“Nosotras no Empezamos a Hacer Eso”:
A Social Semiotic View of a Sheltered
Science Investigation
SUSAN BRITSCH
Purdue University
West Lafayette, Indiana, United States

This article traces the interaction of multimodal semiotic resources
within and across tasks constituting a science unit on surface tension
in an eighth-grade sheltered science classroom. Two discourses, operating simultaneously, realized differently motivated signs. This article
follows the contribution of linguistic and extralinguistic modes that
impinged on science and language learning. These semiotic
resources differently engaged the identities of the teacher and the
focal English learning students. The teacher discourse cued students
to read the available modes in a way that awarded them a behavioral
identity in the classroom, but not a conceptual one. The English
learners recognized and cooperated with the initial focus on behavioral compliance imposed by the teacher discourse while asserting
peer identity, but late in the unit the teacher discourse shifted frames
to require student enactment of conceptual identity through written
language. The teacher discourse foregrounded, or was consistently
undergirded by, the conceptual identity that the teacher brought to
the classroom and by a set of expectations for a learning trajectory
that did not explicitly link key language with students’ investigative
activity. In five of six task-based interactions, the student discourse
foregrounded behavioral or peer identity. As a result, the students
were able to only partially link their science activity to the concept of
surface tension using the language of science.
doi: 10.1002/tesq.546

“ ¡N

osotras no empezamos a hacer eso!” [We girls didn’t start any of
this!]
With this, Graciela, a Level 1 English learner,
defended herself and her lab partner, Beatriz. The teacher had assigned
pairs of students to count the number of paperclips they could suspend
on the surface of the water in a plastic cup. Salvador, also seated at this
lab table, had suddenly pushed all of the girls’ paperclips to the bottom
of their cup. Graciela punctuated her pronouncement about the girls’

310

TESOL QUARTERLY Vol. 0, No. 0, 0000
© 2019 TESOL International Association


innocence with a gesture: She pushed the boys’ paperclips to the bottom
of their cup as well. The teacher’s instructional aim, however, had been
to engage the students in a demonstration of the role of molecular
attraction in surface tension.
This article takes a multimodal, social semiotic view of the science
discourse that took place during one unit of study in an eighth-grade
sheltered science classroom, enrolling only English learners across
multiple levels of English language proficiency (ELP). The aim is to
demonstrate the ways in which shifting configurations of mode and
identity reflected the participants’ experiences of the discourse. The
article suggests that two discourses were actually operating simultaneously throughout the science unit, each enlisting the available semiotic
resources to realize differently motivated signs. As a result, the students partially linked their science activity to the concept of surface
tension, but because the teacher discourse explicitly foregrounded the
students’ identity in terms of their behavior, they did not develop a
conceptual identity either in English or in their first language (L1) in

the context of the science unit. This analysis explores the social semiotic mechanisms undergirding the pedagogical situation in which the
students were placed so as to understand why this outcome was
realized.

THEORETICAL PERSPECTIVE
Discourse participants make principled use of multiple modes of
communication as resources for enacting and representing meaning.
In her discussion of mode, Finnegan (2002) defines communication
as “essentially a purposive, organized and mutually recognizable process in which individuals actively interconnect with each other” (p.
47). On this view, the “modalities used for action and interaction” are
viewed as the resources through which communication takes place
(Finnegan, 2002, p. 48). Mode has subsequently been characterized as
“a socially and culturally shaped resource for making meaning” (Bezemer & Kress, 2008, p. 171). Discourse participants use modes such as
image, talk, writing, gesture, and artifact to formulate particular kinds
of interactions in particular ways (Kress & van Leeuwen, 2001). This
selection not only defines the social practice but also encodes “characteristics of the producer of the sign” (Kress, 1993, p. 172). Discourse
participants not only enact but also recognize identities or ways of
being that will “make visible to others who they are and what they are
doing” (Gee, 2011, p. 37). Thus, signs necessarily reflect an individual’s view of his or her place—his or her identity—in the world of that
discourse. This “motivated relation” (Kress, 1993, p. 73) means that
TESOL QUARTERLY

311


discourse participants may differently link form with meaning, realizing signs that reflect discrepant experiences of the discourse.
This article presents an analysis in which the modes used for meaning making within but not across discourses meet Finnegan’s criteria.
In other words, while using the same modes, the sign makers did not
necessarily interconnect or enact mutually recognizable processes. This
resulted, in part, from the fact that each discourse foregrounded certain semiotic resources to realize signs that drove the interaction while

backgrounding others that supported or rejected those same signs
(Britsch, 2005; Kress, Jewitt, Ogborn, & Tsatsarelis, 2001). This article
further suggests that, in much the same way, variations in identity may
be foregrounded or backgrounded to reflect the participants’ engagement at any given time. This intertwines mode and sign with identity.
On this view, identity becomes multidimensional as one definition of
“what we are doing” (Gee, 2011, p. 31) collides or resonates with
another. Those participants who synthesize the available modes in the
same way may enact or recognize one identity configuration (Britsch,
2005) while an alternative synthesis recognizes a different identity and
proceeds on that basis. As a result, modes may be “loaded with meanings” (Machin, 2007, p. 21) that are not mutually recognizable to discourse participants as resources for enacting mutually defined social
practices (Gee, 2011).

Framing
The perception that participants engaging in the same event may
experience it differently is not new (e.g., Kloesel, 1986; Schutz, 1945;
Tannen, 1993; Wittgenstein, 2009). In fact, both recognition and divergence are involved in Peirce’s conceptualization of the sign vehicle
and its object. On Peirce’s view, “A sign is an object which stands for
another to some mind” [italics added] (Kloesel, 1986, p. 66). Peirce did
not address the objective reality of the sign and suggested that the
object of a sign may be perceptible, imaginable, unimaginable, or even
fictive (N€
oth, 2001). Thus, a sign maker need only experience the
object in some capacity to interpret the sign (N€
oth, 2001). The nature
of that experience need not conform to an objective, or even existing,
reality. This allows for sign makers in a given context to perceive,
frame, construe, and redefine objects differently to other sign makers
and so to derive different interpretations. Peirce used the example of
a painted portrait as a particular rendering of a subject based on the
impression that the subject’s appearance had made on the artist’s

mind. As Peirce put it, “The one caused the other through the medium of the painter’s mind” (Kloesel, 1986, p. 67).
312

SHELTERED SCIENCE


This sort of semiotic subjectivity is also reflected in Wittgenstein’s
(2009) concept of the “language-game” (p. 8e) as a “more or less complicated sharable human activity which might, or might not, have a
utility which could be grasped and stated outside the game” (Kenny,
2006, p. 133). Language-games are played according to rules but are
experienced differently by someone who understands those rules than
they are by someone who does not (Wittgenstein, 1974). Wittgenstein
(1974) concluded that this knowledge causes a person to “have the
particular experience he has” (pp. 49–50). As Goffman (1974) has put
it, “When participant roles in an activity are differentiated—a common
circumstance—the view that one person has of what is going on is
likely to be quite different from that of another” (p. 8). Because different interests may generate different “motivational relevancies” (Goffman, 1974, p. 10), an activity cannot be reductively defined as unitary.
More than one “frame of reference” may be operative at any given
moment (Goffman, 1974, p. 10). Recognized social frameworks are
thus subject to the “serial management of consequentiality”—most
apparent when action is “unexpectedly blocked or deflected” (Goffman, 1974, p. 22). On this view, a participant’s line of activity may be
subordinated to that which “has come to be defined as the main
action” (Goffman, 1974, p. 201). Because a “main track” may co-occur
simultaneously with a “disattend track,” however, conflicting definitions of main actions as opposed to subordinate ones may hold at any
given moment (Goffman, 1974, p. 210). That which is main may even
give way to that which is subordinate, switching the track that drives
the situation as it progresses and then switching it back again.
This means that complete complementarity in an “intersubjectively
established, temporarily shared social world” depends on “a shared
strategy of categorization” across a flow of events that may be experienced visually, kinesthetically, and aurally (Rommetveit, 1974, p. 30).

Adding to this dimensionality, a participant’s understanding of that
which is taking place “requires reference to underlying strategies or
plans employed by both parties” (Levinson, 1979, p. 383). This assessment may be made differently—or may even be opaque—from the
point of view of a particular participant or group of participants in the
discourse. “Divergent strategies of categorization” may thus result in
recurrent partial intersubjectivity in everyday and scientific discourse
(Rommetveit, 1974, p. 39). This deixis navigates both linguistic and
extralinguistic features, including the situation, preceding utterances,
and the listener’s knowledge of both speaker and topic (Rommetveit,
1974). In sum, that which is said along with that which is used or discarded, acknowledged or ignored, depends on participant understanding—or reading—of the essential modes in the discourse in which

TESOL QUARTERLY

313


these are “embedded, over and beyond whatever meaning” words or
other resources may have “in vacuo” (Levinson, 1979, p. 367).

Framing in Science Classrooms
The activity in science classrooms has been addressed specifically in
terms of students’ epistemological framing of that which is taking
place (e.g., Berland & Hammer, 2012; Hutchinson & Hammer, 2010;
Scherr & Hammer, 2009). When students frame the same activity differently, their ideas are also used differently in the discourse; for
example, preservice elementary education majors who framed an activity as a classroom game in a physics class expected to receive instructor-provided information, but when they framed the activity in terms
of sense making about a natural phenomenon, they expected to produce and assess knowledge from their own experience, reasoning, or
schooling (Hutchinson & Hammer, 2010). Teacher cues may also
intentionally or unintentionally contribute to a frame shift if they resonate with prior activity that students have already framed in a particular way. Herbst and Kilpatrick (1999, p. 4) have noted, however, that if
an activity violates the conditions students expect of a frame, the “didactical contract” may be broken and students may respond in ways
that the teacher views as aberrant. This places student activity outside

the teacher’s frame, where it may remain if participant readings of the
discourse continue to conflict.
Moje, Collazo, Carrillo, and Marx (2001) have addressed student
framing of classroom science activity in terms of competing discourses
in which teachers and students often talk “across each other” because
relevant vocabulary words carry technical meanings even as they are
also “embedded in particular Discourses and funds of knowledge” (p.
478). As a result, everyday and scientific discourses may compete,
necessitating the construction of a third space for a new amalgamated
discourse. Differentiating between discourse types may, in fact, be
problematic for some students (Wallace, 2004). Learners “interpret
words and construct meanings that are compatible with at least some
aspect of a knowledge structure in long-term memory” to retain the
integrity of incoming information (Osborne & Wittrock, 1983, p. 496).
Thus, if students recognize a task as matching with a particular knowledge structure or as coincident with an everyday frame, their use of
the available semiotic resources will work in concert with that everyday
frame instead of the new frame. Further, if a frame shift occurs, not
all modes may work as productive semiotic resources in the new frame.
In fact, English learners may experience an additional burden in terms
of both receptive and productive language use if the new frame
314

SHELTERED SCIENCE


requires that knowledge be communicated and demonstrated primarily in the second language (L2) instead of via other modes.
Another perspective on framing has addressed learners’ subjectivities in terms of discourses and counter-discourses. Basing their findings on Bonny Norton Peirce’s (1995) view of motivation, McKay and
Wong (1996) have called for second language learners to be viewed as
“subjects with agency and the need to exercise it” (p. 603). English
learners “must be understood with reference to large and frequently

inequitable social structures that are reproduced in day-to-day social
interactions” (McKay &Wong, 1996, p. 579). McKay and Wong tracked
the negotiation of multiple and contradictory identities by four junior
high school–age Chinese-speaking immigrant students over a 2-year
time frame. While subject to power relations within discourses, these
learners resisted positioning, attempted repositioning, and constructed
counter-discourses to constantly assert their own agency. Similarly, Carhill, Suarez-Orozco, and Paez (2008) concluded that social context factors were central to English language learning experiences for 14- to
19-year-old English learners. At the adolescent level, proficiency in academic English was correlated with (a) opportunities for students to
speak English in informal settings and in school, (b) segregation and
poverty rates at school, and (c) the interplay between the skills students brought to the classroom and the linguistic and educational contexts in which they were placed. Such findings suggest that theoretical
models must capture “the active role of adolescent immigrant students
in constructing their own language environments” (Carhill et al., 2008,
p. 1175) as differentiated from theoretical models that address young
children who are English learners.
Adolescent classroom science discourse has also been explored
from the perspective of multimodal communication (e.g., Carolan,
2016; Goldberg, Welsh, & Enyedy, 2009; Kress et al., 2001; Mavers,
2011; McDermott & Hand, 2016; Richardson Bruna, 2009; Villanueva,
2016). This work confirms that student and teacher readings of a classroom discourse may differ and that these disparate readings may affect
perceived aims and outcomes. Richardson Bruna (2009), for example,
noted the tension between play and learning that emerged in a newcomer classroom when the teacher’s reading of the discourse focused
on what students lacked instead of on the linguistic biliteracy they possessed. Similarly, Villanueva (2016) concluded that although multimodal representation can mediate students’ lack of understanding,
this, too, may be misinterpreted as a language deficit. Disparate readings of classroom discourse are not, of course, based solely on participant readings of verbal modes but also on their readings of visual,
auditory, and kinesthetic modes. Thus, analysis of classroom discourse
that is weighted a priori toward the working of verbal modes not only
TESOL QUARTERLY

315



misses the role of nonverbal experience in the linguistic factuality of
words (Sapir, 1921) but also blinds the analysis to much of that which
constitutes teaching and learning. In contrast, a multimodal perspective, rooted in social semiotics, recognizes actions, objects, and environments (van Leeuwen, 2005) as they “achieve the social uses and
needs of producers” (Pirini, Matelau-Doherty, & Norris, 2018, p. 646).
The acknowledgment of this sort of motivated signifying unearths the
particular relationships between form and meaning that reflect the
“interest” (Kress, 2003, p. 43) of a sign maker in recruiting certain
modes as well that sign maker’s interpretation of the discourse and his
or her role in it. This perspective recognizes every discourse participant as a sign maker and assumes agency on the part of every participant in a classroom discourse. It is this agency that results in the
transformative foregrounding of certain modes and the distancing of
others.
In sheltered classrooms, language choice also works with other
modes as a semiotic resource to mark social identity. Language
choice makes “visible to others” who a participant is and what that
participant is doing (Gee, 2011, p. 37); for example, students may
use their L1 instead of their L2 because of its suitability for expressing an identity as it is recognized in a discourse. Language choice
thus intersects with the perception of self and the ways in which students wish to be viewed and understood (Ige, 2010). Bilingual speakers may switch codes as they switch discourses in a single setting
(Kantone, 2007). In this case, the most appropriate language choice
is also based on the preferred language of the interlocutor. Nonetheless, the use of a particular language in front of someone who does
not understand it is negatively marked for both bilingual and monolingual speakers (Cashman, 2008). In a sheltered classroom where
the teacher is a monolingual English speaker, the students’ use of
their L1 may violate this appropriateness rule to realize a counter
sign. For this reason, language choice may take a central role as discourses are framed.
This article combines the perspectives of discourse framing and
multimodal discourse analysis to trace the involvement of English
learners across the interactions that constituted a classroom science
discourse on surface tension. It does not attempt to detail the syntactic
development of students’ English use, but instead characterizes the
role of language as well as various extralinguistic modes to answer the
following research questions:


•

316

How did discourse enactment impact science learning and English use for the focal English learners throughout the unit on
surface tension?

SHELTERED SCIENCE


•
•

How did the semiotic resources available in the classroom
engage the identities of the teacher and the focal English learners throughout the unit on surface tension?
How did the identities of the discourse participants recognize or
respond to each other?

METHOD
This study followed four focal English learners and the classroom
teacher throughout a 3-day science investigation of surface tension.
This investigation is presented as an exemplar of the interactional patterns that took place across the semester during which the study took
place.

School Site and Participants
The focal school was located in a metropolitan area of approximately 70,000 residents in the midwestern United States. The school
had a diverse student population made up of approximately 60%
White, 12% African American, and 20% Latinx students. Approximately 65% of the students participated in the federal free/reduced
lunch program. The study was conducted in an eighth-grade sheltered

science classroom enrolling 24 students. Sheltered classrooms enroll
only English learners with the aim of providing instruction that simplifies language demands while addressing grade-level content (Wright,
2015). Spanish was the first language of all of the students in the focal
classroom, and all were classified by the school as English learners.
During this study, there was no state-mandated science assessment
at the eighth-grade level; however, data from the National Assessment
of Educational Progress (NAEP) chart the trends in science learning
for English learners in U.S. schools. In 2011, the average score on the
eighth-grade science assessment for English learners as a cross-language category was 106 as compared with 154 for non–English learners
(NAEP, 2011). In 2015, the average score increased to 110 as compared with an average score of 157 for non–English learners (NAEP,
2015a). In that same year, 71% of eighth-graders who were not English
learners scored at the basic level or higher on the NAEP science assessment, whereas only 19% of English learners scored at the basic level
or higher (NAEP, 2015b).
The teacher in the sheltered science classroom was White and a
monolingual English speaker. The teacher had completed a master’s
TESOL QUARTERLY

317


degree in science education and had been teaching science for 10
years. The teacher had volunteered to work with English learners at
the school and was provided with a 2-day training workshop based on
the Sheltered Instruction Observation Protocol (SIOP; Echevarrıa,
Vogt, & Short, 2008). This was conducted by an English as a second
language (ESL) coordinator from another school district. A follow-up
training session took place 5 months later. The training required
teachers to demonstrate the ability to plan and implement a content
area lesson using the SIOP model. The workshop provider also met
with teachers after school at least once per quarter to discuss lesson

plans. The SIOP-based background provided to the teacher was contributed solely by the workshop provider and not by the researchers.
The researchers did not interact with the training provider in any
capacity.
Two researchers acted as co-principal investigators for this study; I
specialized in English learning and the other co-principal investigator
specialized in science education. The researchers acted as participant
observers throughout the study. The teacher planned and carried out
all instruction. The researchers did not stipulate that the teacher was
to cover specific topics or that particular methods of instruction were
to be used. The researchers answered teacher questions if any were
asked following instruction, but the teacher was under no obligation
to consider any of the researchers’ responses in any respect. Thus, the
teacher planned and implemented all instruction throughout the
study.
Four students were selected as focal participants for this study, representing a range of ELP levels based on the standardized language
assessment instrument used by the school district. These students were
also deemed by the teacher to be socially compatible as learners. Identified by pseudonym, the focal students had ELP levels as shown in
Table 1.
In terms of speaking and writing, Level 1 English learners are able
to produce content-related words and phrases when provided with
visual and graphic support; in terms of listening and reading, Level 1

TABLE 1
English Language Proficiency Levels of the Four Focal Students

318

Student pseudonym

Speaking/writing level


Listening/reading level

Beatriz
Graciela
Humberto
Salvador

5/3
1/1
4/3
4/3

4/4
1/1
4/3
4/3

SHELTERED SCIENCE


learners comprehend everyday vocabulary and some content-related
words within simple grammatical structures, again when provided with
visual support (WIDA Consortium, 2012). In contrast, Level 3 learners
may produce simple as well as compound structures containing specific content-area language; they may comprehend ideas related to content areas in discourses that comprise a series of extended utterances.
Finally, Level 5 learners may produce complex grammatical structures
as well as technical and abstract language; they may comprehend
related ideas across content areas as well as technical language
couched within complex grammatical structures (WIDA Consortium,
2012).


Data Collection
Data were collected 4 days a week by video-recording the entire 50minute science class session. All written products were photocopied or
photographed, and all visual products were collected throughout the
semester for students who had submitted signed consent and assent
forms. All written and/or visual materials distributed by the teacher
were also collected. Intermittent informal interviews were conducted
with the teacher to obtain views about classroom processes that were
taking place. Both researchers compiled field notes during and after
each class session to describe activity that was occurring outside the
camera frame and to note relevant theoretical or methodological
points (Flewitt, 2006).

Data Analysis
Data analysis in multimodal discourse is centered on locating patterns (e.g., Flewitt, 2006; Jewitt, 2006; Plowman & Stephen, 2008) to
distinguish the ways in which individual and/or group choices and
actions may “constitute an enacted curriculum—a learning environment” (Erickson, 1986, p. 129). In this study, a process of qualitative
analysis based on a social semiotic view of discourse was used to examine the multimodal sign-making activity in the focal classroom (Bogdan
& Biklen, 1982; Gee, 2011; Hodge & Kress, 1988; Kress & van Leeuwen,
2001; van Leeuwen, 2005). Sign making was defined as the use of multiple modes by discourse participants to realize meaning: (a) oral language, (b) written language, (c) visual language (pictographs,
photographs, objects imaged by a projector), (d) gesture (facial or
manual expression), (e) direction of gaze, (f) full body movement,
and/or (g) artifacts (e.g., classroom and science materials).
TESOL QUARTERLY

319


The unit of analysis was the task-based interaction, defined as a segment of multimodal classroom activity in which a constant group of
participants engaged, either receptively or productively, with a particular set of semiotic resources. Coding categories were derived to characterize the multimodal activity within and across task-based interactions

in which discourse participants engaged via linguistic and extralinguistic modes (Kress, Leite-Garcıa, & van Leeuwen, 1997). These coding
categories were applied to the data to characterize the nature of the
acts that characterized the practice of science investigation within and
across the 3 days of the surface tension unit. Appendix A identifies
and defines the types of multimodal actions that occurred across the
corpus.
Because the study took place in a classroom serving English learners
exclusively, verbal language might be assumed to take on the most
central role. In fact, language is frequently backgrounded to visually
and kinesthetically accessible modes, especially in science classrooms
(e.g., Arnheim, 1969, 1980; Kress, 2003; Kress et al., 2001; van Leeuwen, 2005). For this reason, this analysis uses discourse visualizations
to exemplify all essential modes that operated as mediators within critical patterns of activity (Britsch, 2009). Each task-based interaction thus
encodes a recurring critical pattern. The aim of the visualization is to
demonstrate the simultaneous working of essential modes across each
pattern. A visualization is not a transcript but a multimodal analysis,
summarizing a critical pattern that characterizes, or typifies, a taskbased interaction. Essential modes are defined here as those that carry
the weight of social meaning in a task-based interaction. Critical patterns are representative interactional patterns that recur or that shift
the social or conceptual focus of the task-based interaction. As analytic
graphics, then, discourse visualizations do not report moment-to-moment activity but configure “the essential elements of the event, deliberately leaving out the rest” (Norman, 1993, p. 49).
In each visualization, extralinguistic modes are indicated by pictographs and human participants are indicated by labeled rectangles.
Oral language is transcribed in callouts: Rectangular callouts indicate
student utterances and curvilinear callouts indicate teacher utterances.
Text in bolded italics indicates an English translation of a Spanish
utterance. (A full key is provided in Appendix B.) Each visualization
separates teacher and student discourses into zones. The top zone of
the visualization contains the foregrounded semiotic resources that
drive the interaction, dominating it in a social or organizational way.
In contrast, the lower zone contains those participants and backgrounded resources that support, respond to, or contradict the activity
shown in the top zone (Britsch, 2009). The occupants of a zone are,
however, transitional; that is, the semiotic resources at work in the

320

SHELTERED SCIENCE


teacher discourse may drive one task-based interaction, occupying the
upper zone, but this dynamic may be reversed in the next task-based
interaction. In this case, the student discourse would move into the
top zone, dominating the interaction. Finally, each visualization separates the zones with a broken line if the boundary between them is
permeable (i.e., if the teacher and students participate in, affect, or
respond to the activity in the other discourse). If, on the other hand,
the two discourses function independently and do not interconnect,
the boundary is indicated by a solid line. Participants and modes thus
work in concert with the structure of a discourse to realize its signs
(Rose, 2016).

FINDINGS
This article asserts that the focal students and the teacher experienced the surface tension investigation quite differently, reframing the
discourse in ways that foregrounded and backgrounded different semiotic resources to reflect multidimensional identities. The following sections explore the dynamics of these enacted reframings, beginning
with a description of the teacher’s process for frontloading the key
science vocabulary for the investigation.

Reframing 1: Vocabulary as Behavior
To begin the period, the teacher distributed a gridded worksheet
listing 12 “water words” with cells for students to fill in the “definition,” and a pictographic “diagram” for each key vocabulary item. A
duplicate worksheet was displayed on the ELMOâ projector. The teacher’s key for this word list is shown in Figure 1. The teacher then
directed the students as follows:
T: Today we’ll be working on some surface tension experiments. If you
do not know what this is, you will know by the end of the hour. I don’t
see everybody’s eyes looking up here. For homework I would like you

to practice the vocabulary words one through six. Please write that
down in your assignment notebooks.

The teacher also told the students, “You guys will have a paragraph
to write at the end of our experimenting today.” This alerted the students to the fact that they would need to write about the day’s investigation but did not specify a purpose for or a link to the worksheet
that had been distributed. The teacher then presented a video on the
properties of water. This reiterated elements of a previous unit on
TESOL QUARTERLY

321


FIGURE 1. Teacher’s key vocabulary list for students to copy

sinking and floating and introduced the idea that an elastic skin is
caused by surface tension so that some insects are able to walk on
water. All of the focal students attended to the video. The teacher
replayed the segment explaining the elastic skin twice and began an
oral elaboration that referenced both the video clip and the worksheet:
T: Does everybody have this piece of paper?
[Teacher holds up the worksheet and replaces it on the ELMO.]
322

SHELTERED SCIENCE


T: See the bug sticking on there? We’ll be looking at that idea today.
Okay, why does that bug not sink in the water? Do you remember when
we did all those floating and sinking experiments where we put things
in the water and some of them sunk? Why do you think that bug

didn’t sink like some of the other things we saw?

This open-ended questioning visually referenced the students’ previous experience, which had also involved water, and this seemed to
relate the reasons for sinking and floating to surface tension; however,
the link between the air bubbles the students had observed as objects
sank and surface tension was unclear. Nonetheless, Humberto twice
offered explanations based on recall of the prior unit:
H: Maybe the shape of the leg?
T: Okay the shape of the legs. We talked about that being important so
we’re gonna explore that a little bit more. [Humberto raises his hand
again.] Humberto?
H: It’s light?
T: Okay, that could be a reason. These words are things that we’ve
been kind of talking about. We’re going to talk about some of them.
I’m going to show you a story about water—people actually write books
about things like water—where some of these things are quite important. On this paper, please write your name right now. Okay, your
name and your group name, and we’re gonna look at some words
before we start, okay? One of the words that came up last week was the
word particle, okay? A particle. I think this group used the word particle.
What did you guys mean by particle? [Teacher stands by Graciela.]
B: Small piece.
[S. is trying to peel off the masking tape that holds down the microphone cord. B. taps his hand with her pencil.]
T: A small piece, okay. A small piece. So if we said this table is a large
slice of our class, X. here would be a particle. She would be a small
piece. Beatriz would be another particle so a particle is a piece. Let’s
look at a diagram of a part. [Teacher moves to the ELMO and begins
to write and draw on the worksheet that is displayed.]
H: Do we write—
T: Okay. Yeah. Write it down. What’s another word for piece?


TESOL QUARTERLY

323


B: Object.
S: Particle. [He laughs as he looks at B., who also laughs.]

The teacher’s responses to Humberto did not conceptually address
the explanations he had offered, nor did they address the image from
the video. Instead, the teacher’s questions introduced the procedure
for filling out the worksheet. This foregrounded the students’ behavior
while backgrounding Humberto’s ideas about the possible role of
shape and weight in surface tension. The teacher’s oral elaboration
did reference the students’ actions during the float-and-sink investigation but did not address the fact that surface area instead of shape is
relevant to surface tension. This created a misconception that would
recur on the third day of the unit. Neither did the teacher’s elaboration offer students a chance to elaborate or to question the visual or
verbal science content. The task-based interaction thus moved from
the teacher’s initial questioning to the directions and then to the first
word on the worksheet (i.e., particle), leaving unaddressed Beatriz’s
misinterpretation of “particle” as “object.” This established the students’ behavioral role in the task while the teacher held all of the conceptual cards. As the worksheet task progressed, teacher questioning
became increasingly closed-ended, allowing even students at higher
ELP levels (such as Beatriz and Salvador) to respond with only single
words or brief phrases based on recall or restatement as they copied
whatever was written or drawn on the sample worksheet.
In terms of layout, the position of the ELMO projector itself added
a further mode for Humberto and Salvador because they needed to
turn around completely in their seats to view the teacher’s model. This
meant that they constantly shifted their attention from the teacher to
each other to their worksheets as the teacher faced the ELMO to complete the model; for example, as the teacher defined the word molecule,

Salvador kicked Beatriz under the table and passed her Humberto’s
paper. Because Humberto was known to accomplish tasks successfully,
Salvador’s teasing gesture implied that Beatriz should copy Humberto’s work, but she pushed the paper away and all four students
laughed.
The task that faced the students was to induce a connection
between the words on the model list and the investigative activity that
would follow. At the moment, however, the rules of game that they
were playing simply required them to reproduce new and decontextualized written and visual content. Thus, although the student discourse
in L2 enlisted (a) receptive written language, (b) receptive visual language via the video and “diagrams,” and (c) productive but limited
oral language, the teacher discourse enlisted (a) productive oral
324

SHELTERED SCIENCE


language to define and relate the words to a prior unit, (b) productive
written language for brief definitions of the first six terms (i.e., particle,
molecule, formula, hydrogen, oxygen, attraction), and (c) productive visual
language via the diagrams shown in Figure 1. The fact that the teacher’s use of productive language far exceeded the students’ opportunities to do so began to place the teacher in a zone apart from them.
This separation was reified by the teacher’s existing conceptual understanding of the inductive task at hand, its unarticulated relationship to
the investigation that would follow, and the outcome: a written explanation of the concept of surface tension. Despite the fact that the students had been told they would have to write some type of paragraph
at the end, this distribution of receptive versus productive engagement
awarded the teacher a conceptual role in the task-based interaction
but did not begin to build one for the students.
Reaching attraction, the sixth word on this list, the teacher removed
the model worksheet from the ELMO and began to display photographs from a trade book about water and surface tension, reading
aloud and pointing to each word (Wick, 1997). The teacher then elaborated on the text orally but without writing out a definition; for
example, to discuss weak and strong bonds the teacher said,
T: These oxygen have kind of a negative charge and these hydrogen
have a little positive charge. So what happens is . . . the negatives are

sticking to the positives, okay . . . and we get like a little bond there.
It’s not as strong as the bond between the molecules, but it’s a little
bond—a little stickiness. It’s like you and your friends, okay—you stick
together. And that’s what water is like.

This drew an analogy with the students’ everyday experience primarily via oral language. The teacher addressed the next four words,
including the term elastic skin, in the same way—without a written or
pictographic definition. Although the photographs in the book provided visual description, this process deprived Graciela, who had
arrived late in the school year, of an essential mode. She would write
on her worksheet only when the teacher was also writing on the
model, not while the teacher was reading or orally discussing content.
Graciela relied on Beatriz to quietly translate if the teacher read aloud.
Beatriz also often supported Graciela’s attempts to comprehend teacher talk, including the directions for tasks; for example, when the teacher began to describe the sequence of activities for the period, the
terms before and after were used. Graciela wrote “before antes” and then
“alfe” in her notebook, separate from her worksheet. Beatriz erased
this attempt at spelling “after” and substituted “after despues,” at
which point they both smiled and giggled. Language choice was thus a
key semiotic resource for Graciela as she devised attempts to
TESOL QUARTERLY

325


participate. Although the teacher’s presentation of the key vocabulary
enlisted visual imagery, student comprehension of the photographs
and pictographs as related to surface tension depended on understanding the written and oral elaboration. This effectively excluded
Graciela as an interactional partner in the teacher discourse.
In interview data the teacher noted a concern that the students
know what to do, but also reported that their task was to induce a connection between the key vocabulary and the subsequent science activity
involving materials. Thus, from the vantage point of the teacher’s role

in the discourse, copying the words constituted a conceptual beginning to the investigation, but this was not communicated to the students through either verbally explicit or modeled instructions. The
students read the semiotics of this task-based interaction in terms of
behavioral compliance by watching the teacher and reproducing the
displayed language. This was consistent with the teacher’s warning 3
days earlier that any student who talked or did not look when the teacher spoke would be penalized with detention, but this did little to
carve out a conceptual identity in the discourse for the students.
Figure 2 summarizes the critical pattern that characterized this taskbased interaction. The teacher discourse drove the task through oral
elaboration. The students acknowledged the teacher discourse by
cooperating with its behavioral requirements; however, openings did
exist to contribute limited talk. Although the process was mutually

FIGURE 2. Vocabulary as behavior.
326

SHELTERED SCIENCE


recognizable to all participants, mutual engagement was impeded by a
difference in underlying plans (Levinson, 1979) and by an inequitable
distribution of productive versus receptive roles. One expectation may
have been that the students’ later investigative activity would verify the
science. For the moment, however, the copied pictographs remained
equally as context-free as the copied print definitions. As a result, the
discourse was bifurcated; this is indicated in Figure 2 by a broken line
instead of a solid one because permeability existed. Nonetheless, the
basis for this permeability also threw into question its effectiveness in
initiating a conceptual identity for the students in the investigation.
The students, however, began to reframe the discourse to foreground
their own agency.


Reframing 2: Data Collection as Competition
The teacher next directed the students to suspend as many paperclips as possible on the surface of the water in a plastic cup in order
to observe the effects of surface tension. Beatriz and Graciela worked
together, as did Humberto and Salvador. To introduce this procedure,
the teacher again contrasted the results of the previous float-and-sink
unit with aim of the new investigation but did not invoke any of the
terms from the vocabulary list.
T: Do you remember when we did the floating and sinking experiment? What happened?
B, H, S: They sink.
T: They SUNK, right. This time we’re gonna see if we can get these
paperclips to float and we’re gonna make a prediction. . . . How many
paperclips do you think your team can get to sit on the surface of the
water? . . . Do you think zero or do you think—
B: A certain number.
T: How many do you think you are going to get?
B: Okay four.
S: TWENTY-FIVE!

The teacher’s directions had identified the students as “teams” but
not as science investigators. This was confirmed by the seating

TESOL QUARTERLY

327


arrangement: Beatriz and Graciela on one side of the lab table and
Humberto and Salvador on the other. Further, each pair of students
was provided with a single set of materials. This created a “spatial link”
(van Leeuwen, 2005, p. 228) between the members of each pair. Coupled with the fact that the students’ forward direction of gaze was

toward each other, their peer identity as teams opposing each other
was reified. The teacher’s question about quantity also verbally situated
the task as a competition to achieve a score. As a result, the students
used the science materials and talk as essential modes to carry out the
activity as a contest, not as data collection to observe results. Their subsequent interaction foregrounded dexterous technique with the artifacts; for example, Humberto described the best way to pick up the
paperclips: “Mira, ag
a rralo ası —derecho—mete los dedos ası .” [Look, grab it
like this—straight—then put your fingers like this.] This legitimized Spanish
as an essential mode for participation in the student discourse (Goldberg et al., 2009), but not in the teacher discourse. During the task,
teacher talk did not link either the artifacts or the students’ activity
with the vocabulary on the worksheet or with language functions
related to guided inquiry (Klentschy, 2010). As a result, the students
again complied with the behavioral requirements of the teacher discourse, but its underlying conceptual agenda was neither communicated nor recognized.
For the task itself, each team member used only a small part of the
table layout: the cup and paperclips located between the two team
members. As the students’ activity continued, Salvador suddenly
expanded the boys’ task-based space to push the girls’ paperclips to
the bottom of their cup. This provoked a protest by Beatriz, Graciela,
and Humberto—the girls telling Salvador, “NO!” and Humberto warning, “No manches, Salvador.” [Don’t do that, Salvador.]. In response, Gra IDIOTA
ciela looked directly at Salvador and delivered an insult: “¡EL

ERES TU !” [You are the idiot!]. This was the loudest she had spoken
that day. A string of traded insults followed, culminating in Graciela’s
 fue la que empezό!
use of the feminine “la” to refer to Salvador: “¡EL
Dile que se calme.” [HE’s the one (feminine) who started it! Tell her to chill
out.] She added a denial of the girls’ responsibility in the incident:
“¡Nosotras no empezamos a hacer eso!” [We girls didn’t start any of this!].
She punctuated this with a payback move that invaded the boys’ taskbased space: She pushed all of their paperclips to the bottom of their
cup. Beatriz later commended Graciela: “Graciela, tu tienes buena experiencia” [Graciela, you have good experience (apparently on how to handle

circumstances like this).] This foregrounded talk as an essential mode
for participation in the student discourse, but this was talk that intensified the oppositional nature of the students’ peer identities via a competition of insults, carried out through gesture and artifact.
328

SHELTERED SCIENCE


As this conflict took place, the teacher moved around the room
observing other students. In interview data, the teacher reported that
the focal students were “concentrating fairly well” and seemed to “talk
quietly and play,” but still knew what to do. The teacher characterized
this behavior as bothersome but felt the students could multitask well.
In fact, the student discourse had repurposed the essential modes: the
adversarial use of gesture with the materials now triggered oppositional social talk. Although the teacher’s movement around the room
was meant to monitor all of the students’ observations, each successive
stop narrowed the focus to the materials on one lab table at a time.
When the teacher did arrive at the focal table, Beatriz complained,
“Salvador—Salvador put his finger and he pushed all the way down!”
While looking at the boys’ materials, the teacher responded with a single Spanish word: “Ah—BUENO.” Socially, this use of bueno seemed to
function as would an English evaluation such as “good” (instead of
Spanish bien or buen trabajo). To a Spanish speaker, however, “bueno”
might simply mean “okay then.” Uttered while the teacher’s gaze was
directed toward the boys’ materials, this seemed to validate their compliance with the teacher discourse while overlooking Beatriz’s complaint. The teacher moved away without responding to her. Thus, a
single word coupled with directionality of gaze encoded differential
social meanings for student and teacher (LoCastro, 2012): A dismissal
in the student discourse worked as an indirect imperative in the teacher discourse (i.e., to get on with the activity).
For this reason, the critical pattern shown in Figure 3 places the teacher in a zone both physically and socially distal to the students. A
solid line cuts across the visualization, indicating the low degree of
permeability that remained between the discourses. This time, it was
the student discourse that drove the task-based interaction via gestural

competition and oral insult. The physical movement that characterized
the teacher discourse was set against that dynamic assertion of peer
identity. As the insults between Salvador and Graciela escalated, however, Beatriz and Humberto resumed their behavioral identities outside the zone of the student discourse.
The teacher next moved to the front of the room and directed the
students to tally the number of paperclips suspended in each cup. Graciela counted the girls’ paperclips and Beatriz announced, “Okay,
four.” Salvador counted in Spanish, “21, 22, 23, 24 . . . 32” and then
announced twice in English, “WE HAVE 38.” This created student
engagement but did not link the tally with the reason for it. The students, of course, had carried out the paperclip activity as a social contest and an assertion of peer identity. The investigation had thus
backgrounded but had not completely excluded the behavioral identity mandated by the teacher discourse. After giving directions for
TESOL QUARTERLY

329


FIGURE 3. Data collection as competition

cleanup, the teacher announced that, after all, the students would not
be assigned to write the paragraph that had been referenced at the
start of the period.
T: We’re gonna talk some more about it in the morning before we
write that. You do have six words. These six words I want you to practice tonight. Okay, that is your homework.

The definition of “practice” was unclear because the paperclip task
had not involved the use of the key vocabulary by either the students
or the teacher. This meant that the students’ homework was to memorize the first six words on the worksheet and their still decontextualized written/visual definitions.

Reframing 3: Discourse as Disengagement
The next day’s class began with a review of the key vocabulary from
the word list shown in Figure 1. The teacher placed the word list on
the ELMO and asked for volunteers to explain individual terms in

their own words; for example, a nonfocal student defined molecule by
referring to the pictograph on the worksheet as “that thing with those
round things on it.”
330

SHELTERED SCIENCE


T: Tell us a little more about that. What do you mean by that?
Student: [inaudible]
T: Okay, so they’re so small that you can’t even see them with a microscope. Is that what you mean? Okay, they’re little pieces.
[The teacher continued with the word formula, and Beatriz answered.]
B: It could be like an abbreviation.
T: Okay, that’s another way to say it. The formula symbolizes another
word for an abbreviation—a shortcut—a nickname sometimes we might
say.
H: H is the—like the smallest atom there is in the water.
S: Those are the little balls that someone said—those little things on
here. [He points to the drawing on the model worksheet.]
T: Graciela, do you agree?
G: No, I don’t.
T (to G): Do you agree with Salvador?
B: Yup, we do.
G: Mm-hm.

This questioning invited student elaboration, but the link to surface
tension via the paperclip activity was still unclear. Although the teacher’s question to Graciela required a yes/no response, appropriate to
Level 1, it assumed that Graciela had understood both the definition
of formula and Salvador’s comment. When the teacher asked a second
time, Beatriz took the cue and agreed, speaking up for Graciela. This

helped Graciela to cope with the moment, but not to comprehend the
word formula or what it had to do with the paperclips. Teacher questions continued to invite student response, but with only the definition
itself as a context the students often complied by describing pictographs from the worksheet, by drawing on everyday experience, or
by saying they didn’t know. For example, when the teacher asked why
oxygen is important to water molecules, a nonfocal student responded,

TESOL QUARTERLY

331


“You breathe it?” Another student answered, “The molecules are different.” Humberto and a nonfocal student, however, addressed the idea
of “stickiness”:
T: X. says that oxygen gives it part of the stickiness. See this sticky part
right here? And oxygen is that sticky part that sticks here. Oxygen is
that big part of a water molecule.
H: They attract together. [He bumps two fists together.]

Humberto’s answer did relate molecules to attraction, but other
students’ understandings about the relationship of individual key
vocabulary words to surface tension were unclear. The teacher
addressed this issue in the next task-based interaction by directing
the students to repeat the paperclip task with an enhancement: A
single hand lens was distributed to each table. This time, the students protested to each other in English, loudly enough for the teacher to hear:
S: We’re not doing this AGAIN!
B: We’re gonna do it AGAIN?

In fact, the activity was procedurally nearly identical to the previous
day’s in terms of its essential modes: the science materials, the
required use of these, and the spatial links at the lab table. The hand

lens, however, injected a slightly different visual objective: The students were to observe an elastic skin between the suspended paperclips, thus invoking the term elastic skin from the word list. The
teacher recognized that the students had not linked the paperclip task
to the science concept and as a bridge interjected a series of oral
imperatives along with a description of what the students should be
able to see by using the hand lens; for example, “Look really closely at
the water and there’s almost a skin you can see in between the paperclips. Do you see?” The terms molecule and surface tension, both of
which were relevant, were not used, however.
Humberto took the hand lens first and agreed that he could see
the elastic skin: “Oh yeah, I can see it right here.” This complied with
the behavioral and conceptual requirements of the teacher discourse,
but when Salvador took the lens, he immediately concluded that there
was no difference to the day before. “It’s the same thing!” he said. In
fact, molecules cannot be observed with a hand lens; the students
could only see whether the paperclips floated or not. This made it
difficult to create a conceptual link to the process of molecular
attraction.
332

SHELTERED SCIENCE


To address this, the teacher returned to the ELMO projector and
displayed a set-up that duplicated the students’ materials. This taskbased interaction transducted modes from the student discourse to
the teacher discourse as the teacher added an elaborated a description
of “elastic skin,” including the key vocabulary word molecules:
T: Can you see how the water is in between the little pieces of the
paperclips, okay? That is all of the molecules are sticking and they’re
making an elastic skin. What does “elastic” mean?
B and S: Stretchable.
T: Stretchy. So that skin there is stretchy and flexible.

S: Beatriz, pay attention!
B: I AM!

Teacher questioning had attempted to reinforce the science concept but the molecules were, in fact, still not visible to the students. As
a result, they read the activity as mere repetition. The teacher then initiated another task-based interaction that related what had happened
with the students’ paperclips to similar examples in the trade book.
This engaged the teacher in using image and language to link observation with concept, but not the students. Teacher talk also included
closed-ended questions to clarify vocabulary in the text. These were
often correctly answered by Beatriz with single-word or phrasal
responses, but this assumed that the other students could integrate
her paraphrases into their comprehension of surface tension. As with
the first day’s word list, the teacher’s position at the ELMO and forward or downward direction of gaze narrowed the layout being utilized
to deliver the elaboration (Norris, 2004). This added to the vanishing
permeability between the two discourses. As a result, the students disengaged and reframed their discourse via verbally based peer interaction in both L1 and L2: Salvador began to tease Beatriz about paying
attention and she responded sarcastically; Graciela told a story in
Spanish about a robbery; finally, the students began a discussion in
Spanish about the homework for another class. From the perspective
of the teacher discourse, however, the science materials had provided
indirect evidence of the tension between water molecules. The images
in the book, in fact, clarified this. Thus, oral elaboration and orally
read written language worked as essential modes in the teacher discourse. Although this also referenced visual imagery, comprehension
of the verbal language was central to relating the photographs to the

TESOL QUARTERLY

333


concept of surface tension. Beatriz did not translate for Graciela this
time so, in the student discourse, teacher talk was superseded by competing L1 conversations even as the teacher continued to elaborate. In

the end, Salvador repurposed his pencil for a teasing fight with Beatriz
while the hand lens, cups, and paperclips were left unused. Although
Humberto attended to the teacher and asserted a conceptual identity,
the teacher discourse had not given the students a role in productively
using the key vocabulary with their own observations to explain surface
tension. Information had been presented to them but not scaffolded
with them.
Figure 4 summarizes the interactional pattern that characterized
this task-based interaction. The teacher discourse dominated through
a preponderance of oral elaboration expressing the teacher’s conceptual identity in the discourse. Students, such as Beatriz, contributed
occasional phrasal responses to teacher questions. Teacher elaboration
was countered, however, by the students’ L1 elaborations (e.g., Graciela’s recount of a robbery) and by teasing (e.g., Salvador’s pencil
fight with Beatriz), asserting peer identity. The students’ behavioral
identity was thus pushed to the background. The solid line bisecting
the visualization indicates that the two discourses no longer functioned
together.

FIGURE 4. Discourse as disengagement
334

SHELTERED SCIENCE