could take.
In the domain of temporality, the most extensive crosslinguistic surveys have been carried out by Richard Weist (Weist, 1986, 1990). Weist moved the terminology from âobligatory expressionâ to âconceptualization.â Various researchers had found that the inherent semantics (Aktionsart) of a verb played a role in determining its tense/aspect marking in early speech (e.g., Antinucci & Miller, 1976; Bloom, Lifter, & Hafitz, 1980; Bronckart & Sinclair, 1973). Brown had already noted that: âAppropriate uses of the past begin with a small set of verbs which name events of such brief duration that the event is almost certain to have ended before one can speak. These are: fell, dropped, slipped, crashed, brokeâ (1973, p. 334). Thus it is not clear if such past-tense forms express anteriority (tense) or completion (aspect), and it became commonplace to claim that it was more probable that very young children were marking aspect, rather than tense, with their early past-tense verb forms. Weist, however, found that Polish children distinguished both tense and aspect, since these two temporal notions are marked separately on the Slavic verb. He proposed that children are able to take two kinds of âperspectives on situationsâ â âexternalâ and âinternalâ (1986, p. 365): âWhen a situation is conceptualized from an external perspective, properties such as âcompleteâ, âpunctualâ, and âresultativeâ are salient, and when conceptualized from an internal perspective properties such as âongoingâ (âcontinuativeâ), âdurativeâ, and âincompleteâ are prominent.â
Putting together the functionalistsâ concern with communicative settings for the choice of linguistic forms, and Weistâs crosslinguistic concern with the sorts of conceptualizations involved in such choices, it is evident that the child has to learn much more than whether a particular âobjectively definableâ situation requires the use of a particular grammatical morpheme. The youngest children in the present study have passed their third birthday, and are well beyond the early stages studied by Brown. They already âknowâ many of the morphemes under study, but they do not yet know everything about the uses of these morphemes. And, in the following several years, they are learning not only new uses for the forms they know, but they are also acquiring new forms â in our study, both tense/aspect forms and interclausal connectives.
As we mentioned earlier, we were driven to this âfunctionalist-conceptualâ approach by the use of an elicited narrative task. We picked the âfrog storyâ (first used by Michael Bamberg in his 1985 Berkeley dissertation) because it is rich in opportunities for the encoding of temporal distinctions â sequence, simultaneity, prospection, retrospection, ongoing and completed events, etc. We found â at first to our surprise â that English and Hebrew narratives were quite different in their expression of temporality, even though narrators were, in some sense, telling âthe same story.â Bambergâs dissertation research had been carried out in Germany, and so it became fruitful to add German to our comparisons, leading to new surprises. Later, we were joined by two colleagues who completed the spectrum of languages represented in this book: Eugenia SebastiĂĄn, who gathered Spanish data in Madrid, and Ayhan Aksu-Koç, who gathered Turkish data in Istanbul. (The English stories were gathered in Berkeley by Tanya Renner and Virginia Marchman.) Our surprises, insights, and discoveries are documented in this volume.
1 Form and Function
The Leitmotif of our study is that form and function interact in development. Under âformâ we include a broad range of linguistic devices â from grammatical morphemes and bound inflections to interclausal connectives and syntactic constructions â along with lexical items encoding notions of temporality, manner, and causation. By âfunctionâ we understand the purposes served by these forms in narrative discourse â purposes of constructing a text that is cohesive and coherent at all levels: within the clause, between adjacent clauses, and hierarchically relating larger text segments to one another. The developmental history of any given form reflects the expanding range of functions served by that form, and, at the same time, also reflects the âacquisitional complexityâ of that form as determined by local processing constraints as well as the role played by the form in the overall system of the grammar. (By limiting ourselves to monologic narratives we exclude the considerable range of functions involved in the construction and maintenance of dialogue, including such issues as turn-taking, inter-speaker anaphora, question-answer pairs, and the like.)
Progressive aspect in English. As a mini-case-study, consider the development of progressive aspect in English. The present progressive is the first to be acquired of Brownâs 14 grammatical morphemes. It is used in its âprimitiveâ form, without a reliable auxiliary, throughout Brownâs five developmental stages. He states that it is âalmost always [used to name] an action or state in fact of temporary duration and true at the time of utteranceâ (1973, p. 318), and notes that: âThe children did not attain criterion on the full progressive until long after Stage V, and by then the children could speak of remote past and future times, could use modal auxiliaries as well as catenatives, and could move the temporal reference point into the pastâ (1973, p. 319). The 3-year-olds in our study have reached this point and are exploring additional functions of the progressive, in both present and past tenses.
One function of the progressive (and durative or imperfective aspects in general) is to present the continuing background against which a foregrounded event occurs. This function is evident in the following narrative segment from a child of 3;9, who temporally situates the frogâs escape during the period in which the boy was asleep:
- The frog got out, when heâs sleeping.[E3h-3;9]1
1 Excerpts from the data are followed by a code indicating language, age group, subject indicator, and exact age. Thus [E3h-3;9] indicates the eighth child in the sample of English-speaking 3-year-olds, aged 3 years and 9 months. Adult narrators are all arbitrarily given the age of 20. The languages are: E - English, G - German, H - Hebrew, S - Spanish, T - Turkish.
This usage differs in at least two important ways from the early use of the progressive noted by Brown: (1) It is not âtemporary durationâ that is at issue, but rather the temporal overlap between a durative and a punctual event. (2) The event of the boyâs sleeping is not âtrue at the time of utterance,â but rather true at some narrative reference time.
The separation of speech time and reference time is even clearer when the narrator uses the past progressive. The following segment, from a child of 3;5, shows an attempt to use the progressive to backtrack in time, describing a state â sitting on the edge â that existed before the boy fell into the water:2
- And then he fell over with the dog - into the pond. He was just sitting - on the - edge before - with his dog, and pow - into the water. [E3d-3;5]
2 Transcription conventions: hyphen indicates slight hesitation pause; three dots indicate longer pause; comma is non-final intonation; period is final fall. Words in capital letters indicate heavy stress, while boldface is used to highlight utterance segments that are relevant to the analysis. Some false starts may be omitted in examples for purposes of readability. Deleted clauses are indicated by three dots enclosed in square brackets. Linguistic examples and quotes from the texts are given in italics; glosses are enclosed in single quotes. (See Appendix II for more detail.)
Examples (1) and (2) indicate the backgrounding function of the progressive. Another way in which 3-year-old progressives differ from earlier uses is that they are not strictly tied to inherent verb semantics or Aktionsart. Recall that Brown found that punctual verbs like fall at first were used only in the past tense. But 3-year-olds can take either an âexternalâ or an âinternalâ perspective on such events. For example, Picture 17 shows the boy and dog in mid-fall, while Pict...