Gui A

P1: KNPCUUS443-07 cuus443/Oslon ISBN: 978 0 521 86220 2 Top: 0.5in Gutter: 0.875in October 27, 2008 8:52

CHAPTER 7

The Challenge of Academic Language

Catherine E. Snow and Paola Uccelli 1

Increasingly in recent years, educators haverelated worries about students’ literacyaccomplishments to their lack of “academiclanguage skills” (August & Shanahan, 2006;Halliday & Martin, 1993; Pilgreen, 2006;Schleppegrell & Colombi, 2002). Indeed, itseems clear that control over academic lan-guage is a requirement for success withchallenging literacy tasks, such as readingtextbooks or writing research papers andliterature reviews. As early as the middle-elementary grades, students are expectedto learn new information from content-area texts, so failure to understand theacademic language of those texts can be aserious obstacle in their accessing informa-tion. Accountability assessments requiringwritten essays in persuasive or analytic gen-res are often graded using criteria that referimplicitly to academic-language forms. Evenin the primary grades, students are expectedin some classrooms to abide by rules for“accountable talk” (Michaels & O’Connor,2002 which specify features encompassed inthe term academic language.

Despite the frequent invocations of “aca-demic language” and the widespread con-

cern about its inadequate development,there is no simple definition of what aca-demic language is. What we consider “aca-demic language” in this chapter is referredto in the literature using a variety ofterms: the language of education (Halliday,1994); the language of school, the languageof schooling, the language that reflects school-ing (Schleppegrell, 2001); advanced liter-acy (Colombi & Schleppegrell, 2002); scien-tific language (Halliday & Martin, 1993); or,more specifically, academic English (Bailey,2007; Scarcella, 2003). As suggested by theseterms, one approach to characterizing aca-demic language is to resort to the contextsfor its use – the language used in school,in writing, in public, in formal settings (seeTable 7.1 for a more complete list). Thus,for example, Scarcella (2003) defines aca-demic English as “a variety or register ofEnglish used in professional books and char-acterized by the linguistic features asso-ciated with academic disciplines” (p. 9).Similarly, Chamot and O’Malley (1994)identify it with school, defining it as “the lan-guage that is used by teachers and studentsfor the purpose of acquiring new knowledge

112


THE CHALLENGE OF ACADEMIC LANGUAGE 113

Table 7.1. Contextual Factors

Audience

(Home, friends) (School) (College/professional)Real familiar vs. Pretended distant vs. Alternative communities

cooperative uncooperative with various levels ofinterlocutor interlocutors disciplinary knowledge(assess interlocutor’s (suspend assumptions (become familiar withshared knowledge) of situational knowledge) expectations of audience)

Dialogic/interactive vs. Monologic vs. Delayed dialogue

Activity/Modality

Spontaneous/improvised vs. Highly plannedProcess (dynamic) vs. Product (synoptic)Spoken vs. Written vs. Other additional media

SituationPrivate vs. PublicInformal vs. Formal

Sociocultural match of Primary and Secondary discourses

Closer match vs. Partial mismatch vs. Full mismatch(e.g., home and school)

�same language�similar discourse patterns

(e.g., home and school)�different language�different discourse patterns

and skills . . . imparting new information,describing abstract ideas, and developingstudents’ conceptual understanding” (p. 40).

Whereas identifying contexts of use andpurposes is important, a comprehensive def-inition of academic language requires fur-ther specification. Scarcella (2003) identifiesthree dimensions required for academic-language proficiency: linguistic, cogni-tive, and sociocultural/psychological. Baileydefines being academically proficient as“knowing and being able to use general andcontent-specific vocabulary, specialized orcomplex grammatical structures, and mul-tifarious language functions and discoursestructures – all for the purpose of acquiringnew knowledge and skills, interacting abouta topic, or imparting information to others”(Bailey, 2007, pp. 10–11; in press).

Despite these advances in delineatingacademic language, a conceptualization ofacademic language within a consensual ana-lytical framework that could guide educa-tionally relevant research is still lacking.

Indeed, this topic, which seems as if itshould be located in the exact center ofeducators’ concerns, is notably absent fromthe table of contents in the most up-to-datehandbook of educational linguistics (Spolsky& Hult, 2007). Ironically, although academiclanguage skills are widely cited as the obsta-cle to achievement for struggling readersin general, much of the empirical researchon academic language has been done bythose studying English Language Learners(ELLs). In other words, learning ‘academicEnglish’ is recognized as a challenging taskfor second-language speakers of English, butthe challenges faced by native speakers inlearning the rules, the structures, and thecontent of academic English have receivedmuch less attention.

One line of thinking about academiclanguage started with Cummins’ propos-ed distinction between Cognitive Aca-demic Language Proficiency (CALP) andBasic Interpersonal Communicative Skill(BICS) – a distinction he presented as


114 CATHERINE E. SNOW AND PAOLA UCCELLI

relevant to second-language learners. InCummins’ original formulation of this dis-tinction (Cummins, 1980, 1981), BICS waspresented as easy and relatively automat-ically acquired, whereas acquiring CALPwas seen as a lengthier process. Cumminswas the first to point out cogently thatmany assessments of second-language pro-ficiency focus exclusively on BICS yet areused to place students in classroom con-texts where CALP is required for success(see Kieffer, Lesaux, & Snow, 2008, for anupdated version of this argument as it relatesto the testing requirements of the U.S. NoChild Left Behind Act). Although Cum-mins’ work was crucial in raising aware-ness of the gulf between conversational andacademic language, he did not specify inmuch detail which particular language skillswere encompassed by CALP, in either hisoriginal discussions of it or later, somewhatmore elaborated formulations (Cummins,1984, 2001).

A more theory-based approach that hascontributed centrally to our understandingof language, in general, and of academiclanguage, in particular, is Systemic Func-tional Linguistics (SFL) (Halliday, 1994b).SFL studies language in its social context,understanding language as both shaping andresulting from social circumstances. Withinthis framework, linguists search for system-atic relationships between the social con-text and linguistic features, including lex-icogrammatical and discourse elements, intheir analysis of the registers of particu-lar genres.2 In studying academic language,Halliday (1993) emphasized its multidimen-sional and dynamic nature. On the one hand,he warned us that there is no single aca-demic language, just as there is no singleBritish English, but rather a number of vari-eties that share certain core features. On theother hand, Halliday highlighted that aca-demic language is continually evolving asthe sciences, disciplines, and subdisciplinesthemselves evolve. In fact, he argued thatthe evolution of science goes hand in handwith the evolution of scientific language,so that academic or scientific languages arenot arbitrary sets of conventions but rather

grammatical resources that make scientificthought possible.

Although SFL has proven to be highlyrelevant in studying the language of school(Schleppegrell, 2001), it is a linguisticallysophisticated model originally designedmore as a theory of language than as a frame-work for educational research. An educa-tionally relevant framework would directless attention to the description of linguisticfeatures per se and more to the skillsrequired in the process of mastering aca-demic language and, thus, potentially to thenature of instruction that would promotethose skills. In other words, we argue forthe value of practice-embedded approachesto thinking about academic language thatwould generate more directly useable infor-mation. For example, Bailey (2007) derivedvaluable data about academic-language de-mands on ELLs from an analysis of contentstandards, classroom discourse, and thetests they are expected to pass. Scarcella(2003) and Schleppegrell (2001) also adum-brated the nature of academic languageby describing the typical failures of ELLswho have advanced conversational skillsbut who struggle with high school or uni-versity writing tasks and by proposinginstructional approaches to improving theacademic-language skills of ELL students intertiary-education settings.

Although the problem of academic lan-guage may be particularly visible or acutefor second-language speakers, in fact, weargue that academic language is intrinsicallymore difficult than other language regis-ters and that thinking about the educationalexperiences that promote its developmentis a crucial task for educators of all stu-dents. Furthermore, formulating instruc-tional approaches to academic language isnecessary not just for achievement in thedomains traditionally associated with lan-guage (e.g., literature study, English lan-guage arts) but also for achievement inmath, science, and other areas where all-purpose academic language forms the coreof content-area–specific language. Designinginstruction for academic and discipline-specific language, however, requires having



a convergent view of what academic lan-guage involves, how it should be concep-tualized, where its boundaries are, and howit might be assessed.

Certainly, members of the academy canidentify violations of academic languagein our students’ writing and may havelearned something about the features of aca-demic language by working hard to stampthem out in writings meant to communi-cate with practitioners or the general pub-lic. Despite these practice-based sources ofknowledge about academic language, thecentral concept remains somewhat inchoateand underspecified. In the absence of a con-ceptual framework, it is difficult to designinstruction to promote academic language,to properly assess academic-language skills,or to understand what normal, expectableprogress toward achievement of academic-language skills might look like.

The goal of this chapter is to survey thework on academic language in order to pro-vide an overview of its features as a basisfor proposing a pragmatics-based frameworkthat accommodates those many discretefeatures in a coherent model of communi-cation. Based on this pragmatic framework,we then propose a research agenda focusingon issues that would take our understand-ing of this important topic a step farther.Given the absence of an agreed-upon setof criteria for academic language, we startby presenting an example of middle-schoolstudent writing to illustrate the rules of aca-demic language. We then turn to a moreformal inventory, based on theorizing aboutthe differences between oral and writtenlanguage, between informal and formal lan-guage, and between narrative and expositorylanguage, because these three distinctionsoverlap with and contribute to a sharpen-ing of the definition of academic language. Itis notable that academic language, unlike thecategories of written, formal, and expositorylanguage, has no clear opposite. We start,then, from the assumption that language canbe more or less academic – that is, furnishedwith fewer or more of the traits that are typ-ical of academic language; we have no basisfor postulating a separate category of lan-

guage that has passed some threshold quali-fying it as academic.

Academic Language in Use:Some Examples

The example we analyze herein was an end-of-week paragraph produced by a middle-school student participating in the pilotimplementation of a program intended topromote knowledge of all-purpose ‘aca-demic’ vocabulary in particular and use ofacademic language more generally. The pro-gram was designed for use in an urban dis-trict in which assessment had suggested thatstudents’ reading comprehension challengesmight be related to vocabulary limitations.Classroom observations in this district alsoshowed that the vocabulary instruction thatoccurred was primarily focused on disci-plinary terms (e.g., sonnet, legislation, diges-tion, and rhomboid in English language arts,social studies, science, and math, respec-tively), whereas pretesting showed that asignificant proportion of the students wasstruggling with the more all-purpose vocab-ulary found in their texts, including wordslike dramatic, interpret, sufficient, and decade.

The program consisted of week-longunits in each of which five ‘academic vocab-ulary’ words were targeted. The five tar-get words (and other words of similarly lowfrequency) were introduced in the contextof a paragraph about a topic selected tobe engaging to young adolescents and tobe somewhat open-ended (i.e., supportinga number of different plausible points ofview). The introductory paragraphs werewritten in a style that might be character-ized as ‘serious journalistic’ and each brieflypresented two or more positions on thetopic of the week, with limited elaborationof each position. The instructional programpresented focused teaching about the targetwords (i.e., their varying meanings in dif-ferent contexts, morphological analyses, andvariants of them) as well as contexts for thestudents to use the words. Thus, some formof debate about the issue in the paragraphwas recommended for social studies class



and “taking a stand,” a short argumentativeessay in which each student was asked todevelop and justify his or her opinion aboutthe topic of the week, was the standard Fri-day activity.

It is important to note that these “takinga stand” paragraphs were written in 10 to15 minutes, were not preceded by explicitinstruction in how to construct them, andwere not graded. Thus, we can assumethat they reflect something of the students’natural and unedited writing style, with theexception that students were encouragedto use the words of the week if possible;complying with this request sometimes ledto awkward constructions or even outrighterrors.

EXAMPLE 1: Female Seventh-Grader (12/08/06),responding to the prompt: What do you think thefunction of school is?

What’s the purpose ofschool you tell me!Well first of all, schoolis to get youreducation.[S]o we can learn whatthe teachers learn[,] sowe can be ready forthe 8th grade. Becauseif we don’t get [an]education[,] you can’tbe what you want tobe when [you] growup.Secondly to get usready so we can makeit to the 8th gradeready and prepared[.][T]hey don’t want tosend [us] to the 8thgrade because they likeus or the[y] just feelsorry for us. No![T]hat’s not thereason[.] [T]hey wantto prepare [us,] makesure we understandthe work. When wegrow up we also wantto get a good jobbecause we are the

• involved style• colloquial expressions• redundancy• simple connectives• inconsistent

perspective-taking(you/we)

future leaders of theworld. That’s [why]we need to work withthe function of theschool[,] so you [can]show us [how] theworld should be.

As a piece of writing from a seventh-grader, this falls short of excellence on manygrounds. First, it is inconsistent in attentionto conventions like capitalization, punctu-ation, and spelling (corrections introducedfor readability are indicated in square brack-ets). Second, the major position expressed(i.e., we go to school to get a good job,to be what we want to be when we growup) is somewhat obscured by other claims(i.e., attending school is necessary for pro-motion to higher grades, teachers will notpromote students out of pity, teachers wantstudents to learn) whose relationship to thecentral claim is left unclear. These issuesof form and content are rightly impor-tant in judging this as a piece of writing,but our focus in this chapter is how werespond to it as a piece of academic lan-guage. The key question is: What is themost effective pedagogical response to writ-ings like this? Will academic language fol-low naturally if students are helped to for-mulate their ideas more fully and precisely,or should one teach the academic formsusing the content the students themselveshave generated? If we agree that revisionand rewriting help to improve the qualityof writing (Klein & Olson, 2001), then whatwould be the best strategies to scaffold effec-tive revision and rewriting geared towardimproving academic-language skills morebroadly?

Consider Example 1a, a rewritten ver-sion of the previous example paragraph thatattempts to express the same ideas in a moreacademic form.

EXAMPLE 1A.What is the purpose of school?First, school functions to provide an education,so students can learn what the teachers know andbe prepared to continue their education at higherlevels. Teachers will not promote students who



have not learned the material, so understandingthe work is very important. Without an educa-tion, attaining one’s career goals is very difficult.

Second, getting a good job is dependent ongoing to school. Today’s students are the futureleaders of the world. School could help themunderstand how the world should be.

Analyzing the differences between Ex-amples 1 and 1a reveals some of the keyfeatures of academic language. Example 1aeliminates markers of involvement (e.g.,you tell me! No!); removes redundancy(the point about getting ready for eighthgrade is made twice in Example 1, onlyonce in Example 1a); moves from specificand personal to generic formulations ofclaims (eighth grade becomes education athigher levels; they don’t want to send us . . .becomes teachers will not promote . . . );substitutes metadiscourse markers (first,second) for more colloquial expressions (wellfirst of all); compresses Example 1 clausesinto adverbial phrases (without an educa-tion) and nominalizations (getting a goodjob); and imposes a consistent, distant,third-person perspective, whereas Example1 shifts between first- and second-personperspectives. Although more academic instyle, Example 1a is still not a particu-larly good response to the topic assignedbecause it is restricted to the same ideaspresented in more or less the same order asExample 1.

The paragraphs written by the middle-grade students participating in this programwere not devoid of academic-language fea-tures. Some students provided overarch-ing initial or concluding statements, usedmetadiscourse markers, and incorporatedthe academic vocabulary they were taught.Nonetheless, the paragraphs, particularlyin contrast to the more academic trans-lations one could provide for them, dis-played language features inappropriate foracademic language (e.g., colloquial expres-sions, involvement markers, redundancy)and revealed characteristics of academic lan-guage that the students did not employ (e.g.,grammatical compression, generic state-ments, impersonal stance, a variety of con-nectives).

Finally, comparing the student taking-a-stand paragraph to the original paragraphdesigned to stimulate thinking about thistopic reveals still more academic-languagefeatures.

EXAMPLE 1B. Paragraph prompt.What is the purpose of school?Why do we go to school? One prime goal of edu-cation is to transmit knowledge. Another is toenhance students’ capacities to earn a good liv-ing. Some would argue that schools should orientstudents toward a set of shared values, in orderto facilitate the maintenance of a democraticstate. Others contend that schools should helpstudents develop an understanding of the per-spectives of others, to promote social harmony.Still others think schools should teach students tochallenge authority, reject received opinion, andthink for themselves. Of course, if we accept thislast version of what schools should do, then wewill have to expect that the curriculum will bemassively adjusted and classroom activities radi-cally altered. Whereas thinking for themselves issomething educators value, students don’t alwayshave the license to do so in the classroom.

This adult-written paragraph reveals anumber of features not present in the stu-dent paragraphs, as follows:

� lexical density� modal verbs� endophoric reference� abstract entity as agent (school)� wide variety of connectives� stepwise logical argumentation� evidence of planning� detached stance� authoritative stance� lots of abstract/low-frequency vocabulary� elaborate noun phrases (nominalization)� markers of course of rationale� deductive/inductive inference

Some of these features, such as the highdensity of relatively low-frequency words,were deliberately introduced into the para-graph to serve the purposes of the program.Others were required by the argumentativegenre; these included the logical progressionof the argument and the explicit markingof different points of view. Other features,



such as the use of nominalization, were aproduct of efforts to keep the paragraphbrief. Others, such as the authoritative anddetached stance, are simply the default aca-demic writing style.

Academic Language: An Inventoryof Features

Having explored examples of academiclanguage and its absence in actual practice,we now must confront the issue of how toconceptualize ‘academic language.’ A firstadvantage of a coherent characterizationof academic language might be the valueof sharing it with struggling students.Schleppegrell (2001) argues that only rarelyare the linguistic expectations of school-based tasks made explicit to students,despite the fact that students’ academicperformance is judged considering theseexpectations. Without explicit discussion oflinguistic expectations, academic languageconstitutes an arcane challenge for many,and some explicit teaching about it mightbe useful.

Linguists and educational researchershave revealed features about which studentsmight be taught through contrastive analy-sis of language corpora (e.g., Biber & Rep-pen, 2002; Chafe & Danielewicz, 1987), evo-lutionary analysis of scientific language (Hal-liday & Martin, 1993), explorations of per-formances at different levels of expertise(Schleppegrell, 2001), in different academicdisciplines (Achugar & Schleppegrell, 2005;Schleppegrell, 2007), and in specific gen-res (Halliday & Martin, 1993; Swales, 1990).Table 7.2 represents an effort to summarizethis literature, by organizing the many lin-guistic features identified under the domainsof knowledge involved in academic lan-guage in a way that makes them some-what more tractable.3 The features listedto the left are more characteristic of collo-quial language, whereas the features towardthe right are more typical of academic lan-guage. Linguistic features are divided intothose referring to interpersonal stance, infor-mation load, organization of information,

lexical choices, and representational con-gruence (i.e., how grammar is used todepict reality). Of course, the realizationof all these features requires knowledge ofspecific vocabulary and grammatical struc-tures. In addition to these linguistic skills,three core domains of cognitive accomplish-ment involved in academic-language per-formance are genre mastery, command ofreasoning/argumentative strategies, and dis-ciplinary knowledge.

The typical interpersonal stance expectedin academic language is detached and author-itative. As we saw in Example 1, you tell me!and No! are markers of an involved stylethat in Schleppegrell’s words form part ofa “hortatory style that instantiates a con-text of interaction” (2001, p. 446). In con-trast, academic language requires a nondia-logical and distant construction of opinion,as well as “an assertive author [or speaker]who presents him/herself as a knowledge-able expert providing objective informa-tion” (Schleppegrell, 2001, pp. 444–445) (foran illustration of detached versus involvedwriting styles, see Schleppegrell, 2001,p. 445).

The information load in academic dis-course is characterized by conciseness anddensity. Academic writing or speech isexpected to be short and to the point,conveying information without unjustifiedrepetitions. In Example 1, the repetitionof being ready for 8th grade stands outas a violation to the flow of informationexpected in such a piece of writing. In con-trast to the typical redundancy of sponta-neous speech (Ong, 1982/1995), concisenessis highly valued, with only the minor excep-tions of artful pseudo-redundant moves suchas those included in abstracts or sum-maries and conclusions. Besides, academiclanguage packs a lot of information intoa few words. This informational density isevident in the high proportion of contentwords, usually achieved through nominal-izations and expanded noun phrases (Chafe& Danielewicz , 1987; Halliday & Martin,1993; Schleppegrell, 2001).

At the syntactic level of organization ofinformation, Halliday (1994b) subdivided the



Table 7.2. Linguistic Features and Core Domains of Cognitive Accomplishments Involvedin Academic Language Performance

More Colloquial More Academic

1. Interpersonal stanceExpressive/Involved → Detached/Distanced (Schleppegrell, 2001)Situationally driven personal

stances→ Authoritative stance (Schleppegrell, 2001)

2. Information loadRedundancy (Ong, 1995)/

Wordiness→ Conciseness

Sparsity → Density (proportion of content words per total words)(Schleppegrell, 2001)

3. Organization of informationDependency (Halliday, → Constituency (Halliday, 1994b)/Subordination (Ong, 1995)

1993)/Addition (Ong, 1995)(one element is bound orlinked to another but is notpart of it)

(embedding, one element is a structural part of another)

Minimal awareness of → Explicit awareness of organized discourseunfolding text as discourse (central role of textual metadiscourse markers)(marginal role ofmetadiscourse markers)

(Hyland & Tse, 2004)

Situational support → Autonomous text(exophoric reference) (endophoric reference)

Loosely connected/dialogicstructure

→ Stepwise logical argumentation/unfolding, tightly constructed

4. Lexical choicesLow lexical diversity → High lexical diversity (Chafe & Danielewicz, 1987)Colloquial expressions → Formal/prestigious expressions (e.g., say/like vs. for instance)Fuzziness (e.g., sort of,

something, like)→ Precision ( lexical choices and connectives)

Concrete/common-senseconcepts

→ Abstract/technical concepts

5. Representational congruenceSimple/congruent grammar

(simple sentences, e.g., Youheat water and it evaporatesfaster.)

→ Complex/congruentgrammar (complexsentences, e.g., If the watergets hotter, it evaporatesfaster.)

→ Compact/Incongruentgrammar (clauseembedding andnominalization, e.g.,The increasingevaporation of waterdue to risingtemperatures)(Halliday, 1993)

Animated entities as agents(e.g., Gutenberg inventedprinting with movable type.)

→ Abstract concepts as agents(e.g., Printing technologyrevolutionized Europeanbook-making.) (Halliday,1993)

�Genre masteryGeneric Values (Bhatia, 2002)

(narration, description,explanation. . .)

→ School-based genres(e.g., lab reports,persuasive essay)

→ Discipline-specificspecialized genres

(continued)



Table 7.2 (continued)

More Colloquial More Academic

�Reasoning strategiesBasic ways of argumentation → Specific reasoning moves → Discipline-specific

and persuasion valued at school(Reznitskaya et al., 2001)

reasoning moves

�Disciplinary knowledge• Taxonomies

Commonsense understanding → Abstract groupings andrelations

→ Disciplinary taxonomiesand salient relations

• Epistemological assumptionsKnowledge as fact → Knowledge as constructed

traditional category of subordinated clausesinto “hypotactic” and “embedded” clauses.Hypotactic clauses are subordinated clausesthat are dependent on but not constitutiveof other clauses, such as adverbial clauses orthose introduced by verbs of saying or think-ing (Colombi, 2002). In the following exam-ple, clause a and clause b are hypotacticclauses: I concluded [that the party was a totalfailure]a [because it ended before midnight]b.In contrast, embedded clauses form part ofanother clause, such as clause a and clause bin the following sentence: The party [whichended before midnight]a was a total failure[that we hope will not be repeated]b. Whereassome posit that addition and coordinationare characteristic features of colloquial lan-guage that contrast with subordination andcomplex syntax (Ong, 1982/1995), Halliday(1994b) persuasively argued that the crucialdistinction is dependency (which includeshypotactic subordinated clauses) versus con-stituency (embeddedness). He argued thatembedding is a distinctive feature of sci-entific or academic discourse. If we con-trast the subordinated clauses in Examples1 and 1b, it becomes evident that embed-ded clauses are used only in the lattertext.

Organization of information also involvesexplicit marking of text structures. Explicitawareness of text structure is indexed viadiscourse and metadiscourse markers thathave been widely explored in the literature.

In Hyland and Tse’s (2004) words: “metadis-course represents the writer’s awareness ofthe unfolding text as discourse: how writ-ers situate their language use to include atext, a writer, and a reader” (p. 167). Theseauthors developed a taxonomy of metadis-course markers and their functions by study-ing different types of texts. Additionally,information in academic language needs tobe organized according to a stepwise logi-cal argument structure that makes sophisti-cated use of autonomous endophoric referencestrategies instead of relying on situationalcontext or underspecified references.

At the lexical level, a diverse, precise, andformal repertoire that includes appropriatecross-discipline and discipline-specific termsis desirable.

The final level concerns representationalcongruence, or the correspondence betweenlanguage and the reality it represents. Theconcept of grammatical metaphor plays acentral role in Halliday’s model. Accord-ing to Halliday (1994a), in children’s com-monsense language, nouns refer to things,verbs refer to processes, adjectives denoteattributes, and connectives establish rela-tionships. However, when these grammat-ical categories are extended beyond theirprototypes (e.g., when nouns refer to pro-cesses like evaporation or verbs refer torelationships like precede), a grammaticalmetaphor, which Halliday calls a compactand incongruent form, is created. He argued



that experience is reconstructed when nom-inalized forms such as evaporation are used;this term has the semantic features both ofprocesses (water evaporates) and of things(because a noun prototypically refers tothings). In Halliday’s terms, these processeshave been transformed metaphorically intovirtual objects, “[t]he effect of this is to pro-vide a less dynamic, more synoptic vision ofthe world, in which reality is as it were heldstill, rendered fixed, bounded and deter-minate, so that it can be observed, mea-sured and, if possible, explained” (Halliday,1994a, p. 14). Halliday emphasized that therewould be no noticeable effect of sporadicuses like that but that academic languageis profusely populated by these grammaticalmetaphors (in particular, nominalizations ofprocesses).

Nominalization also creates lexical den-sity. The recursive linguistic principle per-mits nominalizations to function as embed-ded clauses of other propositions, allowinglong, information-packed sentences. Fur-thermore, in examples like The increasingevaporation of water due to rising tempera-tures is alarming, not only is the nominaliza-tion phrase the subject of a longer sentence,but it also constructs the claim of relation-ship between rising temperature and evapo-ration as assumed truth rather than a falsifi-able claim, contributing to the authoritativestance previously discussed.

However, grammatical metaphor is notthe only case of representational incongru-ence. Another incongruent move of aca-demic language involves using abstractconcepts as agents. Whereas in colloquialinteractions, animate entities are typicallythe grammatical agents of sentences, aca-demic language often displays abstract con-cepts as agentive subjects of sentences. Forexample, in Gutenberg invented printing withmovable type, a noun that refers to a person isthe subject and agent of the sentence. How-ever, the sentence, Printing technology rev-olutionized European book-making, presentsa noun that refers to an abstract conceptas agent, a less intuitive construction thatdeparts from our commonsense knowledgeof the world (Halliday & Martin, 1993).

Finally, all these linguistic features mustbe coordinated with at least three additionalcognitive accomplishments: genre mastery(Bhatia, 2002; Swales, 1990); command ofreasoning/argumentative strategies (Reznit-skaya, Anderson, Nurlen, Nguyen-Jahiel,Archodidou, & Kim, 2001); and disciplinaryknowledge (Achugar & Schleppegrell, 2005;Wignell, Martin, & Eggins, 1993). As stu-dents advance in their mastery of thesethree domains of knowledge, they learnto put features of academic language atthe service of genre conventions, persuasiveand clear argumentations, and disciplinary-specific relationships and concepts. Theseare three vast areas of research, which havebeen the focus of work in fields such asEnglish for Specific Purposes, the “Sydneyschool” of genre theory, and the Collabo-rative Reasoning approach, among others.Reviewing these three areas with the detailthey deserve would go beyond the scope ofthis chapter.

As we have seen herein, the claims thathave been made in the literature about thecharacteristics of academic language resultin a lengthy list of features. The mere lengthof the list in Table 7.2 displays the prob-lem with our current conception of aca-demic language: dozens of traits have beenidentified that contrast with primary or col-loquial language and that might functionas markers of academic language, but it isunclear that any of them actually definesthe phenomenon. Any of these traits mightbe present in casual spoken language: Is ittheir co-occurrence that defines some lan-guage as academic? Is it their frequency?How, if at all, do these various traits relateto one another? Are some particularly cru-cial and others merely epiphenomena? Aresome causes and others consequences? Howdoes the list in Table 7.2 help us with thetasks of assessment or instruction?

A Pragmatics-Based Approachto Academic Language

The problem with the inventorizing app-roach reflected in Table 7.2 is the omission



REPRESENTING THE SELF & THE AUDIENCE

Acknowledging the status ofthe audience

Performing a specificcommunicativefunction

Signaling a specificrelationship withaudience

Representing one’s stance onone’s message

Selecting the appropriate voice

Indexing epistemologicalstatus of one’s claims

REPRESENTING THE MESSAGE

Selecting/Participating in a genre

Adjusting message to audience’sknowledge status

Representing/Constructingmeanings and ideas

[Indexing source of information]

ORGANIZING DISCOURSE

Using discourse markers to signal textualrelations

Using discourse markers to signalinteractional relations

Making choices ofreference

CONSTRUCTINGCLAUSES

Choosing lexicon

Choosinggrammaticalstructures

Figure 7.1. Nested challenges within any communicative event.

of attention to the overall rationale for thesefeatures of academic language. In otherwords, we start from the assumption thatlanguage forms represent conventionalizedsolutions to communicative challenges andthat decisions about specific forms consti-tute solutions to those challenges. What arethe communicative challenges to which thefeatures of academic language are meant torespond?

In Figures 7.1 and 7.2, we present a firstattempt to answer this question (and, in theprocess, questions about how the traits listedin Table 7.2 relate to one another). Figure 7.1represents a view of language in which com-municative goals are seen as driving deci-sions about specifics of expression. In thisview, all communicative forms are a simul-taneous solution to two tasks: representingthe self and representing the message. Rep-resenting the self involves selecting (or per-haps simply having) a voice and a relation-ship to the audience; representing the mes-sage requires conceptualizing some thoughtand figuring out what the audience alreadyknows and needs to know about it.

Given a representation of self and mes-sage, then discourse and utterance fea-tures consistent with those prior frames arerealized.

In many communicative exchanges, self-representation is fairly straightforward (e.g.,self as purchaser of a kilo of onions, self asstudent in a first-grade classroom) and themessage is relatively uncomplicated (e.g.,How much do the Vidalia onions cost? 3

plus 2 equals 5). The rules governing dis-course structure, lexical selection, and gram-matical formulation for such exchanges areaccordingly relatively easy to learn and toimplement. Furthermore, formulating somefrequently occurring but potentially chal-lenging messages has been greatly simpli-fied by the availability of conventional formsdesigned to express them (e.g., greetings,requests, apologies, condolences).

We argue, however, that characteristicsof academic language represent an accom-modation to the two ubiquitous features ofcommunicative tasks – representation of selfand of one’s message – under particularlychallenging conditions (see Figure 7.2).

The first condition is the need to formu-late messages that are relatively challengingon any number of grounds – for example,because the content is inherently compli-cated, because some of the concepts beingtalked about are abstract or theoretical,because some of the claims being made havean uncertain epistemological status, and so



REPRESENTING THE SELF AND THE AUDIENCE

Acknowledging status ofintangible non-interactiveacademic audienceand its level of expertise

Displaying one’s knowledge/extending someone’s knowledge

Emphasizing co-membershipwith an expert academic audience

Presenting a neutral,dispassionate stance onone’s message

Selecting an authoritative voice

Explicitly acknowledgingand clarifying when necessarythe epistemologicalstatus of one’s claims

REPRESENTING THE MESSAGE

Selecting one of the approved academic genres

Adjusting level of detail andamount of backgroundinformation provided to level ofexpertise of the intended audience

Representing abstract, theoreticalconstructs, complicated inter-relationships, conditionals, hypo-theticals, counterfactuals, and otherchallenging cognitive schemas

[Explicitly acknowledge sourcesof information/evidence]

ORGANIZING DISCOURSE

Using discourse markers to emphasize theintegration of information, the causal,temporal, or inferential relations beingemphasized

Expressing metatextualrelationships precisely

Using reference termsthat are approvedwithin the discoursecommunity, oftentechnical

Figure 7.2. Nested challenges within a communicative event calling for academic language.

on. It is simply more difficult to explainthe process by which cells replicate, or thetheory of evolution, or the various factorscontributing to global warming than it is tonegotiate the purchase of onions or respondto an addition problem; therefore, the lan-guage required must be more complicated.4

The second challenge is to identify theaudience and the appropriate relationshipbetween self and audience. An early devel-opmental task is to assess the listener’sknowledge so as to provide sufficient infor-mation and to gradually free language fromsituational support. The additional commu-nicative challenges of academic languagerequire learning the traditions that governdiscourse among participants in an intan-gible academic community. The questionsof who the audience is and what theyknow are crucial in appropriately framingthe discourse in academic tasks, yet theyare not always easy to unravel for stu-dents. In face-to-face interactions, speakerslearn language by identifying co-occurrencesbetween language forms and situational con-text via repeated participation in similarspeech events with clearly identifiable par-ticipants. In those situations, children ini-

tially rely on contextual support (e.g., point-ing, enactment, gestures, deictics), but theygradually learn to use language as its owncontext. Of course, autonomous discourseskills develop throughout the school yearsand are needed in many nonacademic sit-uations as well (e.g., talking on the tele-phone, telling a story, writing a letter toa friend). From a communicative perspec-tive, what seems to make academic lan-guage particularly challenging, in additionto the complexity of the message, is that thecomponents of the communicative situationare less obvious and less accessible. In theacademic-discourse world, identifying pat-terns of co-occurrences between specific sit-uations and linguistic forms is a much hardertask. Approaches to this task taken by inex-perienced users of academic language rangefrom borrowing oral-language forms5 to imi-tating experts’ discourse so slavishly as toverge on plagiarism.

Moreover, the producers of academic lan-guage need to establish their own level ofauthoritativeness and negotiate their rela-tionship with a distant and potentially crit-ical or incredulous audience, through thelanguage forms chosen. Impersonal, generic,



and distancing forms are required becauseeven if a personal relationship betweenthe producer of academic language andthe audience does exist, that relationshipis irrelevant to the self being representedunder conditions that call for academiclanguage. Thus, the intrusion of spoken-language involvement markers in Example 1

represents a violation of academic-languagenorms because involvement with the audi-ence is inappropriate under those circum-stances. Control over modals and explicitmarkers of epistemological status (e.g., prob-ably, likely, undoubtedly, evidently, obviously)represents acknowledgment of the need tobe explicit about the credibility of one’sclaims. That need derives partly from theobligation to represent the message accu-rately and partly from the protection of per-sonal authority that comes from making rea-soned and modulated claims.

Thus, the challenge lies not only in theaudience’s physical absence but also moreprofoundly in the somewhat indeterminatenature of this audience. Figure 7.2 describesthe audience as an “intangible noninterac-tive academic audience.” At school, eventhough teachers are the ones who requestassignments, students need to suspend theirpersonal relationship with their teacher andignore what they know their teacher knowsin responding. Instead, they need to imag-ine a nonfamiliar audience with high levelsof language but without specific knowledgeof the target topic.

In line with the pragmatics-based modelproposed herein, we think that two essen-tial starting points for students are to (1)gain an awareness of the desired relation-ship among participants in academic com-munications; and (2) understand that mean-ing resides not only in what they say butalso in how they communicate it. We arearguing, then, that the long list of academic-language markers reviewed in Table 7.2can be sorted out usefully by fitting thevarious items into this pragmatics-basedunderstanding of academic language. Formsthat have to do with the largest task –self-representation – are those that expressauthoritativeness, that perform the function

of displaying knowledge to or for some-one, that acknowledge co-membership withthe audience, that express the speaker’sunique voice within the ‘academic commu-nity,’ and that make explicit the epistemo-logical assumptions under which the speakeris operating. Those markers, then, must beintegrated with language forms imposed byan adequate representation of the messageto be conveyed, which in turn leads to deci-sions about genre (in the broadest sense),about the audience’s level of backgroundknowledge to be presupposed and the levelof detail to be included, about the mecha-nisms for making reference to key conceptsand interrelationships, and about the need toacknowledge sources of information. Havingestablished what self and what informationwill be represented, then text-specific deci-sions at the level of discourse organization(e.g., How will the organizational structureof the discourse be signaled? How will rela-tionships of temporality, causality, depen-dency, conditionality, and so forth be talkedabout? How much anaphoric and exophoricreference is permissible?) and clause con-structions can be made.

Figures 7.1 and 7.2 present a relativelysimplistic view of the nested relationshipsamong these different levels; clearly, muchmore work would need to be done to spec-ify implications of a specific decision at anyof the levels for decisions at lower or higherlevels. Nonetheless, we hope that this rep-resentation makes clear that the clause- ordiscourse-level characteristics typical of aca-demic language may occur under other cir-cumstances, but that the most likely condi-tions for them are in satisfying the demandsthat are particular to self-representing asa member of the ‘academic-language–usingcommunity’ and that are imposed by theneed to express complex content in efficientand effective ways.

Academic Language: A ResearchAgenda

The view presented herein makes no clearpredictions about the order of development



of the various academic-language markersor about an optimal approach to teachingacademic language. Indeed, these are issueswe would prioritize in a research agendafocused on academic language. Clearly, chil-dren start acquiring clause-level skills assoon as they learn to talk; the challengefor teachers is to figure out precisely howthe construction of clauses needs to beadapted to contexts in which academic lan-guage must be produced and/or compre-hended and what new lexical and gram-matical knowledge is needed to succeedin those contexts. Similarly, children pro-duce extended discourses from early in theirlanguage-acquisition trajectories, and theyuse in conversational narratives some fea-tures that may also be relevant to academic-language texts. So, the specific task ofbecoming skilled in academic languagerequires expanding the repertoires availableat those two levels for use in nonacademiccontexts.

As can be inferred from the model pre-sented herein, the skills required for suc-cessful academic-language performance gobeyond the traditionally cited lexicogram-matical skills to include a level of meta-communication. For instance, research withHebrew-speaking children and adults hasshown that whereas knowledge of for-mal sophisticated morphology and syntaxincreases from age five to age seven, onlycollege-educated adults and some older ado-lescents are capable of appropriately dis-playing this knowledge in the constructionof texts (Ravid & Tolchinsky, 2002). Asstated by Berman (2004) and by Ravid andTolchinsky (2002), a crucial aspect of laterlanguage development, in addition to vocab-ulary expansion, is learning a variety ofsophisticated morpho-syntactic structuresand how to use them flexibly for diversecommunicative purposes.

We propose, then, two large categoriesof urgent research questions. One set has todo with the developmental course of aca-demic language and includes attention toissues such as the following: What does nor-mative development look like? How does itrelate to literacy development? Which early-

developing language skills constitute precur-sors to later academic language? How dothe various components of academic lan-guage relate to one another? The secondset has to do with instruction – for exam-ple: What are effective methods for pro-moting academic language? Which aspectsof the system need explicit instruction andwhich do not? How can we best embed(or not) academic-language instruction intoattention to literacy instruction and content-area learning?

We expand briefly on the research basefor these two sets of questions in the follow-ing sections.

The Developmental Course andComposition of Academic-LanguageSkills

What Are the Early Precursors?

Even though the field of academic languageis concerned mostly with the study of laterlanguage development, it is of crucial impor-tance to recognize that academic-languageskills fall on a continuum with earlier lan-guage skills. Within this view, exploringearlier language skills that might predictacademic-language skills later in life is ofparticular educational relevance. Specifi-cally, we need research to explore whichskills are predictive of later mastery of aca-demic language and, in turn, which contextsare most conducive to efficient learning ofacademic-language skills.

Reviewing relevant literature, Blum-Kulka (in press) documents preschool chil-dren’s early development in the areas of con-versation and extended discourse, includingwhat she calls literate discourse. Blum-Kulkadefines literate discourse as “include[ing] allthose uses of language that involve elementsof planning, precision, distancing, internalcoherence, and explicitness. It may appearin discursive events that mainly require theskills for constructing a continuous text,such as public lecture or written articles, aswell as when the main requirement is con-versational skills . . . especially on topics thatare remote from the here-and-now” (p. 9).



Within conversational skills, Blum-Kulkaincludes thematic coherence, frequency of top-ical initiatives, capacity for regulation, cor-rection and metapragmatic comments (e.g.,say it in baby talk), and sociolinguistic skills(i.e., the ability to choose a linguistic styleappropriate to the social circumstances ofthe speech event). Extended discourse skillscomprise structural development (genre fea-tures); enrichment of linguistic means (tex-tual fabric, used to structure the text);conversational autonomy (free from conver-sational scaffold from interlocutors); textualautonomy (ability to correctly assess the stateof knowledge of the interlocutor so thatinformation in text does not assume sharedknowledge); and expansion of range of inter-est, among others.

From a theoretical standpoint, we couldenvision many of these early skills as founda-tional abilities or rudimentary precursors forlater, more sophisticated academic-languageskills.

Research on metadiscourse (Hyland &Tse, 2004) also provides an interesting tax-onomy of markers that might prove relevantfor the study of younger students’ oral andwritten academic language. Whether thesemetadiscourse elements will be sensitive todevelopmental changes, in addition to beingsensitive to different functions of texts,is an open question that deserves furtherinvestigation. Research on the applicabilityof this taxonomy for pedagogical purposesis another potentially fruitful enterprise.Further research looking at these poten-tial associations would be illuminating, bothto construct a comprehensive theoreticalmodel and to inform the design of coherenteducational programs.

What Is the Role of MetapragmaticAwareness?

In line with the conceptualization of aca-demic language presented in Figure 7.2, weurge research attention to the question ofwhether sociolinguistic and stylistic aware-ness plays a pivotal role in the developmentof academic language. We might hypothe-size that sociolinguistic awareness is a pre-

requisite to mastering academic language.Systematic linguistic variation can be dialec-tal, sociolectal, ethnic, or gender-based, aswell as determined by genre, register, andmodality (Ravid & Tolchinsky, 2002). Theability to switch appropriately across lan-guage varieties and registers depends onthe opportunities to participate in variouscommunicative situations (Hymes, 1974).Whereas most speakers can at least par-tially adapt their language forms to spe-cific contexts, expanding these adaptationskills so that students learn how to map lan-guage forms onto a variety of situations ina conscious and reflective way may be acrucial step in fostering academic-languageproficiency. Moreover, stylistic awareness –that is, being aware of a set of linguisticoptions that have the potential to realize avariety of alternative meanings – may alsobe necessary. Schleppegrell’s research con-nects particular language forms with spe-cific expectations in illuminating ways. Forinstance, she documents how the authori-tative stance typical of academic discourseis constructed through impersonal sub-jects, declarative mood structure, and lex-ical realization of meanings; and she relateslexical density and nominalization to thefunction of incorporating more (ideational)content into each clause (Schleppegrell,2001).

What Is the Effect of Mode?

Academic language is understood as a con-struct that goes beyond modes of expres-sion and disciplinary boundaries. Bailey(2007) argues for a core set of academic-language skills that cuts across different dis-ciplines and is complemented with addi-tional discipline-specific skills. Within thisconceptualization, it is relevant to studyhow different modes of expression (writtenversus oral) and skills in specific discipline-based genres (a social studies report versus ascience-lab report) influence each other. Towhat extent skills learned in one mode ofexpression or in one genre transfer to otherdomains is an important question, with rel-evant pedagogical implications.



Whereas nobody would deny a bidirec-tional influence of spoken and written lan-guage, some researchers emphasize one sideas the source of more sophisticated skills.Ravid and Tolchinsky’s (2002) intriguingmodel of linguistic literacy proposes a bidi-rectional influence; however, their modelstates that basic features (e.g., basic syn-tax and phonology) are transferred fromspeaking to writing, whereas sophisticatedfeatures originate in writing and, therefore,exposure to and production of written lan-guage is the main factor in enriching linguis-tic literacy. However, some complex fea-tures might also transfer from spoken towritten language, as Collaborative Reason-ing studies demonstrate (Reznitskaya et al.,2001). Reznitskaya and colleagues show thathigher levels of argumentation or reason-ing can be achieved through the scaffold-ing of explicit discourse stratagems. Toconstruct a theoretical model that estab-lishes associations or predictive relationshipsacross modes of expression, research needsto assess later language development so thatwe can begin to understand which skills gettransferred under which conditions.

Is Academic Language Truly MoreGrammatically Complex?

Findings on syntactic complexity of aca-demic language are not uncontroversial.Whereas many authors have pointed toa higher degree of subordination in aca-demic writing versus colloquial speech, oth-ers (Poole & Field, 1976) have reported moreembedding in spoken language. Tolchinskyand Aparici (2000) found a higher degreeof embedding in written than spoken narra-tives in Spanish but more frequent center-embedded relative clauses in subject posi-tion in spoken than written expository texts.As pointed out by Ravid and Tolchinsky(2002) and previously emphasized by Biber(1995), language features should be studiedtaking into account the influence of register,degree of formality, and planning. In thestudy of academic-language skills, then, thethree domains of knowledge identified inTable 7.2 and all contextual factors men-

tioned in Table 7.1 should be considered todevelop a precise picture of which skills aredisplayed under which circumstances.

What Is the Normative DevelopmentalCourse and the Ultimate Goal?

Teachers’ expectations and students’ skillsvary not only by grade but also by disci-pline and specific genres within disciplines.In addition, academic-language skills canprogress to reach highly sophisticated lev-els such as those used in sharing profes-sional knowledge among a community ofexperts. Within this range of possibilities,what should be considered the ideal devel-opmental endpoint for academic-languagedevelopment and, just as important, theminimal educational standards for differentgrades and content areas?

Teaching Academic-Language Skills

Which Academic-Language Skills ShouldBe Instructed?

Teaching about mechanisms for represent-ing complex information – both as an ap-proach to reading comprehension and asan input to academic writing – could wellbe helpful in supporting students’ devel-opment of academic-language skills. Hereagain, however, the task may be primar-ily one of expanding the learner’s reper-toire of useful stratagems for formulatingmessages because children from their firstmonths of talking understand the challengeof trying to express complex thoughts withlimited language skills. Consider the childlexical forms formerly seen as overgeneral-izations, such as calling the postman daddyor calling horses cows; most child-languageresearchers would now argue that these aresimply immature attempts to comment onsimilarities or to refer despite lexical gaps(Gelman, Croft, Fu, Clausner, & Gottfried,1998). Their occurrence suggests that evenyoung children can solve the problem ofexpressing complex ideas, although in waysthat may be unconventional and thus oftenunsuccessful.



Exposure to talking styles that display fea-tures of academic discourse and participa-tion in academic genres is probably essentialfor mastering academic language. Childrenwho come from families that value the accu-mulation and display of knowledge for itsown sake, who require warrants for claims,and who model and scaffold the organiza-tion of extended discourse and sophisticatedutterances will probably have a much eas-ier transition into academic language. How-ever, documentation of how some familiessupport their children’s academic-languageskills is sorely needed.

For school instruction, attention to lin-guistic form may be a powerful mechanismfor improving students’ academic-languageskills. A traditional grammar approachmight be effective, but the value of a discus-sion about self, audience, purpose, and theappropriate lexical and grammatical meansto represent information in specific schooltasks should be ascertained. Assuming thatteaching grammatical and lexical devices isessential, we agree with a little-cited claimmade by Bakhtin (1942) decades ago:

Without constantly considering the stylis-tic significance of grammatical choices, theinstruction of grammar inevitably turnsinto scholasticism. In practice, however, theinstructor very rarely provides any sort ofstylistic interpretation of the grammaticalforms covered in class. Every grammaticalform is at the same time a means of rep-resenting reality. Particularly in instanceswhere the speaker or writer may choosebetween two or more equally grammat-ically correct syntactic forms, the choiceis determined not by grammatical but byrepresentative and expressive effectivenessof these forms. Teaching syntax withoutproviding stylistic elucidation and with-out attempting to enrich the students’ ownspeech does not help them improve thecreativity of their own speech productions(quoted in Bazerman, 2005).

What Are Effective PedagogicalApproaches?

Research-based pedagogical approaches toteaching academic-language skills within

specific disciplines or genres are startingto emerge (Lemke, 1990; Schleppegrell,Achugar, & Oteiza, 2004). For instance,content-based instruction (CBI) is a ped-agogy for English as a Second Languagethat integrates language and content-areaknowledge with the purpose of improvingboth dimensions within specific disciplines(Schleppegrell & de Oliveira, forthcoming).These emerging approaches are promising,yet their design and effectiveness are still inneed of further study. How to make the lin-guistic expectations explicit to students, atwhat level of precision, and how to furtherdevelop sociolinguistic and stylistic aware-ness skills to improve academic-languageperformance in the classroom are still openresearch questions.

A related challenge is how to pro-vide instruction without prescription. Manygenre-based classroom pedagogies havebeen critiqued because of their prescriptiveand hierarchical nature and the low trans-ferability of skills produced (Fosen, 2000;Kamler, 1994). Developing students’ soci-olinguistic competence, stylistic awareness,familiarity with linguistic expectations, andcommand of lexical and grammatical fea-tures of specific genres while emphasiz-ing the individual creativity required foran expert mastery of the interplay betweenform and meaning is a major challenge.

How Do Planning, Revision, andRewriting Improve the Advancement ofAcademic-Language Skills?

In a conceptualization of academic languageas a construct that cuts across modes ofexpression, exploring the effect of editingas a way of fostering acquisition and aware-ness of academic skills seems promising.Whereas encouraging students to edit theirown texts seems to be a successful approachto improving writing skills, little research hasexplored the effect of rereading and revisingon students’ learning (Klein & Olson, 2001).Research suggests that frequent opportuni-ties for authentic writing improve the qual-ity of students’ written products (see Klein &Olson, 2001, for a brief review). Thus, would



frequent opportunities for editing texts havea positive effect on academic-language per-formance? If so, would the skills learnedtransfer from writing to speaking? Howmuch guidance do students need so thatediting can effectively improve academic-language skills?

How Can Schools Provide Intervention inAcademic-Language Skills to StudentsWho Start Far Behind?

Children enter school with different linguis-tic, sociolinguistic, and pragmatic experi-ences, and not all of them have been exposedto the forms of communication valued atschool. Strategies to make children feel com-fortable in expressing who they are and whatthey bring to school should be at the coreof any instructional program. At the sametime, schools have the moral obligation toprovide all children with equal opportuni-ties to participate in the discourse of aca-demics that is a requisite for later academicsuccess. Children’s education can be basedin their own culture while also providingexplicit teaching of the skills required forsuccess in the academic context of schools(Delpit, 1995). Snow, Cancini, Gonzalez,and Shriberg (1989) found that meeting theexpectations of a formal academic regis-ter, such as definitional discourse, correlatedwith academic success. Therefore, childrenwho are less skillful in academic language areless likely to succeed at school. How to pro-vide all children – ELLs and also strugglingnative English speakers – with equal oppor-tunities of mastering academic language in away that incorporates and values their pri-mary discourses is yet another challenge.

How Can the Role of Language inSelf-Presentation Be Taught?

It is not obvious that all children auto-matically see language as a form ofself-representation. Evidence from childrengrowing up bilingual suggests that theychoose the language that is effective forcommunication (i.e., for formulating a mes-sage that is likely to be successful) from a

very early age (e.g., Genesee, 2005, 2006;Genesee & Nicoladis, in press; Taeschner,1983) but that they become aware of the‘otherness’ imposed by speaking a minoritylanguage in public only somewhat later. Fur-thermore, understanding the relationship ofa language to an identity is rather differ-ent from understanding how features withina language express identity. Certainly, stu-dents do identity work through languagein adolescence (Eckert, 1989), but it is notclear how much metalinguistic awarenessthey have about those linguistic decisions.Thus, it is worth exploring whether stu-dents might benefit from teaching designedto make the problem of self-representationexplicit because that is a source of impor-tant academic-language features – but, atthe same time, a pragmatic force to whichthey may be blind. One approach to this taskmight be sociolinguistic exploration of ques-tions like “How does the language of peoplein power differ from the language of those insubordinate positions?” Another approachmight involve text analysis to determine, forexample, which markers readers use to inferthe writer’s level of certainty or to decidewhether they consider the writer trust-worthy.

Do Students Need Instruction inMetasociolinguistic Awareness?

Another somewhat different approachwould be taken if we assumed that studentsknew the importance of linguisticallymanaging self-representation but lacked afull understanding of the cues signaling theappropriate representation for academicsettings. In that case, a metasociolinguisticcurriculum might be appropriate, one thatspecified the factors leading to the needfor greater care in representing oneself asknowledgeable or trustworthy (see Table 7.1for a preliminary list of the situations thatdo/do not call for academic language). Howshould talking to one’s friends in class differfrom talking to them on the playground?How does pursuing an intellectual disputediffer from arguing with one’s boyfriend?6

Charting students’ knowledge about these



issues might be a research undertakingworth pursuing and might shed light onhow to gradually bring them to deeperunderstandings of the interactions betweenform and meaning.

A related research area involves exploringthe best strategy to help students understandthe importance of continuing to expandtheir language knowledge. For example,would it be fruitful to teach teachers and/orstudents explicitly about the concept of‘academic language’? Should we also teachstudents about the multidimensionality oflanguage – discussing, for example, how hav-ing a conversation with friends requires adifferent set of skills than a formal presenta-tion? Would this knowledge be helpful and,if so, at what level of specification? Whichpurposeful activities would best help pro-mote this learning?

In What Informational Context ShouldTeachers Teach Academic-LanguageExpression?

Studying the development of definitionalskill, Snow (1990) reported no age differ-ences in the amount of information chil-dren provided but significant age differencesin “the way they organize that informa-tion into the formal structure required” (p.708). These findings lead us to reflect aboutwhether is it too much to ask of students thatthey simultaneously learn content and lin-guistic organization. Should academic lan-guage perhaps be taught initially in the con-text of highly familiar topics or topics forwhich students have abundant backgroundknowledge?

Which Genres Are the Most Important?

Which discourse varieties should beincluded under the label “academic lan-guage” for the purposes of improving school-relevant linguistic skills? What are the cru-cial discourse varieties students need to mas-ter in school? Should we study mainly thelanguage of the most traditional academicsubjects, or should we also include otherprofessional discourses, such as journalistic,

legal, medical, or business language? Ravidand Tolchinsky (2002, p. 421) note that dis-course varieties “can be thought of as mul-tidimensional spaces within which speakersand writers move, and which can be definedat different depths of focus: for exam-ple, . . . the genre of a high school physicstextbook versus the less specific genre ofnatural sciences.” What should be the depthof focus in defining academic registers? Inother words, should we focus on highly spe-cific genres, such as a laboratory report, aproject proposal, and a biography, or shouldwe direct our efforts to clusters of genresthat share register features, such as scientificversus persuasive discourse?

Conclusion

We have suggested several possible linesof research focused on understanding theorigins of academic-language skill, probingthe differential success of different groupsof students with it and evaluating differentapproaches to helping all students master it.The basic question underlying all of thesesuggestions is one about the source of thechallenge: academic language, like all lin-guistic communication, involves challengesat the level of self-representation, represent-ing a message, constructing discourse, andcomposing utterances. Where in this nestedprocess do students encounter particular dif-ficulties, and are those difficulties primarilyones of understanding or of performing thetask? If we had the answer to these ques-tions, then we would be well on our wayto devising effective instruction for studentsand professional development for teachersto ensure universal improvement in this cru-cial aspect of academic functioning.

Notes

1 The authors’ names are in alphabetical order.The authors would like to thank The SpencerFoundation, which has supported the firstauthor’s work on this topic, and the Insti-tute of Education Sciences, which has sup-ported the second author through the grants



“Diagnostic Assessment of Reading Com-prehension: Development and Validation”and “Improving Reading Comprehension forStruggling Readers.”

2 Register is a central notion in SFL. Registeris defined as “the constellation of lexical andgrammatical features that characterizes par-ticular uses of language” (Halliday & Has-san, 1989). As elaborated by Schleppegrell(2001, pp. 431–432): “A register reflects thecontext of a text’s production and at the sametime enables the text to realize that context.In other words, the grammatical choices aremade on the basis of the speaker’s percep-tion of the social context, and those choicesthen also serve to instantiate that social con-text. . . . Registers manifest themselves bothin choice of words or phrases and also in theway that clauses are constructed and linked.”Each genre has its own register features anddifferent genres can share many common reg-ister features. Genres are purposeful, stageduses of language that are accomplished inparticular cultural contexts (Christie, 1985,as quoted in Schleppegrell, 2001). As statedby Schleppegrell, certain lexical and gram-matical features are common to many schoolgenres because they are functional for “doingschooling” (Schleppegrell, 2001, p. 432).

3 This table is organized in categories imposedby the authors of this chapter.

4 Note that we are not arguing here that aca-demic language is more complex overall thanother forms of language. Language forms con-stitute adequate responses to a variety ofcommunicative challenges; thus, complex-ity can be manifested at different levels invarious language exchanges. We are simplyhighlighting one specific dimension of com-plexity. More colloquial forms can be morecomplex in other dimensions – for exam-ple, in how linkages among clauses are indi-cated from one part of a discourse to another(Schleppegrell, 2001).

5 For example, a paper submitted to a specialissue of Hormones and Behavior that reviewedhow the functioning of pheromones as socialcues is mediated by brain structures includedthe sentence, “We thus conclude there issomething funky going on in the amygdala.”

6 It is worth noting that in one fifth-grade class-room that implemented the Word Gener-ation curriculum, the teacher often closeddown the heated student debates on the topicof the week by saying “but we are still going to

be friends, right?,” thus explicitly marking thedistinction between the academic argumentsand the normal classroom relationships.

References

Achugar, M., & Schleppegrell, M. J. (2005).Beyond connectors: The construction of causein history textbooks. Linguistics and Educa-tion, 16 (3), 298–318.

August, D., & Shanahan, T. (2006). Developingliteracy in second-language learners: Reportof the national literacy panel on languageminority children and youth. Mahwah, NJ:Lawrence Erlbaum; Washington, DC: Centerfor Applied Linguistics.

Bailey, A. (2007). The language demands of school:Putting academic English to the test. NewHaven, CT: Yale University Press.

Bailey, A. (in press). From Lambie to Lam-baste: The conceptualization, operationaliza-tion, and use of academic language in theassessment of ELL students. In K. Rolstad(Ed.), Rethinking school language. Mahwah,NJ: Lawrence Erlbaum Associates.

Bazerman, C. (2005). An essay on pedagogy byMikhail M. Bakhtin. Written Communication,22 (3), 333–338.

Berman, R. A. (2004). The role of context indeveloping narrative abilities. In S. Stromqvist& L. Verhoeven, (Eds.), Relating events innarrative: Typological and contextual perspec-tives. Mahwah, NJ: Lawrence Erlbaum, 261–280.

Bhatia, V. J. (2002). Applied genre analysis: Ana-lytical advances and pedagogical procedures.In A. M. Johns (Ed.), Genre in the classroom:Multiple perspectives. Mahwah, NJ: LawrenceErlbaum Associates.

Biber, D. (1995). Dimensions of register varia-tion: A cross-linguistic comparison. Cambridge:Cambridge University Press.

Biber, D., & Reppen, R. (2002). What does fre-quency have to do with grammar teaching?Studies in Second Language Acquisition, 24 (2),199–208.

Blum-Kulka, S. (in press). Language, communi-cation and literacy: Steps in the developmentof literate discourse. In P. Klein & K. Yablon(Eds.), Modes of early education and their effecton improving education in school. The Initia-tive for Applied Research in Education, IsraeliAcademy of Science.

Chafe, W., & Danielewicz, J. M. (1987). Prop-erties of spoken and written language. In



R. Horowitz & S. J. Samuels (Eds.), Compre-hending oral and written language (pp. 83–113).San Diego, CA: Academic Press.

Chamot, A. U., & O’Malley, J. M. (1994). TheCALLA handbook: Implementing the cog-nitive academic language learning approach.Reading, MA: Addison-Wesley.

Colombi, M. C. (2002). Academic languagedevelopment in Latino students’ writing inSpanish. In M. J. Schleppegrell & M. C.Colombi (Eds.), Developing advanced literacyin first and second languages: Meaning withpower, 67–86. Mahwah, NJ: Lawrence Erl-baum Associates.

Colombi, M. C., & Schleppegrell, M. J. (2002).Theory and practice in the development ofadvanced literacy. In M. J. Schleppegrell &M. C. Colombi (Eds.), Developing advancedliteracy in first and second languages: Meaningwith power (pp. 1–20). Mahwah, NJ: LawrenceErlbaum Associates.

Cummins, J. (1980). The cross-lingual dimen-sions of language proficiency: Implications forbilingual education and the optimal age issue.TESOL Quarterly, 14, 175–187.

Cummins, J. (1981). Age on arrival and immi-grant second-language learning in Canada:A reassessment. Applied Linguistics, 2, 132–149.

Cummins, J. (1984). Language proficiency andacademic achievement revisited: A response.In C. Rivera (Ed.), Language proficiency andacademic achievement. Clevedon, UK: Multi-lingual Matters.

Cummins, J. (2001). Negotiating identities: Educa-tion for empowerment in a diverse society. 2ndEdition. Los Angeles: California Associationfor Bilingual Education.

Delpit, L. (1995). Other people’s children: Culturalconflict in the classroom. New York: The NewPress.

Eckert, P. (1989). Jocks and burnouts: Social cate-gories and identity in the high school. New York:Teachers College Press.

Fosen, C. (2000). Genres made real: Genre the-ory as pedagogy, method, and content. Paperpresented at the Annual Meeting of the Con-ference on Composition and Communication,Minneapolis, MN.

Gelman, S., Croft, W., Fu, P., Clausner, T., &Gottfried, G. (1998). Why is a pomegranatean apple? The role of shape, taxonomic relat-edness, and prior lexical knowledge in chil-dren’s overextensions of apple and dog. Journalof Child Language, 25, 267–291.

Genesee, F. (2005). The capacity of thelanguage faculty: Contributions from stud-ies of simultaneous bilingual acquisition. InJ. Cohen, K. T. McAlister, K. Rolstad, & J.MacSwan (Eds.), Proceedings of the 4th Inter-national Symposium on Bilingualism (pp. 890–901). Somerville, MA: Cascadilla Press.

Genesee, F. (2006). Bilingual first language acqui-sition in perspective. In E. Hoff & P. McCar-dle (Eds.), Childhood bilingualism (pp. 45–67).Clevedon, UK: Multilingual Matters.

Genesee, F., & Nicoladis, E. (in press). Bilingualacquisition. In E. Hoff & M. Shatz (Eds.),Handbook of language development, Oxford:Blackwell.

Halliday, M. A. K. (1987). Spoken and writtenmodes of meaning. In R. Horowitz & S. J.Samuels (Eds.), Comprehending oral and writ-ten language (pp. 55–82). San Diego, CA: Aca-demic Press.

Halliday, M. A. K. (1993). Some grammaticalproblems in scientific English. In M. A. K.Halliday & Martin, J. R., Writing science: Lit-eracy and discursive power. London: Falmer(Critical Perspectives on Literacy and Edu-cation); Pittsburgh: University of PittsburghPress. (Pittsburgh Series in Composition, Lit-eracy, and Culture).

Halliday, M. A. K. (1994a). A language devel-opment approach to education. In N. Bird,et al. (Eds.), Language and learning. Paperspresented at the Annual International lan-guage in Education Conference, Hong Kong,1993.

Halliday, M. A. K. (1994b). An introduction tofunctional grammar. London: Edward Arnold.

Halliday, M. A. K., & Martin, J. R. (1993). Writ-ing science: Literacy and discursive power. Lon-don: Falmer (Critical Perspectives on Liter-acy and Education); Pittsburgh: University ofPittsburgh Press. (Pittsburgh Series in Com-position, Literacy, and Culture).

Hemphill, L. (1989). Topic development, syntax,and social class. Discourse Processes, 12 (3), 267–286.

Hyland, K., & Tse, P. (2004). Metadiscourse inacademic writing: A reappraisal. Applied Lin-guistics, 25 (2), 156–177.

Hymes, D. (1974). Foundations in sociolinguistics:An ethnographic approach. Philadelphia: Uni-versity of Pennsylvania Press.

Kamler, B. (1994). Gender and genre in early writ-ing. Linguistics and Education, 6 (2), 153–182.

Kieffer, M., Lesaux, N., & Snow, C. E. (inpress). Promises and pitfalls: Implications of



No Child Left Behind for defining, assessing,and serving English language learners. In G.Sunderman (Ed.), Holding NCLB accountable:Achieving accountability, equity, and schoolreform. Thousand Oaks, CA: Corwin Press.

Klein, P. D., & Olson, D. R. (2001). Texts, tech-nology, and thinking: Lessons from the GreatDivide. Language Arts, 78 (3), 227–236.

Lemke, J. L. (1990). Talking science: Language,learning, and values. Norwood, NJ: AblexPublishing.

Michaels, S., & O’Connor, M. C. (2002). Account-able talk: Classroom conversation that works[CD-ROM]. Pittsburgh, PA: University ofPittsburgh.

Ong, W. J. (1982/1995). Orality and literacy: Thetechnologizing of the world. London/New York:Routledge Press.

Pilgreen, J. (2006). Supporting English learners:Developing academic language in the contentarea classroom. In A. Terrel & N. L. Hadaway(Eds.), Supporting the literacy development ofEnglish learners. Newark, DE: InternationalReading Association.

Poole, M. E., & Field, T. W. (1976). A comparisonof oral and written code elaboration. Languageand Speech, 19, 305–311.

Ravid, D., & Tolchinsky, L. (2002). Developinglinguistic literacy: A comprehensive model.Journal of Child Language, 29 (2), 417–447.

Reznitskaya, A., Anderson, R., Nurlen, B.,Nguyen-Jahiel, K., Archodidou, A., & Kim,S. (2001). Influence of oral discussion on writ-ten argument. Discourse Processes, 32, 155–175.

Scarcella, R. (2003). Academic English: A concep-tual framework. The University of CaliforniaLinguistic Minority Research Institute. Tech-nical Report 2003–1.

Schleppegrell, M. J. (2001). Linguistic features ofthe language of schooling. Linguistics and Edu-cation, 14 (4), 431–459.

Schleppegrell, M. J. (2007). The linguistic chal-lenges of mathematics teaching and learning:A research review. Reading & Writing Quar-terly: Overcoming Learning Difficulties, 23 (2),139–159.

Schleppegrell, M. J., Achugar, M., & Oteiza,T. (2004). The grammar of history: Enhanc-ing content-based instruction through a func-tional focus on language. In TESOL Quarterly:A Journal for Teachers of English to Speakers ofOther Languages and of Standard English as aSecond Dialect, 38, 67–93.

Schleppegrell, M. J., & Colombi, M. C. (2002).Developing advanced literacy in first and sec-ond languages: Meaning with power (pp. 1–20).Mahwah, NJ: Lawrence Erlbaum Associates.

Schleppegrell, M. J., & de Oliveira, L. C. (forth-coming). An integrated language and con-tent approach for history teachers. Journal ofEnglish for Academic Purposes.

Snow, C. E. (1990). The development of defi-nitional skill. Journal of Child Language, 17,697–710.

Snow, C. E., Cancini, H., Gonzalez, P., &Shriberg, E. (1989). Giving formal definitions:An oral language correlate of school literacy.In D. Bloome (Ed.), Classrooms and literacy(pp. 233–249). Norwood, NJ: Ablex.

Spolsky, B., & Hult, F. M. (2007). Handbook ofeducational linguistics. Oxford: Blackwell Pub-lishing.

Swales, J. (1990). Genre analysis: English in aca-demic and research settings. Cambride: Cam-bridge University Press.

Taeschner, T. (1983). The sun is feminine: A studyon language acquisition in bilingual children.Berlin: Springer-Verlag.

Tolchinsky, L. & Aparici, M. (2000). Is writ-ten language more complex than spoken lan-guage? International literacy project workingpapers in developing literacy across genres,modalities, and languages, Vol. III. Barcelona:Institute of Educational Sciences.

Wignell, P., Martin, J. R., & Eggins, S. (1993). Thediscourse of geography: Ordering and explain-ing the experiential world. In M. A. K. Hal-liday & J. R. Martin (Eds.), Writing science:Literacy and discursive power (pp. 136–165).London: Falmer (Critical Perspectives on Lit-eracy and Education); Pittsburgh: Universityof Pittsburgh Press (Pittsburgh Series in Com-position, Literacy, and Culture).

Gui A

Documents

delineating academic

term academic language

academiclanguage forms

language of schooling

language speakers of

academic disciplines

language of education

academic english bailey