Journal of Reading Behavior

1991, Volume XXIII, No. 2

ON THE OUTSIDE LOOKING IN: A STUDY OF REMEDIAL

READERS' MEANING-MAKING WHILE
READING LITERATURE

Victoria Purcell-Gates
University of Cincinnati

ABSTRACT

This study compared remedial readers to more proficient readers in their

meaning-making processes during the reading of literary text. Six children
attending a university-based Literacy Center read aloud two stories and verbalized
their thoughts during reading. Langer's categories (1990) were used to compare
the resulting think alouds. Langer found readers assume four stances as they make
meaning while reading literature: (a) being out and stepping into an envisionment,
(b) being in and moving through an envisionment, (c) stepping back and rethinking
what one knows, and (d) stepping out and objectifying the experience. Langer
described readers as always working toward an evolving understanding of
the whole, and their envisionment of the whole affects their momentary
understandings. A qualitative analysis was also conducted as other categories
emerged and all the responses were recategorized for difficulties experienced by
remedial readers. Analysis revealed that the remedial readers spend a dispropor-
tionate amount of being out of envisionments—either attempting to step into one
or failing to step in. Also, these readers consistently failed to construct evolving
wholes and struggled with the language of literary discourse. The overall picture
of remedial readers gained from this study is one of being on the outside looking
in. Implications for and of instruction are discussed.

The relationship between the reader and the text during the act of reading has
been the focus of a great deal of work over the last few decades. Theorists and
researchers from such areas of inquiry as literary theory and reading comprehension
have attempted to describe what happens when readers enter the world of text.
Although each field has perceived and approached this question somewhat differ-
ently, the resulting consensus paints a picture of a dynamic construction of meaning
by the reader using the text as a guide for this interpretive process (Anderson,

235
236 Journal of Reading Behavior

1984; Fish, 1989; Iser, 1978; Langer, 1990; Rosenblatt, 1978; Smith, 1985). That
comprehension relies upon the activity of the reader in the construction of a whole
is generally now taken as a precept among reading theorists.
Accepting this description of the reader/text relationship, we can begin to
examine the ways in which children who are performing well below grade level
on measures of reading achievement go about constructing meaning while reading.
Fresh insights into the difficulties experienced by remedial readers are desperately
needed. Years of research have yielded little information about operative factors
involved in reading difficulties beyond the word identification level. Results from
studies of poor readers which looked beyond this factor, however, suggest a prob-
lem in this area of the reader/text relationship.
Comparisons of good and poor readers reveal that poor readers read in a
piecemeal, word-by-word manner, are slower and less accurate, and fail to access
or utilize larger language and idea units in the written text (Cromer, 1970; Daneman
& Carpenter, 1980; Golinkoff, 1975-1976; Perfetti & Lesgold, 1977, 1979). Re-
medial readers exhibit problems processing text at all levels and suffer severe
difficulties with comprehending written text. Explanations for reading difficulties
center around two different theories. One theory suggests that remedial readers'
comprehension problems are the result of difficulties with accessing semantic infor-
mation during reading (Zecker & Zinner, 1987). Claiming that one area which
could be affected by a verbal deficit (Vellutino, 1979) is semantic processing,
Zecker and Zinner suggested the following scenario:
(a) poor readers and good readers may have the same overall amount of semantic
code information available to them, but poor readers may be slower in accessing
that information; (b) poor readers may have an impoverished semantic code relative
to good readers, perhaps in the sense of an incomplete semantic network (Collins
& Loftus, 1975); or (c) poor readers may have both an impoverished code and
slower semantic access relative to good readers; or (d) good and poor readers may
differ in ability on automatic or controlled processing of semantic information, or
both. (p. 178)
Another theory postulates that remedial readers fail to use self-regulatory strate-
gies to monitor comprehension progress and to take remedial action when compre-
hension is impeded (August, Flavell, & Clift, 1984; Bos & Filip, 1982; Flavell,
1981; Garner & Reis, 1981; Meichenbaum & Asarnow, 1979; Owens, Peterson,
Bransford, Morris, & Stein, 1980; Ryan, 1982; Wagoner, 1983). Johnston (1985)
built on this by suggesting that at least some portion of reading disability is the
result of a combination of conceptual difficulties, rational and irrational use of
self-defeating strategies, and negative affective responses.
Poor readers seem to perceive reading as mainly a decoding process rather
than one of meaning construction or comprehension (Gambrell & Heathington,
1981). They are characterized by a passive approach to meaning making (Johnston
& Winograd, 1983). From the perspective of reading response theory, this passivity
 On the Outside Looking In 237

distorts the reader/text relationship by perceiving the text as the purveyor of mean-
ing rather than as the signifier of the meaning to be constructed by the reader.
Whether or not poor readers' failure to actively construct meaning during
reading is inextricably linked to problems at the word level is not fully understood.
This study is an attempt to achieve greater understanding of this aspect of remedial
readers' processes of meaning making during reading. The perspective of this study
is somewhat unique in that the phenomenon of remedial readers' meaning making
during reading is viewed through a reader response lens as well as from the perspec-
tive of the literature on remedial reading. It is hoped that this will allow for greater
validity of results and interpretation by broadening the perspective on the issue.
Langer's (1989, 1990) recent research into the nature of meaning-building
processes during the reading of literary and nonliterary text provides a vehicle for
such a study. Seeking to discover ways in which readers develop their understand-
ings during the reading of literature and text material in other area classes, Langer
asked students in middle and high school (Grades 7-12) to engage in think alouds
as they read several short stories, poems, and science and social studies pieces.
Langer's perspective is that reading is sense-making during which understandings
develop and change as the reading progresses. These understandings involve the
creation of local envisionments, " . . . a personal text-world embodying all she or
he understands, assumes, or imagines up to that point in the reading" (Langer,
1989, pp. 4-5). Since envisionments are evolving wholes which are subject to
change even after the reading is finished, Langer claimed that it is necessary to
examine readers' sense making across time throughout the course of the text.
Analysis of the think alouds in Langer's study revealed four stances the readers
took as they developed their meanings across time with both literary and nonliterary
texts. These stances reflect varying reader/text relationships and are recursive rather
than linear. The four stances are: (a) being out and stepping into an envisionment,
(b) being in and moving through an envisionment, (c) stepping back and rethinking
what one knows, and (d) stepping out and objectifying the experience.
Being out and stepping into an envisionment occurs as the reader attempts to
establish initial understanding of the content, genre, characters, language, and so
on. It happens at any point where unfamiliar or unexpected content is encountered.
When encountering such content in the text, the reader may ask questions, make
associations, and create a general context for understanding the reading. An exam-
ple is an initial description of a character: His grey head gently bobbing, William
made his way haltingly down the street. While reading, a reader may speculate,
"He must be a very old man." Langer stated that it is a time when readers'
envisionments do not cohere because not enough text knowledge has been estab-
lished from which to predict.
Being in and moving through an envisionment occurs when readers use the
ongoing text to build an evolving envisionment. They elaborate and make connec-
tions among their ideas. Within this stance, a reader may (building on the previous
238 Journal of Reading Behavior

example) express, "He's very lonely now and starting to feel depressed." This
stance occurs after the initial understanding is established and the reader is re-
sponding to additional text and elaborating upon the envisionment. As Langer
(1989) defined it, "Being in and moving through an envisionment describes the
engaged moments when readers used personal experiences and knowledge as well
as the text to push their envisionments along—where meanings begot meanings
(p. 10).
Stepping back and rethinking what one knows describes what readers do when
they use their growing understandings to step outside of the envisionment to reflect
upon their own lives or their own knowledge. Langer explained that in other
stances, readers use their background knowledge and experiences to help them
understand the text, whereas in this stance the envisionments are used to help the
reader reflect on thought, actions, or feelings outside of the text—the reader's
personal world. "I never thought about old people getting depressed like this," a
reader may reflect during this stance.
Stepping out and objectifying the experience occurs when readers distance
themselves from their final envisionments and reflect on the reading activity, their
understandings, and their reactions. They may also comment on the text and the
reading experience. "This story was so intense!" a reader could conclude, or " I
never did like science fiction; there's never enough emotions." Langer emphasized
throughout her analysis of the data that readers are always working toward an
evolving understanding of the whole and their envisionment of the whole affects
their momentary understandings.
Langer's sample for this study included students judged by their teachers as
performing below grade level as well as average and above average. The results
indicated that although most of the students engaged in the full range of stances,
there was some initial indication that the poorer readers often spent more time in
the being out and stepping into an envisionment stance than did the other readers.
This is suggestive of a locus of difficulty for poor readers. A replication of part of
Langer's study with students with severe reading difficulties could explore this area
in greater depth. As such, this study was designed to: (a) examine identified reme-
dial readers' meaning making processes through the lens of Langer's description
of stances taken by readers of literary text, and (b) search for factors which may
be related to difficulties experienced by these readers as they attempt to construct
meaning during the reading of literature.

METHOD

Participants and Setting

The student-informants for the study all attended a university-based Literacy
Center, a remedial clinic for children Grades 1-8 where children's literature is
 On the Outside Looking In 239

used as the material for instruction. Children attend the Center two times a week
for diagnosis and instruction in reading and writing. The children each work one-
to-one with a graduate student enrolled in a three-quarter course sequence in diagno-
sis and remediation of reading/writing problems. This sequence is taught by the
author who also directs and administers the Literacy Center. Instruction and diagno-
sis in the Center reflect a wholistic perspective with the children reading and writing
for self-selected purposes with real-world materials, consisting mainly of a full
range of children's literature, fiction and nonfiction. The Center was designed to
provide access to quality instruction for children from the economically depressed
urban neighborhoods surrounding the university. Thus, three quarters of the chil-
dren have parents helping in the Center in exchange for instruction. The other
parents pay nominal fees, based on a sliding scale of family income.
Six children in Grades 6-8 were identified for the study. Out of the 21 children
enrolled in the Center, only these 6 met the requirements for the study: (a) as close
as possible to Langer's sample in terms of grade level in school, and (b) capable
of reading children's literature without decoding problems.
All of the children in the sample were reading at an independent level of at
least 3.0, according to the Analytical Reading Inventory (Woods & Moe, 1989).
According to this inventory, the independent level is the level at which the student
can read with no more than one uncorrected miscue in each 100 words (99%) and
with at least 90% comprehension. In some cases, the students reading at this level
may exceed this criterion for miscues if comprehension is maintained at the 90%
level or higher.Additionally, all of the children in the sample were performing at
least 2 grade levels below norm for their individual grade levels, according to the
Stanford Achievement Test (1987) and the Analytical Reading Inventory.
Four boys and two girls comprised the sample, four black and two white. They
all lived within the limits of a large midwestern city. They were all enrolled in
public schools, and none of them received special help outside of the Center for
their reading problems, although they were all assigned to the lower academic
tracks in their respective schools. It was agreed upon by the graduate-student
teachers and the author that, although each child in the sample exhibited different
types and degrees of reading disabilities, taken together they represented the span
of remedial readers typically encountered in schools.

Materials
Langer's (1990) initial findings on the type of stances taken by poorer readers
were based on data collected during the reading of material which, according to
Langer, could well have been too difficult for the low readers. To unconfound the
factors of poor word knowledge and general proclivity toward a particular stance
during reading, material was chosen for this study which was on the independent
reading level of all of the students, according to the Fry Readability Formula
240 Journal of Reading Behavior

(1972). The two short stories used for data collection were "The House on the
Hill" (Minturn, 1973), and "The Real Princess" (Moore, 1969).
"The House on the Hill" is about a teenage boy who feels unwanted and
alienated. He has recently been placed in a foster home, due to the break-up of his
family, and has begun shoplifting and storing the items in an abandoned house.
The plot relates an event in which he is nearly caught stealing and his subsequent
realization that his foster parents do care for him.
"The Real Princess" is a version of Hans Christian Anderson's "The Princess
and the Pea." This is the tale of the princess who proves her royalty by feeling a
pea through many layers of mattresses.

Procedure
The students were introduced to the study by their individual teachers at the
Center. The think aloud procedure (as implemented by Langer, 1990) was taught
to them through explanation and modeling. The teachers explained to the students
that they were being asked to participate in a study which would help reading
teachers understand what readers thought about as they read stories. One way to
do this, explained the instructors, was to have readers say aloud what they are
thinking as they read orally. Their readings, complete with their think alouds,
would be tape recorded for later analysis. The students were assured that their
performance on this task would not affect them personally in any way, such as
through evaluations or reports to parents. Each teacher then modelled the think
aloud procedure while reading a story.
The use of think-aloud protocols is often employed by researchers investigating
students' orchestration of reading and writing strategies over time (Flower & Hayes,
1980; Haas, 1989; Hynds, 1989; Langer, 1989; Long, Winograd, & Bridge, 1989;
Newell, 1984; Seife, 1984). This procedure has students verbalizing their thoughts
at given or self-selected points during the reading or writing process and allows
researchers access to otherwise hidden cognitive processes. Although think-aloud
protocols provide information unattainable through other measures, their use does
have drawbacks. Naturally occurring reading and/or writing events do not include
the verbalization of cognitive plans, imagings, or reflections. Thus, ecological
validity is compromised (Cooper & Holtzman, 1983). Additionally, the impact of
the necessity to verbalize thought on the process under study, whether it be reading
or writing, is unclear but undoubtedly present to some degree (North, 1987). There-
fore, all conclusions drawn from the use of think-aloud protocols should be inter-
preted with these limitations in mind.
For the think alouds, the students were encouraged to verbalize their thoughts
when they felt comfortable rather than at prearranged text breaks to minimize the
disruptive effect of the think aloud procedure on the development of meaning.
However, if the teacher felt a student was going for too long without offering a
 On the Outside Looking In 241

think aloud, a tap on the shoulder (teachers sat slightly behind each student to
minimize distraction) was used to remind the student about the procedure. Each
student practiced the procedure with three practice stories written at the same
readability level as the data collection stories. The students wore wireless micro-
phones and their readings were taped during the practice sessions to ensure familiar-
ity with the taping procedure to be used during data collection. The practice sessions
extended over several weeks, with the teachers modelling the think aloud procedure
each time.
When each teacher felt the student was familiar and comfortable with the
procedure, data were collected on the two stories. Only one story was read per
session to avoid fatigue. All of the data collection sessions were audiotaped for
later analysis.

Analysis. Analysis was partly descriptive in that Langer's categories were used
initially to categorize as many of the think aloud responses as possible. The unit
of analysis was each response. A response was defined as each time a reader
stopped reading'to offer a single, unitary thought. Each response was bounded by
text, that is, no pause in reading for reader comment contained more than one topic
of response. Although it is possible for one response to contain evidence of more
than one stance, none of the students in the sample did this; each response in the
data reflected only one stance. A more qualitative analysis also resulted as: (a)
other categories of response emerged, and (b) responses viewed through the stance
lens were reviewed and categorized for qualities of the responses which seemed to
offer insights into difficulties experienced by remedial readers.
Responses were assigned to one of Langer's categories through the following
procedure: (a) each transcript was read with Langer's definitions and examples of
each stance in mind; (b) each response was examined for the reader's concern at the
time (e.g., did the response indicate a concern with establishing an envisionment or
with developing an already established envisionment?); and, (c) if the response
reflected one of Langer's stances, it was thus labeled. Responses which did not fit
one of Langer's stances were examined again for other patterns of response; one
other pattern was identified. All responses were reexamined again to verify catego-
ries. This recursive procedure continued until the researcher was convinced as to
the uniqueness of each classification. Responses not falling into a pattern were
withheld from further analysis (7 out of 270). These responses were either off-task
comments or one which made no sense to the researcher.
Reliability of category assignment was assessed by having a trained graduate
student mark six randomly selected transcripts (three for each story). The Pearson
product moment correlation between the first and second category assignment was
r=.87.
Percentages were derived by dividing the total number of responses in a cate-
gory by the total number of all responses assigned to categories. The results are
242 Journal of Reading Behavior

based on analysis of 263 think-aloud responses garnered from six students reading
two short stories.

RESULTS

Analysis revealed that many of the think-aloud responses reflected Langer's

stances, with the majority falling into the being out and stepping into an envi-
sionment stance. Another pattern of response emerged, however, which was not
reported by Langer. This type of response reflected a passive, reiterative orientation
toward the text and was exemplified by a repeating or restating of the text.
A qualitative analysis of the content of the responses revealed difficulties for
these readers on two levels: (a) failure to construct wholes, and (b) struggle with
the language of literary discourse.

Stances
Figure 1 illustrates the pattern of response by the remedial readers in this study.
Langer's initial indicator of poor readers spending more time trying to get into
envisionments was reinforced by the results of the present study. Forty eight percent
of the responses were of the being out and stepping into an envisionment stance.
Examples of this stance follow, showing readers' attempts to make initial acquain-
tance with local events throughout the stories. They are from the readings of
"House on the Hill" (student comments are italicized whereas the text remains
unmarked):
He could see his high school near his old home. He started to think of the days
when he lived with his parents.
So he don't live with his parents anymore and he don't care about them.
From its window, Mark could see down to the town. The room seemed far away
from everything. Nobody could reach him here.
Maybe it was a mansion.
In addition, 13% of the total number of responses were of the repeating/
restating text type which can be interpreted as a being out and failing to step into
an envisionment, a category not reflected in Langer's data. These responses took
the form of either a verbatim rendering of text language just read or a mere re-
statement of the just-read text. All of the readers produced this type of response.
Examples of this follow from the story "House on the Hill":
He was like the house on the hill.
He was like the house on the hill.
He walked through the dusty halls, up the dirty stairs, and into the front room.
 On the Outside Looking In 243

100
LU
CO

o
CL
CO
LU
DC 48%
LU 50 —
O *
f- 36%
Z
LU
Ü
DC 13%
LU
Q.
1%
•si'
Being Out and Being Out and Being In and Stepping Back Stepping Out
Stepping Into Failing to Step Into Moving Through and Rethinking and Objectifying
an Envisionment an Envisionment an Envisionment What One Knows the Experience

Figure 1. Percentage of responses for stance categories.

So he walked . . . he go up there . . . to the front room.

It was Sam, looking like he didn't know just what to say. He put his hands in his
pockets, and stood there for a second, trying to say something.
Sam didn't know what to say.

Thus, compared to Langer's total sample, these readers spent much more time
being out of an envisionment.
Thirty-six percent of the responses, however, did reflect the being in and
moving through an envisionment stance. In this stance, readers further their con-
struction of an already established envisionment. Elaborations and a sense of con-
nectedness to other ideas distinguish this stance from the being out and stepping
in stance. Examples of these responses for "House on the Hill" follow:
He must have taken it just for kicks, but he sure didn't look like he was getting
any kicks now.
He was very . . . I mean he looked sad 'cause he got busted.
Mark felt sick, and his knees were shaky. The lunch box was suddenly very heavy.
He found himself moving back to the camera counter.
244 Journal of Reading Behavior

He was so scared he was gonna put it back.

Mark wanted to run, but his feet wouldn't move. They were frozen to the floor.
"This will give you a record, you know," the voice continued. Mark turned
around slowly, but there was no one behind him.
// was his conscience.

The readers in this study revealed very few attempts to connect the text world
to their own. Only a few of the total think-alouds revealed a stepping back or out
of an envisionment by these readers. One percent of the total responses resembled
Langer's stepping back and rethinking what one knows stance. Although there was
no evidence of rethinking in these responses, they did reveal a stepping back from
the text stance. A more accurate term for this stance would probably be stepping
back and connecting to own world. Examples from "The Real Princess" follow:
One night there was a storm. And what a storm! Thunder rolled. Lightning flashed.
And the rain came down as if it would never stop.
That's the way it was all this weekend.
Lightning flashed. And the rain came down as if it would never stop. It was a
frightful night.
It's not that frightening. Lightning's just not that frightening.
Two percent of the responses reflected the stepping out and objectifying the
experience stance. Most of these occurred after the readings of "The Real Prin-
cess," the text of which ended with the question "That's some story, isn't it?"
which invited readers to assume this type of stance. Examples follow:
(From "House on the Hill") Sam was all right. He would probably understand
things. Maybe, with Sam around to help, Mark would make it, too.
Good. Wheee. Good story; good story; good story.
(From "The Real Princess") That's some story, isn't it?
Yeah. That's a pretty good story.

Failure to Create Wholes

The readers in this study consistently revealed a strong tendency to create
meaning on local event-levels rather than working from an understanding of the
whole as they read. In this way, they differed markedly from Langer's more general
sample. Langer (1989) concluded that readers in her study " . . . took each idea
they read and tried to understand it in terms of their sense of the whole, rather than
as a stepping stone along the way. They clarified ideas as they read and related
them to the growing and changing horizon—that horizon modified the parts and
the parts modified the horizon" (p. 16).
Langer's portrait of the meaning construction process implies a proactive
 On the Outside Looking In 245

stance by readers. The think alouds produced by the problem readers in this study
reflected more of a reactive approach to the reading of literature. These readers did
not take charge of the reading experience, creating wholes and establishing varying
relationships to the text as they developed their meanings across time. Rather, they
seemed to engage in reactions to the text on the local level with little attempt to
tie what they were reading to their own experiences or to reflect upon what they
had read. They were so tied to the text that they spent an unexpected amount of
time merely restating or repeating the text language they had just read.
The relative amount of time they spent in the being out and stepping into an
envisionment stance (48% of responses) is a clear indicator of this tendency to
create meaning on the local level rather than work with an evolving concept of the
whole. Following is an excerpt from Melissa's transcript which provides a compel-
ling example of this. This example comes from "The House on the Hill" and
occurs near the end of the story. Mark has survived nearly being caught stealing,
has decided not to steal again, and is establishing a relationship with his foster
father (Sam). They are out for a walk and notice that the abandoned house has
been purchased:

"Well," said Sam, "someone finally bought this place. It's a good place." "You
like it, too?" Mark asked. "I always have," Sam answered. "It just never had
the right owners."
Maybe he wants to move in it also.
Sam walked on slowly.
Probably so they could get closer and closer.
Mark stood there and thought of the things in the closet.
Probably there were ghosts.
The new owners would think the last people had left them.
Probably was his parents.
The things weren't his any longer.
Probably those who moved into that house.
He smiled. They never had been his. Mark looked again at the house, shining
with its new coat of paint.
Maybe somebody owns that paint.
Someone had seen through the weeks and the dust.
Maybe somebody was cleaning up the place.
Someone had seen that a really nice house was under it all.
Maybe there were some ghosts.
"You made it, house on the hill," Mark said softly. Then he ran to catch up with
Sam. "Wait a minute, Sam!" he called. Sam was all right.
246 Journal of Reading Behavior

Maybe he had been sick.

He would understand things.
One has the sense, reading this, that Melissa has read this story up to this
point with no concept of the whole. Although she had already read that Mark stored
his stolen items in the closet, she thinks of ghosts at this mention of items in the
closet. Two lines later, she considers the possibility that the items in the closet
(ghosts?) may belong to the new owners of the house. Her think-aloud comments
reveal almost stream-of-consciousness reactions to the preceding text, event-by-
event, with no attempt to make sense of the text in relation to a meaningful whole.
Although it is possible that the think-aloud procedure encouraged this, the readers
in Langer's study, working with the same procedure, did not respond in the same
way.

Struggle With Literary Language

Perhaps related to this disabling strategy were the readers' struggles with the
less-than-completely-explicit language of the literary text. Many of the think alouds
revealed a focus on the meaning of the words or language just read. This struggle
took several forms.
First, all of the readers exhibited problems with assigning appropriate meanings
to the figurative language in the text. Figures of speech were often interpreted in
ways which did not make sense in the context of the story being read. For example,
Melissa responds in the following manner during her reading of "House on the
Hill":
"All right, kid, let's see what you have," said a loud voice behind Mark. "Come
on, hand it over." Mark wanted to run, but his feet wouldn't move. They were
frozen to the floor.
Probably there was ice on the floor.
Readers had similar problems with words with multiple meanings, often assum-
ing an inappropriate one. An example from Alan's reading of "House on the Hill"
follows:
Mark wanted to run, but his feet wouldn't move. They were frozen to the floor.
"This will give you a record, you know," the voice continued.
Probably a music record.
The readers in this study also revealed a need to translate the language of the
text into more explicit language through their think alouds. This often took the
form of the filling in of ellipted information at the text level (from "The House on
the Hill":
 On the Outside Looking In 247

Mark left the window and went to the small hall closet.
To put his clothes in there.
He kept thinking about being caught.
By some police.
(From "The Royal Princess") What did they do with the pea? Oh, they put it in
a glass box, and they put the box in a museum where everyone could see it.
For whoever came first.

This type of response was characterized as filling in at the text level because of
the reading intonation nature of the think alouds. If one were only listening to the
taped readings, it would be easy to assume that the responses underlined above
were actually part of the text being read. This quality of the intonation distinguished
these responses from others.
This explicit stating of pronominal referents was also an indicator of this need
to make the language explicit (from "The Royal Princess"):

The old king himself went out to open it.

The king went to open it, the door. His door.
He even had dreams about it.
Probably about the policeman.

Many of the think alouds revealed the readers' struggles with inferred informa-
tion. They would often state the inference explicitly as if explaining to themselves
(from "The House on the Hill"):

As they ate they asked him about school. He answered their questions by saying
just " y e s " or " n o . " He was too busy thinking to talk to them.
So he like . . . so they asked what he did at school but he didn't go to school, he
was at the old house.
Mark looked again at the house, shining with its new coat of paint. Someone had
seen through the weeds and the dust. Someone had seen that a really nice house
was under it all.
So somebody wanted that house badly and they bought it.

At times, their incorrect inferences or their questions would reveal confusion

(from " T h e House on the Hill"):
"Well," said Sam, "someone finally bought this place. It's a good place." "You
like it, too?" Mark asked. " I always have," Sam answered. "It just never had
the right owners."
Sam bought the house.
248 Journal of Reading Behavior

(From "The Real Princess") What in the world was in that bed? It was so hard
that I am black and blue all over.
/ don't know how she got all black and blue.

Finally, the tendency of these readers to simply repeat or restate text was also
taken as evidence of a struggle with the language of the text. Their behavior of
repeating the language verbatim and/or restating it appeared to function as transla-
tions into more familiar, thus more comprehensible, language or to repeat it in
such a way as to try again for the meaning. This behavior was pervasive throughout
the data. It is possible that the nature of the think aloud procedure may have
contributed to this repeating/restating behavior, that is, the readers may have re-
sponded to the requirement to say something by simply repeating the text, espe-
cially when reminded to respond with a tap on the shoulder. This argument weak-
ens, however, with a review of the data which reveals that 81 percent of the
repeating/restating responses were the result of spontaneous comments rather than
of reminders.

DISCUSSION

This study compared remedial readers (Grades 6-8) to Langer's (1990) more
proficient readers (Grades 7-12) in their meaning-making processes during the
reading of literary text. It also provided insights into some of the problems experi-
enced by problem readers as they read literature. The overall picture of remedial
readers gained from this study is one of being on the outside and looking in. They
rarely, if ever, "evoke a poem" in the words of Rosenblatt (1978, p. 48). They
find it difficult to move into an envisionment, and when they do, they elaborate
upon it only momentarily before they again find themselves outside trying to get
in. Langer reports similar patterns for the poor readers in her study. These readers
did move into envisionments, she relates, but seemed unable to maintain them
(personal communication, November 24, 1989). The results of this study suggest
that problems with the text may be related to the reported appearance of an overall
passive, reactive posture assumed by the remedial reader. These problems are not
at the word identification level, but rather lie within the reader/text relationship.
Text plays an essential semiotic function in meaning construction during read-
ing. It serves to provide guideposts and clues to the reader who must create the
envisionment (Iser, 1978; Rosenblatt, 1978). Rosenblatt assigned two major func-
tions to the text: (a) it serves as a stimulus in activating elements of the reader's
past experience with both literature and life, and (b) it serves as a blueprint, "a
guide for the selecting, rejecting, and ordering of what is being called forth; the
text regulates what shall be held in the forefront of the reader's attention" (p. 11).
The ordering of signs, or the blueprint, provided by the literary texts in this study
 On the Outside Looking In 249

did not serve this function for these remedial readers. Rather, their struggles with
the language—". . . the unique pattern of words which constitutes the text"
(Rosenblatt, 1978, p. 11)—revealed more of a gatekeeping role for the text, keep-
ing these readers out of a created literary experience.
It may be that the passive posture toward text reported of remedial readers
may reflect not only a behavioral predisposition but may also be an artifact of this
difficulty with the semiotics of the text. This is suggested by the fact that the
readers in this study and in Langer's did at times attempt to maintain envisionments.
Their continued failure to do so may be indicative of years of such failure which
would account for the behavioral aspect of passivity reported by others. In other
words, poor readers may have experienced difficulty with moving in and out of
text through the creation of wholes and local envisionments from the beginning
and, in essence, given up and assumed a passive posture toward meaning making
while reading.
In a very real sense, the responses of the remedial readers reflected more of an
efferent approach than the aesthetic which is integral to literary reading (Rosenblatt,
1978). Rosenblatt defined efferent reading as reading in which the reader concen-
trates on what the symbols (the text) designate, the information he or she will take
away from the reading, rather than on the actual experience the reader is living
through during the reading event as in the aesthetic approach. The focus on the
literal meaning of the text by the readers in this study, taken to the extreme of
simple repetition or restatement, reflected a concern with efferent reading of literary
text, almost as if they were answering anticipated questions of the what happened
in the story type. (Langer [1989, 1990] described a similar distinction contrasting
readers' differing orientations to literary and nonliterary readings. In this system,
it can be considered that the poor reader read a literary text in a nonliterary manner.)
The nature of the difficulties with the language of the text experienced by the
readers in this study suggests provocative insights into the comprehension difficul-
ties experienced by remedial readers. Two distinct explanations for these difficulties
arise from the literature. Johnston (1985) characterizes these two types of theories
as similar to the state/trait tension in personality psychology. One explanation of
the readers' difficulties with the text language would be that remedial readers
possess inherent language disabilities which make the processing of figurative and
implicit meanings more difficult. This places the locus of the problem within the
reader and thus is a trait explanation. Another theory posits that remedial readers
have been taught, or interpreted instruction in such a way as, to focus only on
surface aspects of text and, thus, have never learned to actively construct meaning
through the use of the text as a blueprint. This explanation stems from a state-
orientation. "In this case, the expected cause is less long-term and more amenable
to instruction than when the trait notion is the guiding one," according to Johnston
(p. 154).
It may be that the truth lies somewhere between these two poles. Enough data
250 Journal of Reading Behavior

exist to suggest that many remedial readers show evidence of general language
difficulties which exhibit themselves beyond the reading situation. Studies also
show that most remedial reading instruction focuses on the literal text level when
it does go beyond the word identification focus so predominant in the field. It is
possible that general language difficulties contribute to the comprehension problems
experienced by many remedial readers and that these readers find their dysfunc-
tional meaning-making strategies reinforced by the instruction they receive in reme-
dial reading classes.
The qualitative analysis of the data in this study revealed a difficulty with
figurative language and with implicit meanings requiring inferencing. Taking the
case of figurative language alone, several reading researchers have acknowledged
its role in comprehension difficulties (O'Brien & Martin, 1988). It an examination
of Ortony's (1979) salience imbalance model of figurative language comprehen-
sion, Readence, Baldwin, Martin, and O'Brien (1984) hypothesized a series of
steps a reader must make in order to correctly process figurative language. Although
a complete treatment of this subject would be too involved for this discussion,
suffice it to say that one of the steps that they posit as necessary for comprehension
of figurative language is a disregard of matching attributes (of the term being
compared and the term doing the qualifying, e.g., his feet were like ice) that are
trivial or that lead to fantastic or contextually inappropriate interpretation. The data
from this study suggest that this was a crucial step which these readers failed to
make as they attempted to process the figurative language of the texts. Thus, it is
possible that the failure of the readers in this study to get into envisionments and
stay there to elaborate upon them as they created an evolving vision of the whole
is at least partially related to their difficulties with figurative and inexplicit language
so typical of written text.
Suggesting a state explanation for a phenomenon does not necessarily mean
that it is resistant to alteration through education, as suggested by Johnston (1985).
Actually, problems with figurative language and inferencing are typical of certain
language disorders and are amenable to language therapy (B. Kuvshinoff, February
7, 1989, personal communication). Instruction for remedial readers which is
process-based and which focuses on facilitating the active meaning-construction
basic to the reading process may very well lead to advances in our success rate with
remedial readers. Given further research on the suggested link between language
dysfunction and reading comprehension, this instruction could also include explicit
teaching of ways to understand figurative language and implicit meanings carried
by the text. Surely, at the very least, the results of the present study imply that
remedial instruction must focus beyond word-level difficulties to the very nature
of the reader/text relationship. Additionally, research on such instruction would
contribute to our knowledge in this domain.
Several limitations to the results of this study also suggest further research.
One is the age difference of the subjects between the sample for this study and
 On the Outside Looking In 251

Langer's sample, to which they were being compared. Although an age overlap
exists (Grades 7 and 8) between the two samples, the average ages of the subjects
differ, with Langer's average age being two and a half years older. This may well
contribute to the student's ability to better create evolving wholes as they make
meaning while reading literary selections. The factors of skill level and age need
to be unconfounded in further research. Also, the relationship between general
cognitive development and strategies of meaning making during reading is yet to
be examined through studies of children of all ages reading literature.
Research is also needed to compare remedial reader's meaning-construction
processes during non-literary reading to those documented in this study during
literary reading. Also, the contribution of choice and ownership over reading mate-
rial to meaning-construction processes must be examined. It may very well be that
the picture of reader/text relationship will look differently for remedial readers if
they are reading self-selected literature for their own purposes.
Another limitation to this study is the small sample size, dictated in this case
by available remedial students accustomed to reading literature. As literature-based
reading instruction expands in the schools, further studies can investigate the find-
ings of this study with larger samples.

REFERENCES

Anderson, R. C. (1984). Role of the reader's schema in comprehension, learning, and memory. In
R. C. Anderson, J. Osborn, & R. J. Tierney (Eds.), Learning to read in American schools: Basal
readers and content texts (pp. 243-258). Hillsdale, NJ: Erlbaum.
August, D. L., Flavell, J. H., & Clift, R. (1984). Comparison of comprehension monitoring of skilled
and less skilled readers. Reading Research Quarterly, 20, 39-53.
Bos, C., & Filip, D. (1982). Comprehension monitoring skills in learning disabled and average students.
Topics in Learning and Learning Disabilities, 2, 79-85.
Collins, A. M., & Loftus, E. F. (1975). A spreading-activation theory of semantic processing. Psycho-
logical Review, 82, 407-428.
Cooper, M., & Holtzman, M. (1983). Talking about protocols. College Composition and Communica-
tion, 34, 284-293.
Cromer, W. (1970). The difference model: A new explanation for some reading difficulties. Journal of
Educational Psychology, 61, 471-483.
Daneman, M., & Carpenter, P. (1980). Individual differences in working memory and reading. Journal
of Verbal Learning and Verbal Behavior, 19, 450-466.
Fish, S. (1989). Doing what comes naturally. Durham, NC, and London: Duke University Press.
Flavell, J. H. (1981). Cognitive monitoring. In P. Dickson (Ed.), Children's oral communication skills
(pp. 35-39). New York: Academic Press.
Flower, L., & Hayes, J. R. (1980). Identifying the organization of writing processes. In L. W. Gregg
& E. R. Steinberg (Eds.), Cognitive processes in writing (pp. 3-30). Hillsdale, NJ: Erlbaum.
Fry, E. (1972). Reading instruction for classroom and clinic. New York: McGraw-Hill.
Gambrell, L. B., & Heathington, B. S. (1981). Adult disabled readers' metacognitive awareness about
reading tasks and strategies. Journal of Reading Behavior, 13, 215-222.
252 Journal of Reading Behavior

Gardner, E. G., Rudman, H. C., Karlsen, B., & Merwin, J. C. (1987). Stanford achievement test
(Forms E/F). San Antonio, TX: Psychological Corporation.
Garner, R., & Reis, R. (1981). Monitoring and resolving comprehension obstacles: An investigation of
spontaneous text lookbacks among upper grade good and poor comprehenders. Reading Research
Quarterly, 16, 569-582.
Golinkoff, R. M. (1975-1976). A comparison of comprehension processes in good and poor compre-
henders. Reading Research Quarterly, 11, 623-659.
Haas, C. (1989). How the writing medium shapes the writing process: Effects of word processing on
planning. Research in the Teaching of English, 23, 181-207.
Hynds, S. (1989). Bringing life to literature and literature to life: Social constructs and contexts of four
adolescent readers. Research in the Teaching of English, 23, 30-61.
Iser, W. (1978). The act of reading. Baltimore, MD: Johns Hopkins University Press.
Johnston, P. (1985). Understanding reading disability: A case study approach. Harvard Education
Review, 55, 153-177.
Johnston, P., & Winograd, P. (1983, December). Passive failure in reading. Paper presented at the
annual meeting of the National Reading conference, Austin, TX.
Langer, J. (1989). The process of understanding literature (Report Series 2.1). Albany, NY: University
at Albany, State University of New York, Center for the Learning and Teaching of Literature.
Langer, J. (1990). The process of understanding: Reading for literary and informative purposes. Re-
search in the Teaching of English, 24, 229-260.
Long, S. A., Winograd, P. N., Bridge, C. A. (1989). The effects of reader and text characteristics on
imagery reported during and after reading. Reading Research Quarterly, 24, 353-372.
Meichenbaum, D., & Asarnow, J. (1979). Cognitive-behavior modification and metacognitive develop-
ment: Implications for the classroom. In P. Kendall & S. Hollan (Eds.), Cognitive research and
procedures (pp. 11-35). New York: Academic Press.
Minturn, S. (1973). The house on the hill. In K. Kleiman (Ed.), Double action (pp. 64-71). New
York: Scholastic.
Moore, L. (1969). The real princess. In The ugly duckling and two other stories by Hans Christian
Anderson. New York: Scholastic.
Newell, G. (1984). Learning from writing in two content areas: A case study/protocol analysis. Research
in the Teaching of English, 18, 265-287.
North, S. M. (1987). The making of knowledge in composition: Portrait of an emerging field. Upper
Montclair, NJ: Boynton/Cook.
O'Brien, D. G., & Martin, M. A. (1988). Does figurative language present a unique comprehension
problem? Journal of Reading Behavior, 20, 63-88.
Ortony, A. (1979). Beyond literal similarity. Psychological Review, 86, 161-180.
Owens, R. A., Peterson, G. A., Bransford, J. D., Morris, C. D. & Stein, B. S. (1980). Spontaneous
monitoring and regulation of learning: A comparison of successful and less successful fifth graders.
Journal of Education Psychology, 72, 250-256.
Perfetti, C. A., & Lesgold, A. M. (1977). Discourse comprehension and sources of individual differ-
ences. In M. Just & P. Carpenter (Eds.), Cognitive processes in comprehension (pp. 141-183).
Hillsdale, NJ: Erlbaum.
Perfetti, C. A., & Lesgold, A. M. (1979). Coding and comprehension in skilled reading and implication
for instruction. In L. B. Resnick & P. A. Weaver (Eds.), Theory and practice of early reading
(Vol. 1, pp. 57-84). Hillsdale, NJ: Erlbaum.
Readence, J. E., Baldwin, R. S., Martin, M. A., & O'Brien, D. G. (1984). Metaphorical interpretation:
An investigation of the salience imbalance hypothesis. Journal of Educational Psychology, 76,
659-667.
Rosenblatt, L. (1978). The reader, the text, the poem. Carbondale, IL: Southern Illinois University
Press.
 On the Outside Looking In 253

Ryan, E. B. (1982). Identifying and remediating failures in reading comprehension: Toward an instruc-
tional approach for poor comprehenders. In T. G. Waller & G. E. MacKinnon (Eds.), Advances
in reading research (Vol. 3, pp. 223-261). New York: Academic Press.
Seife, C. L. (1984). The predrafting processes of four high and four low-apprehensive writers. Research
in the Teaching of English, 18, 45-64.
Smith, F. (1985). Reading without nonsense. New York, NY: Teachers College Press.
Vellutino, F. (1979). Dyslexia: Theory and research. Cambridge, MA: MIT Press.
Wagoner, S. A. (1983). Comprehension monitoring: What it is and what we know about it. Reading
Research Quarterly, 18, 328-346.
Woods, M. L., & Moe, A. J. (1989). Analytical reading inventory (4th ed.). Columbus, OH: Merrill.
Zecker, S.C., & Zinner, T. E. (1987). Semantic code deficit for reading disabled children on an auditory
lexical decision task. Journal of Reading Behavior, 19, 177-187.

AUTHOR NOTE

Preparation of this report was supported in part by the Center for the Learning
& Teaching of Literature, SUN Y Albany, through grant number G00870278, which
is cosponsored by the U.S. Department of Education, Office of Educational Re-
search and Improvement (OERI/ED), and by the National Endowment for the
ARTS (NEA).