Content uploaded by Pamela Rosenthal Rollins
Author content
All content in this area was uploaded by Pamela Rosenthal Rollins on Mar 28, 2014
Content may be subject to copyright.
January 1994 AJSLP 45
ike a French braid, language development com-
prises multiple strands, including phonology,
semantics, syntax and morphology, and discourse,
all of which must intertwine for literacy to emerge
(McCabe, 1992). The relationship is such that delays in
development of any of the language strands may have
detrimental effects on learning to read. In fact, studies of
children with language disabilities have found impairments
in each strand to be linked to reading problems. Further-
more, prognosis for eventual language and literacy
functioning depends on the particular aspect of language
affected. Phonological problems, for example, are most
likely to have minimal lasting impact (Bishop &
Edmundson, 1987).
In contrast, narrative discourse development has strong
implications for emergent literacy in preschool children
(Dickinson & McCabe, 1991). The ability to tell a coherent
narrative predates and predicts successful adaptation to
school literacy (e.g., Feagans, 1982). Because early
intervention to prevent learning disability is predicated on
early assessment of language-related literacy problems,
narrative assessment holds promise in this regard. At
present, however, few systematic protocols tied to good
developmental research on narrative are available to
accomplish this assessment.
Information about preschool narration is relatively new
and has received much less attention than, for example, the
link between difficulties in phonological awareness and
emergent literacy. The purposes of this article are (a) to
Clinical Focus
Assessment of Preschool Narrative Skills
1058-0360/94/0301-0045$01.00/0 © American Speech-Language-Hearing Association
Allyssa McCabe
University of Massachusetts-Lowell, and
Harvard Graduate School of Education, Cambridge, MA
Pamela Rosenthal Rollins
Harvard Graduate School of Education, Cambridge, MA, and
Braintree Hospital Pediatric Center, Braintree, MA
The assessment of discourse skills in young
children is an important responsibility facing
clinicians today. Early identification of problems
in discourse skills and, more specifically, narra-
tive abilities is especially important for identify-
ing children at risk for later learning and liter-
acy-related difficulties. Despite this, few tools
are available for assessing narrative skills in
preschoolers. In this article we provide informa-
tion concerning preschool narrative develop-
ment in typically developing, North American,
Caucasian, English-speaking children. Methods
are suggested for assessing narrative skill of
children with language impairment and children
developing language normally. Transcripts of
narratives from these children are presented,
along with specific recommendations for
evaluating these narratives.
document the importance of oral narrative ability for
achieving full literacy; (b) to discuss normal narrative
development, presenting age norms; (c) to describe
assessment procedures that are mindful of cultural differ-
ences and appropriate for preschool and early elementary
school-age children; (d) to distinguish normal individual
differences from individual differences due to deficits; and
(e) to touch briefly on the therapeutic implications of our
assessment procedure.
Links Between Preschool Narration and
Literacy Acquisition
It now appears that it may be possible to identify
children at risk for language-based reading problems in the
early phase of language acquisition. Children who develop
reading disability in grade 2 have been found to be
deficient in the length, syntactic complexity, and pronun-
ciation accuracy of their spoken language at age 2:6
(Scarborough, 1990). Narrative was not assessed in these
children perhaps because no developmentally appropriate
narrative assessment device was available, a constraint we
seek to redress in this article.
Cross-sectional studies of older children who are
readers with dyslexia have identified a number of dis-
course-level deficits. Readers with disabilities have been
found deficient in tasks that tap oral skills, such as the
ability to recall highly familiar sequences of events
(Feagans & Short, 1984), directions (Feagans & Short,
L
on February 10, 2010 ajslp.asha.orgDownloaded from
46 January 1994 AJSLP
1986), and narratives (Graybeal, 1981; Weaver &
Dickinson, 1982). Deficits have been found in the ability to
construct narratives as well (Roth & Spekman, 1986).
Children with dyslexia seem to have problems—not
necessarily in constructing sentences, but rather in connect-
ing sentences using conjunctions. When asked to recall
stories, they tend to omit causal and temporal links (Liles,
1985; Roth & Spekman, 1986; Weaver & Dickinson, 1982)
and to provide fewer of the important details of stories that
are likely to be tested in schools (Graybeal, 1981; Griffith,
Ripich, & Dastoli, 1986; Hansen, 1978; Johnston, 1982;
Levi, Musatti, Piredda, & Sechi, 1984; Roth & Spekman,
1986; Weaver & Dickinson, 1982). In general, oral
narrative performance predicts literacy achievement
(Feagans, 1982; Michaels, 1981).
Discourse-level difficulties associated with reading
problems have been found at least as early as kindergarten
age. Children’s ability to tell a complete version of the
Three Bears predicts later reading success (de Hirsch,
Janksy, & Langford, 1966). Cross-sectional work also
finds a strong relationship between narrative comprehen-
sion among kindergartners and other measures of early
literacy, such as the ability to define words, phonemic
awareness, and early print skills (Dickinson & Snow,
1987). Deficiencies in discourse skills linked to reading
have been found prior to age 5 years. One longitudinal
study followed 87 children with language impairments and
children developing language normally from age 4 to 5:6
years and found that the ability to recall a short story was
the best predictor of language development (Bishop &
Edmundson, 1987).
Normal Discourse Development
The most prominent aspect of language acquisition
during early elementary-school years is the development of
extended discourse (Karmiloff-Smith, 1986). The onset of
this process is much earlier, beginning only a few months
after sentences are first formed. At about 22 months
(Eisenberg, 1985; Sachs, 1982), children begin to refer to
real past events, at first with much assistance from adults.
At 2 years, their narratives often concern negative past
events, especially injuries (Miller & Sperry, 1988). Even
when they produce fantasy stories, children aged 2–5 years
are preoccupied with themes of aggression, death, hurt, or
misfortune (Ames, 1966; Pitcher & Prelinger, 1963).
Between 3 and 5 years of age, children tell each other
longer and more complex personal narratives, and increas-
ingly respond to narratives from peers (Umiker-Sebeok,
1979).
Children tell each other many forms of narrative
(personal anecdotes, parodies, film retellings, fantasies),
but more than half of their conversational narratives
concern real personal experiences (Preece, 1987). McCabe
and her colleagues (McCabe & Peterson, 1991a; Peterson
& McCabe, 1983) have analyzed personal event narratives
of young children using high point analysis to describe the
developmental sequence of the narrative macrostructure
(see Table 1). At 3 1/2 years children generally combined
only two events even in their longest narratives, resulting
in what is called the Two-Event Narrative. By 4, children’s
narratives tend to consist of more than two events that
occurred on one occasion, but they narrate the events out of
sequence in what is called a Leap-Frog Narrative. Children
who use Leap-Frog Narratives also often omit some events
necessary for the listener to make sense of their personal
narratives. By age 5, however, children rarely have trouble
sequencing events in oral narratives. Five-year-olds do,
nonetheless, tend to end their personal narratives prema-
turely, dwelling on a climactic event at the end of their
narration in what is called an End-at-High-Point Narrative.
Six-year-olds tell a well-formed story that orients a listener
to who, what, and where something happened, narrates a
sequence of events that builds to some sort of climax or
high point, and then goes on to resolve itself by telling how
things turned out. That is, children age 6 and older use
what is called a Classic Narrative.
Although there is a developmental sequence of narrative
macrostructure from Two-Event to Leap-Frog to End-at-
High-Point to Classic Narrative structure, not all personal
event narratives fall along the developmental continuum.
On the contrary, Chronological Narratives are produced by
children and adults of all ages. These undeveloped stories
are relatively unevaluated laundry lists of actions, much
like the stereotypical adult slide show narrative. For
example, one 4-year-old girl responded to our question,
“Have you ever been to the zoo?” with the following
Chronological Narrative: “Yeah, before, I, and we saw
TABLE 1. Results of McCabe and Peterson’s high-point analysis for normally developing North American, Caucasian, English-
speaking children: The percentage of structural types at each age. Ten children were assessed at age 3 1/2 (McCabe, in press), and
16 children at each of the other age groups (Peterson & McCabe, 1983).
Age in Years Two-Event Leap-Frog End at High Point Classic Chronology‡ Misc.‡
3:6 63.3†10 3 320
41529
† 2 122318
510 429
†21 25 10
610 62335
†15 10
7 2 01748
†25 8
8 0 01762
†21 0
9 6 01758
†13 6
† Most common structure produced by children at each age.
‡ Narrative type found at all ages.
on February 10, 2010 ajslp.asha.orgDownloaded from
January 1994 AJSLP 47
gorillas, scary ones. And I was afraid. They had a swing
set up higher. But I don’t know if they swing on it. And I
saw a monkey too, and a lion.” The hallmark of this kind
of narrative is that children (and adults) do not select some
part of an overall event on which to focus and provide
details. Sometimes this kind of narrative seems to reflect a
relative lack of excitement about a topic.
Preschool children are more able to structure their oral
personal narratives in a sophisticated way than to structure
general scripts of personal experience or fictional stories
(Hudson & Shapiro, 1991), although this comparative
advantage of fact over fiction in terms of structural sophis-
tication does not apply to the story-writing of older school-
aged children (Freedman, 1987), perhaps as a result of
increased exposure to fictional stories from books. In spon-
taneously told fantasy stories, the plots of children between
the ages of 7 and 9 approximate those of fairy tales (Botvin
& Sutton-Smith, 1977; Hudson & Shapiro, 1991).
It should be emphasized that the developmental
sequence described here is that found in the personal event
narratives of young, North American, Caucasian, English-
speaking children. This type of narrative was chosen
because it shows the clearest developmental progression
during the preschool years. Variations expected for
children from diverse cultures are discussed later.
Current Assessments of Narrative Skill
Formally and informally, both in research and in clinical
practice, the construction of fictional stories is the primary
genre employed at present to assess the narrative skills of
children at risk for language problems. As we have noted,
however, such a narrative task is inappropriate for pre-
school children.
Formal Assessments
Unfortunately, existing formal tests of narrative skill do
not contain normative data for preschool-aged children, yet
it would be desirable to assess problems with narrative
discourse as an early indication of later reading problems.
At present, clinicians are unable to assess reading problems
until children start to read and experience failure. Children
aged 6:6 years and above might be assessed using either
the Story Construction subtest of the Detroit Test of
Learning Aptitude (DTLA-3; Hammill, 1991) or the Test
of Word Finding in Discourse (TWF-D), recently devel-
oped by Diane German (German & Simon, 1991). In the
DTLA-3 children are shown three pictures and asked to
make up a story about each. The story is then scored for its
conceptual, insightful, and coherent qualities. The TWF-D
asks for three stories about three pictures, followed by
probes to extend children’s stories, including one that asks
them to relate a personal experience about a carnival
(German & Simon, 1991). These stories are then scored
primarily for word-finding characteristics.
Informal Assessments
Informal clinical assessments of narrative skill have
often followed research paradigms. Regrettably, research-
ers have also used fictional stories to assess the narrative
skills of young children. Often researchers have scored
children’s productions or recollections for the extent to
which these display story grammar structure, meaning the
extent to which the stories open with setting information,
and proceed to tell how some problem was precipitated and
subsequently resolved. Ironically, researchers have
established the fact that school-aged children with learning
disabilities do not have problems with story grammar
(Graybeal, 1981; Griffith, Ripich, & Dastoli, 1986;
Hansen, 1978; Johnston, 1982; Jordan, Murdoch, &
Buttsworth, 1991; McConaughy, 1985; Merritt & Liles,
1989; Ripich & Griffith, 1988; Roth & Spekman, 1986;
Weaver & Dickinson, 1982). Stated differently, a number
of published studies have found no differences in the
narrative production of children with dyslexia, specific
language impairment, and traumatic brain injuries. All such
children seem cognitively competent in reporting problem-
solving episodes. Some researchers (e.g., Jordan, Murdock,
& Buttsworth, 1991) have attributed their failure to find
differences in the story grammar analysis of narratives
produced by children with language impairment to mean
that their narrative discourse skills are not impaired. This
conclusion may be erroneous, however, because of the use
of insensitive means of scoring narratives, as well as an
inappropriate genre.
In addition to problems posed by genre and scoring
systems, there may also be problems with elicitation
procedures. That is, story retelling was used as a means of
eliciting narratives in many of the studies above that found
no differences between children who were achieving
normally and children diagnosed as having specific
language impairment or learning disabilities (e.g., Gray-
beal, 1981; Griffith, Ripich, & Dastoli, 1986; Hansen,
1978; McConaughy, 1985; Ripich & Griffith, 1988; Strong
& Shaver, 1991; Weaver & Dickinson, 1982). When story
retelling is compared to story generation in response to a
picture by the same individuals, story retelling results in
longer, more detailed productions, and productions that
contain more frequent complete story grammar episodes
(Liles, Coelho, Duffy, & Zalagens, 1989; Merritt & Liles,
1989). Children with and without language disorders often
become confused and exhibit different word-finding
difficulties as they generate stories around picture stimuli,
so that reliable scoring is difficult. Thus, it is additionally
problematic that currently available formal means of
assessing narrative skill ask children to generate stories in
response to pictures.
Narrative Problems in Children With Specific
Language Impairments
Using a developmentally sensitive approach to analyze
narrative macrostructure (which is described below),
Miranda, McCabe, and Bliss (1993) found that 8- to 9-
year-old children with specific language impairment do
have substantial difficulty producing narratives. Although
boys without language impairments deliver chronologically
sequenced chains of actions, children with language
impairments deliver what might be termed Leap-Frog
on February 10, 2010 ajslp.asha.orgDownloaded from
48 January 1994 AJSLP
Narratives. That is, their narratives resemble those told by
normally developing 4-year-old boys: either the narratives
leap backward as well as forward in time, omit important
events, insert unrelated scripts or events, or all of these.
This means that listeners must do considerably more work
to comprehend the stories they tell. Children with specific
language impairments sometimes engage in a kind of
pseudo-development of their narrative topics by departing
from true narrative incidents to laundry lists of scripts
given in future tense or nominally in the past tense. For
example, during a narrative about taking the family dog to
the vet, one child continued, in the same turn, to discuss
what he had for lunch and then whom he kissed before he
went to bed. Other children with specific language impair-
ments generate happenings related to the discourse theme
but do so in an attempt to cover up their inability to say
more about a particular happening. Perhaps most confusing
of all are those instances when such children haphazardly
and for no explicit purpose narrate a seemingly unrelated
happening intrusively in the midst of another happening.
All these interminglings of seemingly unrelated events are
forms of Leap-Frog narration.
High-Point Analysis: A Developmentally
Sensitive Means of Assessing Narrative Skill
In order to assess the narrative skills of preschool
children, the appropriate genre of storytelling must be
elicited: personal narratives about real past events. This is
the genre that preschool children naturally compose quite
frequently in conversation with each other (e.g., Preece,
1987) as well as with parents (McCabe & Peterson,
1991b). Nonprofessional adults have provided evidence
that high-point analysis is valid; that is, adults’ judgments
of the quality of children’s personal narratives and their
memory of those narratives are both related to high-point
analysis (McCabe & Peterson, 1984, 1990a). Reliability of
speech-language pathologists scoring high-point analysis
using the method we will present is estimated to be kappa
= .90 for 18 narratives, which represents “almost perfect”
agreement (Landis & Koch, 1977). Thus, high-point
analysis is both valid and reliable. In the next section, we
discuss important aspects of eliciting personal narratives.
In addition, a scoring procedure using a developmentally
sensitive approach is described.
The Conversational Elicitation Procedure
The protocol for eliciting personal event narratives is as
important in the assessment process as the narrative itself.
The protocol suggested here is a conversational technique
developed by Peterson and McCabe (1983), called the
Conversational Map. Figure 1 is an example of a clinician
eliciting a personal event narrative from a child using the
conversational map. In addition to using this conversa-
tional map, clinicians must remember to: (a) use story
prompts, (b) try to collect at least three narratives, (c) use
relatively neutral subprompts in response to the child’s
answers, and (d) minimize the child’s self-consciousness.
The latter four points are crucial to narrative collection; a
discussion of each follows.
Use a story prompt. In spontaneous interactions you
have to tell a story to get a story. Almost everyone has
experienced awkward silences in social situations. No one
can think of a thing to say. However, the minute one
person launches into a tale about locking keys in his or her
car or leaving lights on in a parking lot, virtually all others
in the group share a similar incident that happened to them.
The exact content of a story prompt is not important per
se. What is important is that children are asked to talk
about experiences that mean something to them. In general,
children are likely to tell their best stories about being hurt
or scared. These are experiences that almost all children
have had but are significant enough to any particular child
to be worth talking about.
Collect at least three narratives from each child. No
matter what story prompt is used, no story appeals equally
to all children. Not every child has been stung memorably,
although most have. Hence, to increase the chances of
reminding the child of something he or she really wants to
talk about, it is essential that at least three brief anecdotal
story prompts be told.1
Unfortunately, adults routinely ask children to talk
about things that tend to produce poor narratives, even
from gifted narrators. One potentially unsuccessful probe is
to ask a child about a recently acquired sibling. Also,
professionals should avoid prompts about birthday parties,
which yield scripts that amalgamate many past parties, not
true narratives of specific, singular parties. Adults should
FIGURE 1. Conversation map for narratives of real experience.
1. Trip to doctor’s office PROPS: COUGH
I went to the doctor’s office the other day. I had to wait 3 hours
to see him. There were twin brothers about 5 years old waiting too.
They kept trying to read magazines. But every time one brother
picked out a magazine, the other brother wanted to read the same
magazine. They would start fighting, and their mother would take
the magazine away from them. They went through the whole pile of
magazines and didn’t get to read any of them.
- Do you have any brothers or sisters? (Follow child’s answer
with subprompts.)
- Do they fight or argue?
I’m still coughing from the cold I had (COUGH). When I finally
got in to see the doctor, he gave me some pink pills that were
about the size of a penny!
- Have you ever been to the doctor’s office or the hospital?
When I go home, I have to visit my aunt who is in the hospital.
She broke both of her legs and has to have them hooked up to
some wires from the ceiling.
- Have you ever visited anyone in the hospital?
- Have you ever gotten hurt?
1Note that you can’t give the same story prompt to the same child over
and over again. In fact, children will resist telling you the same story
twice. Evaluation over the course of treatment must be based on, for
example, whether high points begin to appear in whatever narrative the
child tells you rather than on whether the child inserts such a point into a
narrative he or she has told you before. Perhaps the best way to document
progress is to say that “Johnny will exhibit high point evaluation while
delivering a narrative in 7/10 novel narratives.”
on February 10, 2010 ajslp.asha.orgDownloaded from
January 1994 AJSLP 49
also avoid prompts about trips, which tend to yield
chronological accounts with little more coherence than a
travel itinerary.
Occasionally children tell about an experience they have
had when some relative, stranger, or pet died. These death
narratives are structured differently from other kinds of
narratives by children who are developing normally.
Specifically, virtually no evaluation is used to describe
deaths, and this lack is more pronounced the younger the
child and the closer the child was to the person who died.
Furthermore, even older children, who are otherwise fully
capable of producing well-formed narratives, are likely to
produce confusing, jumbled sequences in narrating events
that led up to a death of a person or pet who was dear to
them (Menig-Peterson & McCabe, 1977). Thus, narratives
of deaths should not be used for clinical assessment of
language problems.
Use relatively neutral subprompts in response to the
child’s answers. Narratives are constructed between two
people; what is said in part determines what will be heard.
Children are accustomed to telling stories to parents who
respond to them freely. As professionals, when we elicit
narratives from children, we need to respond too. One of
the most important reasons for responding is that no
response is a response. Children who meet with no
response are likely to say very little. Some responses are
less leading than others. That is, they encourage narration
while not directing the narrative. Specifically, clinicians
are encouraged to (a) repeat the exact words of the children
when they pause, (b) say “Uh-huh,” (c) say “Tell me
more,” or (d) ask “Then what happened?” Clinicians may
notice that these responses are relatively neutral, really just
indications of interest in hearing whatever the child wants
to say. This is quite deliberate on our part. Just as it is
important to avoid too little response, it is also important to
avoid saying too much, such as evaluating children’s
stories. The more adults do, the less chance there is to see
what children can do on their own. Unfortunately, it is
particularly difficult to restrain commentary when dealing
with children who are having a hard time conveying their
story. These are the very children who are most likely to
have problems and need careful assessment. With practice,
however, it becomes more natural to avoid overt commen-
tary.
Minimize the child’s self-consciousness. In order to
collect narratives worth analyzing, it is critical that the
focus be removed from the language and narrative ex-
change per se. A successful technique is to have the child
draw a picture (Peterson & McCabe, 1983). This allows
the clinician to become acquainted with the child and
comes quite naturally at the beginning of an evaluation
session.
Alternatively, during a hectic diagnostic session,
conversational narratives could be collected during a break
in more formal testing. Wherever placed, children say
more if they do not see the interaction as a test.
Don’t rush the child. It is important not to rush through
the collection of personal event narratives. Even though
they may be collected during breaks between more formal
tests, narratives are an important part of the assessment
procedure and care needs to be taken to allow the child
time to communicate his or her message. These narratives
from children are not test scores taken from a standardized
battery. Instead, the stories are meaningful expressions of
who the child is and what he or she has experienced. For
many reasons, it is important to take time in listening to
children’s narratives.
Scoring
Select the longest narrative. McCabe & Peterson
(1990b) found that the length of a narrative was a rough
indicator of its complexity. Therefore, in order to obtain a
fair estimate of the upper bounds of a child’s performance,
the longest narrative should be used for scoring. Occasion-
ally the best narrative will not be the longest one; conse-
quently, clinicians may want to score all narratives elicited
and give the child credit for the best narrative structure
produced. Length of a narrative should be determined by
counting the number of lines in a transcription of it. If two
or more narratives are close in terms of the number of
lines, choose the one that has the most uninterrupted text
and/or makes the most sense when you read it. Individuals
may disagree about exactly where a narrative begins and
ends, but this disagreement will not affect their assessment
of overall structure. Finally, the narrator should have been
present during the events narrated; occasionally, children
will tell a story about something that happened when they
were not present. These should not be used for assessment.
FIGURE 2. Questions for scoring narrative structure. The
North-American-Caucasian-English-speaking model.
One-Event
Narrative
Two-Event
Narrative
Miscellaneous
Narrative
Leap-Frog
Narrative
Chronological
Narrative
End-at-High-Point
Narrative
Classic
Narrative
No
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
No
No
Are there two past tense events?
Are there more than two past tense
events?
In the real world is there a logical or
causal sequence to these events?
Is there a High Point?
(event(s) that occur after a high point and
wind up crisis)
Is there a resolution?
Does the narrator's order of the events
mirror the sequence in which the events
must have logically occurred?
(concentration of evaluative comments)
Does the narrator's order of the events
mirror the sequence in which the
events must have logically occurred?
▼
▼
▼
▼
▼
▼
▼
▼
▼
▼
▼
▼
(concentration of evaluative
comments)
(event[s] that occur after a high
point and wind up crisis)
on February 10, 2010 ajslp.asha.orgDownloaded from
50 January 1994 AJSLP
It should also be noted that identifying the beginning
and ending of narratives will be more challenging for
disordered narratives than for well-formed ones. In fact, the
struggle to identify the boundaries of a narrative is one
potential indicator of disordered narration.
Identify the narrative macrostructure. Figure 2 depicts a
series of questions to be used for scoring narrative struc-
ture. You may notice that yes answers allow you to proceed
to the next question; no answers result in the determination
of the type of narrative structure displayed. For example,
compare the following two narratives: 2,3
Narrative One (8-year-old girl)
E: Have you ever gotten jabbed by anything?
B: By a bee.
E: By a bee. Oh, tell me about it.
B: It got kind of cool one day and my grandma came.
She called me and she wanted to know where Dennis was.
E: Where Dennis was?
B: Yeah, and I ran outside to tell her and I was running
and I stepped on a bee.
E: You went outside to tell her and you were running
and you stepped on a bee. Ah. Then what?
B: Nothing. I just went in the house and had to have
something on it.
Narrative Two (5-year-old girl)
E: Have you ever gotten jabbed with anything?
L: Uh huh. I got jabbed with a bee.
E: By a bee. Oh, tell me about it.
L: See, I got jabbed on my foot. I was barefooted. I
screamed and I screamed and I cried and I cried. I
screamed and I screamed. Until my next door neighbor
came out and my Dad came out and my brother came out.
And, they all carried me into the house but after that
happened I got to sleep overnight with my neighbor.
Both narratives have more than two past events or
action words. Both of these narratives are about being
stung by a bee, an event that has a temporal sequence in the
real world. Both narratives unfold in such a manner that the
events mimic the order in which they happened. Narrative
One, however, does not have a high point. There is no
emotional heart of the story. No evaluation takes place.
Narrative One is a Chronology, a listing of unevaluated
events. In contrast, Narrative Two does have a high point.
The child uses repetition, evaluating her response to the
bee sting, “I screamed and I screamed. I cried and I cried. I
screamed and I screamed.” In addition, Narrative Two has
a resolution. The child indicates that everything turned out
all right; in fact, “I got to sleep overnight with my neigh-
bor.” In scoring Narrative Two, we answered yes to all of
the questions, which indicated a Classic Narrative struc-
ture. It must be stressed, however, that although resolutions
bring narratives to a formal closure, the content of resolu-
tions may involve no solution to any problem (Peterson &
McCabe, 1983) and may be quite negative. In Narrative
Two above, for example, the resolution has nothing to do
with solving the problem of the bee sting. Another resolu-
tion to a bee sting story might be a massive allergic
reaction and hospitalization, which are highly negative
events. In short, resolutions are formally defined as actions
that occur after the high point event.
Identify the evaluated high point. Identifying the narra-
tive structure is relatively straightforward except for identi-
fying the evaluated high point. Evaluation doesn’t disclose
information about what happened in the past. Rather,
evaluation reveals the meaning a happening had for the
narrator. There are many kinds of evaluation. Some are
obvious (e.g., “That was good”); others, such as negative
events, are quite subtle. As Labov (1972) pointed out, an
infinite set of things did not happen on any occasion. When
a narrator uses negation to tell what did not happen (e.g.,
“My dad didn’t keep his eyes on the road”), the narrator is
identifying deviations from what people expected would
happen, which is evaluative rather than informative per se.
Figure 3 displays a list of evaluation types found in the
narratives of 4- to 9-year-old North American, Caucasian,
English-speaking children, along with examples.
Children normally evaluate their experiences. Children
begin to use evaluation as young as 2 years (Miller &
Sperry, 1988), and it becomes increasingly more frequent
with age. In fact, Peterson & McCabe (1983) found that
children aged 4–9 years who are developing normally
evaluate half of all comments in some way. Although
evaluation is prevalent in many narratives of young
children, not all evaluation constitutes an evaluated high
point. What distinguishes an evaluated high point from
other types of evaluation is that the evaluation at the
narrative’s high point is concentrated in some way. For
example, the evaluated high point or emotional heart of
Narrative Three is italicized:
Narrative Three (8-year-old girl)
K: I really fight Pete. Well whenever we get home from
school, he starts an argument. He says, “You got to do that.
You got to do that.” Which is, when it’s really his turn. I
get so mad I punch him in the stomach about that hard and
he goes screaming, “Mommy, Kim hit me.” I mean it. But
when he gets in trouble and I have arguments with Pete,
especially one. Well, it was the day he went into my room.
I was at school still, this was when I was in kindergarten
and he went in my room and tore all my pictures down that
I painted and he tore them up. And he broke one of my
best, my very best doll, my Raggedy Ann, she was my
favorite. I got another one. I love Raggedy Ann dolls. Then
I told my mother and she came and he got it. Then we
started arguing all over again. Boy once Pete talks, he can’t
stop talking.
Here, the narrator leaves no room for doubt about the
heart of this experience for her, although there is some
evaluation earlier in the narrative (e.g., stomach is stressed,
“I mean it,” etc.).
2 For all narratives presented, the initial E indicates the examiner.
3 It was necessary for pedagogy to include narratives of children older
than preschool age in the body of the text. Additional narratives of
preschool children may be found in the Appendix.
on February 10, 2010 ajslp.asha.orgDownloaded from
January 1994 AJSLP 51
Identify age appropriateness of narrative structure.
Table 1 may be used to identify whether a child is using
narrative structures that are typical of his or her age group.
For example, Narrative Two, a Classic Narrative, was
produced by a 5-year-old child. From Table 1, we see that
although the dominant structure for children aged 5 is end-
at-high-point, 21% of the 5-year-olds studied produced
more advanced Classic Narratives. The child who produced
Narrative Two is functioning well within normal limits on
narrative structure. On the other hand, if a 5-year-old boy
produced only Two-Event Narratives at best, he would fall
into the lowest 10% of normal children. This would be
cause for concern about his narrative skills. In short, if
children produce a narrative that falls into the most
common category for their age (or more advanced catego-
ries), all is well. If, however, children consistently produce
narratives with less advanced structures, clinicians might
advise monitoring. In particular, assessment of a Leap-Frog
Narrative in a child over the age of 6 is quite a serious
concern because none of the normal children interviewed
produced narratives that fell into this category (Peterson &
McCabe, 1983); in contrast, all of the children with a
diagnosis of specific language impairment produced
narratives that would fall into this category (Miranda,
McCabe, & Bliss, 1993).
The 8-year-old child who produced Narrative One, a
Chronology, has produced a structure that is normal for
individuals of all ages. Hence, to determine whether she is
functioning within normal limits, we would need to
analyze a different narrative. In other words, if the narra-
tive selected for scoring is a Chronology, a second or third
narrative should be selected in order to find a narrative
structure that falls along the developmental continuum.
Issues That Complicate the Identification of
Narrative Macrostructure
Sometimes children provide the main event of a story to
see whether the listeners are interested in hearing more. If
listeners indicate their willingness to do so, narrators back
up, tell the events that led up to the high-point event, and
resolve it. Children should not be penalized for such
normal departures from timeline sequencing. For example,
one 7-year-old boy began his narrative, “I got bit by a
crab.” He then waited for a response, indicative of interest
in hearing more, and eventually backed up to begin with,
“We went down to Wisconsin.”
Problems with pronouns and word-finding make
narratives generally less comprehensible. Clinicians might
notice that word-finding problems are prominent in the
narratives they collect, but this is a different issue, requir-
ing a different type of analysis. Clinicians should separate
word-finding issues from issues of narrative construction,
even though the flow of narratives is affected by word-
finding problems.
Children with word-finding problems may sound to
naive listeners as if they have problems with narrative
construction; the job of speech-language pathologists is to
distinguish between general discourse problems, such as
word-finding deficits, and problems specific to narrative
construction.
Also note that the assessment of the structural maturity
of narratives among children should be distinguished from
assessment of the truth value of what they say. Children
developing normally will exaggerate or lie from time to
time. One 5-year-old asked the first author, “Do you know
what? Every single tree fell down on our house because
Evaluation Type Example
Onomatopoeia “It went Bam.”
Stress “I screamed and I
screamed
” with heightened tone of voice
Elongation “We had to stay a loong time.”
Exclamation “Oh boy!”
Repetition “I screamed and I screamed and...I screamed and I screamed”
Compulsion words “We had to come in then.”
Similes and metaphors “His eyes got as big as tomatoes.”
Gratuitous terms “very,” “really,” “just”
Attention getters “I got to tell you the important part.”
Words per se
finally, accidentally, squished, scared
Exaggeration and fantasy “I picked them [trees] up with my pinky.”
Negatives “He didn’t shot me or nothin’.”
Intentions, purposes, desires, or hopes “I hoped Santa would bring me a new one.”
Hypotheses, guesses, inferences, predictions “We didn’t think it would rain.”
Results of high-point action ”I cut myself with the knife. Blood came running out.”
Objective judgments “My brother liked my snowman much better than he liked my sister’s.”
Subjective judgments “That was my favorite.”
Facts per se “I caught the biggest fish.”
Internal emotional states “She didn’t care about me.”
Tangential information that is evaluative “She gave me ten dollars for going in there.
Ten dollars is a lot of
of main narrative information
money when you’re little.”
Causal explanations “He hit me in the head with a rock, so I threw one at him.”
FIGURE 3. Evaluation types found in 4- to 9-year-old white English-speaking children.
on February 10, 2010 ajslp.asha.orgDownloaded from
52 January 1994 AJSLP
there was a snowstorm. I picked them up with one, with
my pinky. All of them with my pinky. Do you believe
that?” If children tell such coherent, albeit untrue, narra-
tives, they do not have narrative language problems. They
may have other problems, but those are outside the
purview of speech-language pathologists.
As professionals become accustomed to this scoring
procedure, they recognize structures as children talk to
them, at least well enough to determine whether narratives
heard are chronologically or miscellaneously structured (it
will be obvious to everyone if the narrative concerns
death). In such cases, the clinician should make doubly
sure to collect more than three narratives.
Individual Differences Versus Deficits
Several sources of individual variation must be consid-
ered while evaluating narrative skill. First, there are
individual differences within the normal population. Some
variation in performance is normal. Children developing
normally may produce a Chronological Narrative followed
by a Classic one at any given time. Moreover, the same
topic may elicit a Leap-Frog Narrative from one 4-year-old
but a Classic Narrative from another 4-year-old. Some
normal variation can be expected among age peers in
optimal performance. For example, compare the Classic
Narrative Two above produced by one 5-year-old, and the
more typical End-At-High-Point Narrative Four produced
by another 5-year-old below. This type of variation is
common and should be expected in narrative production.
Narrative Four (5-year-old girl)
E: Did you ever go to the doctor’s office?
D: Uh-uh. No, yes, over Dr. Graham’s house, night.
E: You went there? What happened?
D: Nothing. Just I sticked around and he told me to
come in first and then he, and, that’s all I had to do. And
taked me out, out, and and he put me in the doctor office.
And I had a cold.
E: You did?
D: Last night.
E: Right.
D: And I, I was scared to come in. And he didn’t shot
me or nothing.
E: He didn’t shot you or anything?
D: Uh-uh. He didn’t even shot me.
E: He didn’t shot you?
D: He gave me them, them tiny pills too, just like you.
That’s only reason I had.
Individual differences also may vary according to
culture. Most of the research reviewed in this paper has
been done with North American, Caucasian, English-
speaking children. It is critical to realize that substantial
cultural differences influence the ways children structure
their narratives. Such cultural differences must be carefully
distinguished from individual deficits. For example,
African American children often tell what has been called a
Topic-Associating Narrative (Michaels, 1981). Such
narratives thematically combine narration of events that
happened at different times and places into one narrative
(Rodino, Gimbert, Perez, Craddock-Willis, & McCabe,
1991), as in Narrative Five.
Narrative Five (7-year-old African American girl)
E: Have you ever been in a car accident?
C: Yes when I was with my aunt and my mother. And
my mother was driving the car but there was a truck in the
way and she was trying to move over and pass him, but the
truck was too big. And when she, and she moved over. And
when she was driving, she moved back the other way. And
the mirror on the outside of the door—it bumped into the
side of the car—not the mirror. But on the side of the car it
bented. My father got mad at her because it wasn’t her car.
It was my father’s. And he, when my, we got home, my
mother said, “Go tell your father it’s time to eat.” And I
told my daddy. And he said leave him alone. And he didn’t
come to eat until we were sleeping. But he didn’t. He did
eat, but while we were asleep. But he was mad. So he
moved out. Cause my mother bent the car, but only on the
side.
E: Okay. Well that was very good.
C: And one day somebody threw a rock and hit my
daddy’s, my father’s car. And the mirror—it broke off.
And me and my cousin saw it, and we were mad too. And
after that he [father] moved out.
In a manner reminiscent of haiku, Japanese children tell
extraordinarily succinct collections of experiences, often
given in sets of three lines, rather than narrating the details
of what happened on one occasion in the manner that
North American, Caucasian, English-speaking children do
(Minami & McCabe, 1991), as in the following example:
Narrative Six (8-year-old Japanese boy)
A: As for the first, you know, got at Ehime, you know,
hurt a lot. As for the second, you know, knew, you know,
hurt, you know. Well, you know, didn’t hurt so much, you
know. The next was the same again. As for the very last,
you know, didn’t hurt at all.
In the present framework, the above narrative would be
classified as a One-Event Narrative. It would be found
atypical of productions by children of this age. However,
such a classification would be inappropriate because this
narrative was told by a Japanese child. Its structure is quite
different from that of North American, Caucasian, English-
speaking children (see Minami & McCabe, 1991, for more
details on Japanese children’s narrative structure).
Children from Latino cultures rarely focus on sequenc-
ing events in their narratives, foregrounding instead
narration of family connections and relationships (Rodino
et al., 1991). In fact, almost 50% of the narratives produced
by children from Latino cultures contained no sequencing
of events (Rodino et al., 1991). In story retellings, Hungar-
ian children extensively embellish their recapitulations,
unlike American children (John-Steiner & Panofsky,
1992). Hawaiian children tell talk-stories that weave
teasing and fantasy into repetitive routines for a number of
participants (Watson, 1975). Narratives from different
cultures contrast with narratives produced by children with
on February 10, 2010 ajslp.asha.orgDownloaded from
January 1994 AJSLP 53
specific language impairments, yet without careful
attention the former could be mistaken for Leap-Frog
Narratives, which would be developmentally inappropriate
narration for their age group. Such misdiagnoses must be
avoided.
Many important educational implications might be
noted regarding cultural differences in narrative structure.
However, they are quite complex and would require
explanation far beyond the scope of this article (for more
information see McCabe, in press).
Finally, individual differences vary according to deficit.
As exemplified by Narrative Seven, some departures from
normal developmental sequence reflect important deficits
at the level of discourse. In Narrative Seven, the narrator
jumps from talk of asthma attacks, to getting his tonsils
out, to breaking his knee, to his friend’s accident, without
developing any of these or relating them to each other.
Thus, this narrative is a Leap-Frog structure, very atypical
of productions of North American, Caucasian, English-
speaking children who are developing normally. In fact,
this child was diagnosed as having specific language
impairment.
Narrative Seven4 (9-year-old boy)
E: Last week, I had a sore throat. I went to the doctor
and I had to get a shot. Have you ever gotten a shot at the
doctors?
J: Yeah.
E: Tell me about it.
J: I was losing my voice, I was having asthma attack
real bad. So, my friend go it and he got to me while I was
coughing in the middle of the night, and he got a shot right
on my leg and I had to take my tonsils out, I didn’t like it.
And...
E: You had to get your tonsils out?
J: That’s what the doctor said.
E: And then what happened?
J: I went into the hospital for a week. And...because a I
had a real bad asthma thing, and they just put me in the
hospital for a week. And and I and I broke my knee.
E: You broke your knee?
J: Yeah. While I was ridin my bike, I was jumpin on a
curb and and I was it was I was slidin through the air, and
it looks like I broke my knee. I couldn’t move nothin.
E: What happened?
J: My mom took me to the hospital and said um, said
uh...the doctor said we he might be, we might be, we’re
going to...he has to do, can’t ride his bike in the street, I
can’t ride my bike in the street anymore. Cause uh I get
hurt. My friend David had a car accident.
E: He did?
J: Yeah, and he um he’s in the hospital right now
be...and (pause) he stopped and said I forgot. He said uh,
they said uh, he can’t he couldn’t he can’t, they put a metal
thing on his head. He will not talk. He’s deaf now.
E: He’s deaf?
J: Uh-uh. And they, they’re gonna take it off soon. So
they did, not he can talk now. And doin OK now, they took
it off already. Uh...my friend, my mom’s friend works in a
hospital and in somewhere in Roseville. I forgot what’s her
name (pause). That’s all I can think of.
Leap-Frog Narratives have one or more of the following
features: (a) events on a specific occasion are given out of
sequence (e.g., J says “they’re gonna take it off soon...they
took it off already.”); (b) important information appears to
be omitted (e.g., J omits a connection between his asthma
attack and his tonsillectomy); (c) scripts are inserted into a
narrative about past events; (d) instead of giving more
information about one specific occasion, the narrator may
leap into generating narratives of a similar occasion (e.g., J
talks about his friend’s accident, his own accident, and
other hospitalizations); and (e) unrelated occasions are
narrated intrusively in the midst of another occasion (e.g.,
one child with specific language impairment began a story
about a birthday party, leaped to a story about his aunt’s
accident and back to the birthday party). This is the kind of
variation, mindful of age, culture, and normal variation,
that should be registered to indicate a child at risk for
reading. A considerable body of literature reviewed above
links problems with narrative skill to reading comprehen-
sion difficulties. If a 3-year-old has not yet referred to past
events, there may be cause for concern. If a 4-year-old
never chains two events together in any kind of narrative
sequence, there may be cause for concern. If a 6-year-old is
still struggling with sequencing events, or not chaining two
events together in most instances, there may be cause for
concern.
As speech-language pathologists, it is critical to
separate out issues that explain narrative differences
exhibited by children. Is individual fluctuation characteris-
tic of normal performance? Is it individual deficit? Or is it,
instead, a matter of cultural differences? To assess narra-
tives from children coming from different cultures,
professionals must appreciate how stories are valued in
these cultures. If a 7-year-old North American, Caucasian,
English-speaking child produces Leap-Frog or Single-
Event Narratives at best, however, an assessment of
narrative deficit is definitely warranted.
Treatment Possibilities/Implications
Given that a narrative deficit is diagnosed, intervention
efforts for narrative deficits should take place throughout
the child’s day. Past efforts to improve narrative skills by
talking one-on-one with a friendly adult for several hours a
week over the course of several months have not been
successful (e.g., McCabe & Peterson, 1989). Hence,
intervention should focus on consultation with teachers and
staff within the school environment and consultation with
the child’s parents (or primary caretaker). Furthermore,
intervention with parents and child together is integral to
the language treatment of preschool children with narra-
tive-level difficulties. The reasons for this are twofold:
4 Narratives # 1, 2, 3, and 7 are from the Peterson and McCabe (1983) cor-
pus. Narrative # 4 is from the Rodino et al. (1991) corpus. Narrative # 5 is
from the Minami and McCabe (1991) corpus. Narrative # 6 is from the
Miranda, McCabe, and Bliss (1993) corpus.
on February 10, 2010 ajslp.asha.orgDownloaded from
54 January 1994 AJSLP
(a) preschool children spend most of their waking hours
with their parents (or primary caretakers), which makes
such adults the logical choice to facilitate narrative skills
through the child’s day; and (b) some parental styles of
talking about the past predict more optimal narrative
development than other styles.
In a longitudinal study, McCabe and Peterson (1990b)
studied 10 children talking about past events at home with
their parents. The study began when the children were 2:0
and continued until they were 6:0. Some parents did not
take “Nothing” for an answer to questions like, “What did
you do in school today?” They talked at length about a
variety of topics with their children, especially ones that
seemed to interest the children (e.g., “Who had an accident
in nursery school?”). Other parents responded to their 2-
year-olds’ inept narration by changing the subject fre-
quently, never dwelling on any one topic about the past.
The first group of parents—those who habitually extended
topics concerning past events, using a variety of techniques
for doing so—were the ones whose children were the best
narrators as measured by high-point analysis 4 years later.
Parents who switched topics or habitually corrected their
children (e.g., “That’s not the way it happened”) had
children who eventually refused to tell narratives to their
parents (e.g., “You tell the story, Mommy.”)
These findings suggest that many parents may need
training concerning optimal language facilitation tech-
niques. It is not enough for parents to be told to elicit
narratives from their children; they may need to be
educated and supervised in facilitation techniques. For
example, parents may be told to emphasize discussion of
things that happened during times when they were sepa-
rated from their children, because children see this as a real
communication exchange rather than the kind of test or
school assignment they do poorly on (e.g., “Tell me what
we did when our class went to the science museum.”).
From the age of 31 months, children are more likely to
respond to prompts about events that occurred when they
were not with their parents than when they were (McCabe
& Peterson, 1991b). Parents should also be encouraged to
interweave narrative exchange among other language
games (see McCabe, 1992, for various games prerequisite
for narrative exchange presented in a format intended for
use by parents).
Conclusion
A number of researchers have found that preschool
children who are unable to tell a personal narrative as well
as their peers may be at risk for difficulty acquiring
literacy. We have proposed a method of eliciting and
scoring personal narratives from preschool children and
presented some normative data that would enable early
detection of problems at the narrative level of language.
The authors have attempted to carefully distinguish
between the kind of jumbled and incomplete event se-
quences told by children after the age of 4 years who have
problems with language in general and narrative construc-
tion in particular, and the kind of alternative storytelling
styles of children from other cultures that may value
structures other than the kind of sequencing found in North
American, Caucasian, English-speaking narratives. We
invite readers to try our method of narrative analysis on the
additional narratives of young preschool children included
in the Appendix.
Acknowledgments
Portions of this article were presented at a miniseminar during
the annual Convention of the American Speech-Language-
Hearing Association in Atlanta in November 1991. We gratefully
acknowledge Lydia Greene and Laurie Bozzi for performing
reliability estimates and for their very helpful suggestions. We
would also like to acknowledge Nickola Nelson and Marilyn
Newhoff for their helpful editing.
References
Ames, L. B. (1966). Children’s stories. Genetic Psychology
Monographs, 73, 337–396.
Bishop, D. V. M., & Edmundson, A. (1987). Language-
impaired 4-year-olds: Distinguishing transient from persistent
impairment. Journal of Speech and Hearing Disorders, 52,
156–173.
Botvin, G. N., & Sutton-Smith, B. (1977). The development of
structural complexity in children’s fantasy narratives.
Developmental Psychology, 13, 377–388.
de Hirsch, K., Jansky, J. J., & Langford, W. S. (1966).
Predicting reading failure. New York: Harper and Row.
Dickinson, D. K., & McCabe, A. (1991). A social interactionist
account of language and literacy development. In J. Kava-
naugh (Ed.), The language continuum (pp.1–40). Parkton,
MD: York Press.
Dickinson, D. K., & Snow, C. E. (1987). Interrelationships
among prereading and oral language skills in kindergartners
from two social classes. Early Childhood Research Quarterly,
1, 25.
Eisenberg, A. R. (1985). Learning to describe past experiences in
conversation. Discourse Processes, 8, 177–204.
Feagans, L. (1982). The development and importance of
narratives for school adaptation. In L. Feagans & D. Farran
(Eds.), The language of children reared in poverty (pp. 95–
116). New York: Academic Press.
Feagans, L., & Short, E. J. (1984). Developmental differences
in the comprehension and production of narratives by reading
disabled and normally achieving children. Child Development,
55, 1727-1736.
Feagans, L., & Short, E. J. (1986). Referential communication
and reading performance in learning disabled children over a
three-year period. Developmental Psychology, 22, 177–283.
Freedman, A. (1987). Development in story writing. Applied
Psycholinguistics, 8, 153–170.
German, D. J., & Simon, E. (1991). Analysis of children’s
word-finding skills in discourse. Journal of Speech and
Hearing Research, 34, 309–316.
Graybeal, C. M. (1981). Memory for stories in language-
impaired children. Applied Psycholinguistics, 2, 269–283.
Griffith, P. L., Ripich, D. N., & Dastoli, S. L. (1986). Story
structure, cohesion, and propositions in story recalls by
learning disabled and non-disabled children. Journal of
Psycholinguistic Research, 15(6), 539–549.
Hammill, D. D. (1991). Detroit Test of Learning Aptitude.
Austin, TX: ProEd.
Hansen, C. L. (1978). Story retelling used with average and
learning disabled readers as a measure of reading comprehen-
sion. Learning Disability Quarterly, 1, 62–69.
on February 10, 2010 ajslp.asha.orgDownloaded from
January 1994 AJSLP 55
Hudson, J. A., & Shapiro, L. A. (1991). From knowing to
telling: The development of children’s scripts, stories, and
personal narratives. In C. Peterson & A. McCabe (Eds.),
Developing narrative structure (pp. 89–136). Hillsdale, NJ:
Erlbaum.
John-Steiner, V., & Panofsky, C. (1992). Narrative competence:
Cross-cultural comparisons. Journal of Narrative and Life
History, 2(3), 219–234.
Johnston, J. R. (1982). Narratives: A new look at communica-
tion problems in older language-disordered children. Lan-
guage, Speech, and Hearing Services in Schools, 13, 144–155.
Jordan, F. M., Murdoch, B. E., & Buttsworth, D. L. (1991).
Closed-head-injured children’s performance on narrative
tasks. Journal of Speech and Hearing Research, 34, 572–582.
Karmiloff-Smith, A. (1986). Some fundamental aspects of
language development after age 5. In P. Fletcher & M.
Garman (Eds.), Language acquisition (2nd ed., pp. 455–474).
Cambridge, MA: Cambridge University Press.
Landis, J. R., & Koch, G. G. (1977). Measurement of observed
agreement for categorical data. Biometrics, 33, 159–174.
Labov, W. (1972). Language in the inner city. Philadelphia:
University of Pennsylvania Press.
Levi, G., Musatti, L., Piredda, L., & Sechi, E. (1984). Cogni-
tive and linguistic strategies in children with reading disabili-
ties in an oral storytelling test. Journal of Learning Disabili-
ties, 17(7), 406–410.
Liles, B. Z. (1985). Cohesion in the narratives of normal and
language-disordered children. Journal of Speech and Hearing
Research, 28, 123–133.
Liles, B. Z., Coelho, C. A., Duffy, R. J., & Zalagens, M. R.
(1989). Effects of elicitation procedures on the narratives of
normal and closed head-injured adults. Journal of Speech and
Hearing Disorders, 54, 356–366.
McCabe, A. (1992). Language games to play with your child
(rev. ed.). New York: Insight/Plenum.
McCabe, A. (in press). Chameleon readers: Some problems that
cultural differences in narrative structure pose for
multicultural literacy programs. New York: McGraw-Hill.
McCabe, A., & Peterson, C. (1984). What makes a good story?
Journal of Psycholinguistic Research, 13, 457–480.
McCabe, A., & Peterson, C. (1989, March). Strategies for
developing narrative structure in a preschool setting. Paper
presented at the annual meeting of the American Educational
Research Association, San Francisco.
McCabe, A., & Peterson, C. (1990a). What makes a narrative
memorable? Applied Psycholinguistics, 8, 73–82.
McCabe, A., & Peterson, C. (1990b). Keep them talking:
Parental styles of interviewing and subsequent child narrative
skill. Paper presented at the Fifth International Congress for
the Study of Child Language, Budapest, Hungary.
McCabe, A., & Peterson, C. (Eds.) (1991a). Developing
narrative structure. Hillsdale, NJ: Erlbaum.
McCabe, A., & Peterson, C. (1991b). Getting the story: A
longitudinal study of parental styles in eliciting oral personal
narratives and developing narrative skill. In A. McCabe & C.
Peterson (Eds.), Developing narrative structure (pp. 217–
253). Hillsdale, NJ: Erlbaum.
McConaughy, S. H. (1985). Good and poor readers’ comprehen-
sion of story structure across different input and output
modalities. Reading Research Quarterly, 20(2), 219–232.
Menig-Peterson, C., & McCabe, A. (1977). Children talk about
death. Omega—Journal of Death and Dying, 8, 305–317.
Merritt, D. D., & Liles, B. Z. (1989). Narrative analysis: Clinical
applications of story generation and story retelling. Journal of
Speech and Hearing Disorders, 54, 429–438.
Michaels, S. (1981). “Sharing time”: Children’s narrative styles
and differential access to literacy. Language in Society, 10,
423–442.
Miller, P. J., & Sperry, L. L. (1988). Early talk about the past:
The origins of conversational stories of personal experience.
Journal of Child Language, 15, 293–315.
Minami, M., & McCabe, A. (1991). Haiku as a discourse
regulation device: A stanza analysis of Japanese children’s
personal narratives. Language in Society, 20(4), 577–599.
Miranda, E., McCabe, A., & Bliss, L. S. (1993). Jumping
around and leaving things out. Manuscript submitted for
publication.
Peterson, C., & McCabe, A. (1983). Developmental
psycholinguistics: Three ways of looking at a child’s narra-
tive. New York: Plenum.
Pitcher, E. G., & Prelinger, E. (1963). Children tell stories.
New York: International Universities Press.
Preece, A. (1987). The range of narrative forms conversationally
produced by young children. Journal of Child Language, 14,
353–373.
Ripich, D. N., & Griffith, P. L. (1988). Narrative abilities of
children with learning disabilities and nondisabled children:
Story structure, cohesion, and propositions. Journal of
Learning Disabilities, 21(3), 165–173.
Rodino, A. M., Gimbert, C., Perez, C., Craddock-Willis, K., &
McCabe, A. (1991, October). “Getting your point across”:
Contractive sequencing in low-income African-American and
Latino children’s personal narratives. Paper presented at the
16th Annual Boston University Conference on Language
Development.
Roth, F. P., & Spekman, N. J. (1986). Narrative discourse:
Spontaneously generated stories of learning-disabled and
normally achieving students. Journal of Speech and Hearing
Disorder, 51, 8–23.
Sachs, J. (1982). Talking about the there and then: The emer-
gence of displaced reference in parent-child discourse. In K. E.
Nelson (Ed.), Children’s language (pp. 1–28). Hillsdale, NJ:
Erlbaum.
Scarborough, H. S. (1990). Very early language deficits in
dyslexic children. Child Development, 61, 1728–1743.
Strong, C. J., & Shaver, J. P. (1991). Stability of cohesion in the
spoken narratives of language-impaired and normally
developing school-aged children. Journal of Speech and
Hearing Research, 34, 95–111.
Umiker-Sebeok, D. J. (1979). Preschool children’s
intraconversational narratives. Journal of Child Language, 6,
91–109.
Watson, K. A. (1975). Transferable communicative routines:
Strategies in group identity in two speech events. Language in
Society, 4, 53–72.
Weaver, P. A., & Dickinson, D. K. (1982). Scratching below the
surface structure: Exploring the usefulness of story grammars.
Discourse Processes, 5, 225–243.
Received April 22, 1992
Accepted August 24, 1993
Contact author: Allyssa McCabe, PhD, Harvard University,
HGSE, 3rd Floor, Roy E. Larsen Hall, Appian Way, Cam-
bridge, MA 02138-3752
Key Words: narratives, assessment, preschool, literacy,
specific language impairment
on February 10, 2010 ajslp.asha.orgDownloaded from
56 January 1994 AJSLP
Narrative Ten (31-month-old boy)
M: Did you like the puppy?
N: He taste my knee.
M: He tasted your knee?
N: Theth, an puppy chase me!
Questions:
1. Are there two past events?
Yes, “taste” and “chase.”
2. Are there more than two events?
No.
This Two-Event Narrative is typical of 2- to 3-year-old children.
Narrative Eleven (2 1/2-year-old boy)
N: ’Member my book? My baby sitter b(r)oke it.
Questions:
1. Are there two past events?
No, only one, “b(r)oke.” Again,
even though the pronunciation is not exactly correct, this is not a
narrative problem.
Although the chart does not depict narratives by such young
children, from the literature review in the paper we found that such
One-Event Narratives were quite normal productions from 2-year-
old children.
Narrative Twelve (3-year-old boy)
N: I go to Janie’s school and da man hid a white rabbit.
Questions:
1. Are there two past events?
Yes, there are two events, “go” and
“hid.” Even though they are not marked morphologically as past
tense, that would be seen as a grammatical issue rather than a
narrative one.
Such Two-Event Narratives are common from 3-year-old children.
Narrative Thirteen (4 1/2-year-old boy)
E: (rubs elbow) Oh, I hurt my elbow.
M: Hurt head. (boy touches head as he says this)
E: You did? Tell me about it.
M: Fell down.
E: Yeah? What else?
M: That’s all.
Questions:
1. Are there two past events?
Yes, “hurt” and “fell.”
2. Are there more than two events?
No.
This Two-Event Narrative is atypical for a 4 1/2-year-old boy.
Comparison with the other narratives reveals that it is more similar
to the productions of the 2 1/2-year-old (Narrative Ten) or the 3-
year-old (Narrative Twelve) than to the typical multiple-event, Leap-
Frog Narrative told by 4-year-olds (e.g., Narrative Eight). This little
boy is diagnosed as having specific language impairment. His
narrative skills are also delayed, unfortunately.
Appendix
Narrative Eight (4-year-old girl)
E: When I go home I have to visit my aunt who’s in the hospital.
She broke both of her legs. And she has to have them kind of hung
up, suspended from the ceiling with those little wires.
B: She had to have cast on.
E: That’s right.
B: My sister had, she’s had. She broke a arm when she fell in
those mini-bike.
E: Tell me about what happened.
B: She broke her arm. She had, she went to the doctor, so I, my
Dad gave me spanking, and I
E: Your Dad gave you what?
B: A spanking to me.
E: A spanking?
B: Yeah. And she had to go to the doctor to get a cast on. She had
to go get it, get it off and, and it didn’t break again.
E: And then it didn’t break again?
B: No. She still got it off. She can’t play anymore.
E: She can’t play anymore?
B: She can’t play we, she can play rest of us now.
E: Oh good. ... Have you ever had a shot?
B: Mm, she has cast on. When she was home. When she came
back and she, and she, and she hadda go back and, take off the
cast.
E: She had to go back and take off the cast?
B: Yeah. The doctor.
Questions:
1. Are there two past events?
Yes: broke, went, gave, had to go.
came back, had to go and take.
2. Are there more than two such events?
Yes.
3. Is there a logical, perhaps causal sequence to those events in
the real world or could they have occurred in any order?
Yes,
events precipitating an injury and the follow-up treatment have a
tight causal sequence in the real world.
4. Does the narrator’s order of delivering those events mirror the
sequence in which the events must have logically occurred?
No,
narrator jumps around in time and leaves out what we infer was her
iniquitous act that actually caused her sister to fall off her bike and
break her arm.
Thus, this narrative is termed a Leap-Frog Narrative, quite typical
of 4-year-old narration.
Narrative Nine (23-month-old girl)
J: I hied the big boy.
Questions:
1. Are there two past events?
Again, no there is only one event,
and that is an idiosyncratic formation
.
Such One-Event Narratives are really quite good productions for
children not even 2 years old.
on February 10, 2010 ajslp.asha.orgDownloaded from
45-563, American Journal of Speech-Language Pathology,
1994
Allyssa McCabe and Pamela Rosenthal Rollins
Assessment of Preschool Narrative Skills
This information is current as of February 10, 2010
http://ajslp.asha.org
located on the World Wide Web at:
This article, along with updated information and services, is
on February 10, 2010 ajslp.asha.orgDownloaded from
