Theories of second language

acquisition: Three sides, three angles,

three points

Author: Margaret Thomas

Persistent link: http://hdl.handle.net/2345/3223

This work is posted on eScholarship@BC,

Boston College University Libraries.

Post-print version of an article published in Second Language Research 21(4): 393-414.

doi:10.1191/0267658305sr258ra.

These materials are made available for use in research, teaching and private study,
pursuant to U.S. Copyright Law. The user must assume full responsibility for any use of
the materials, including but not limited to, infringement of copyright and publication rights
of reproduced materials. Any materials used for academic research or otherwise should
be fully credited with the source. The publisher or original authors may retain copyright
to the materials.
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra

Theories of second language acquisition: Three sides, three angles, three points

Margaret Thomas

Program in Linguistics

Slavic & Eastern Languages

Boston College

Chestnut Hill MA 02467

U.S.A.
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra

Abstract

Three recent books take up different positions in the on-going debate about how,

and out of what, to construct a theory of second language acquisition. Johnson (2004)

advocates a “dialogically based approach,” inspired by Vygotsky’s sociocultural theory

and Bakhtin’s “dialogized heteroglossia,” with which she would replace what she views

as a prevailing “cognitive bias” in the field. Block (2003) similarly supports a “more

interdisciplinary and socially informed orientation” to second language acquisition. But

Block wants to reform rather than replace certain assumptions of what he represents as

the best existing theory of second language acquisition, namely, Susan Gass’ Input-

Interaction-Output model. Jordan (2004), on the other hand, argues forcefully that

theorizing about second language acquisition must be based on a rationalist

epistemology. He provides a set of “Guidelines” for theory construction, including six

assumptions foundational to rationalist inquiry in general, and a five-point evaluation

metric against which rival theories can be judged. He also passes on a list of six

“practices to be avoided.” Jordan encourages the cultivation of many, varied, theories so

long as they observe the rationalist Guidelines. He goes on to criticize a broad sample of

L2 research, commenting on whether specific proposals do or do not adhere to the

Guidelines. This article reviews all three scholars’ positions in this important debate,

which has the potential to sharpen second language theorists’ sense of what they are

doing and how they should do it.

 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
3

Review article

Theories of second language acquisition: Three sides, three angles, three points

Margaret Thomas Boston College

Block, D. 2003: The social turn in second language acquisition. Edinburgh: Edinburgh

University. In the series Edinburgh textbooks in Applied Linguistics. viii + 162 pp.

UK£19.99. ISBN (EU) 0 7486 1552 0 (paperback). Simultaneously published by

Georgetown University, Washington DC. US$26.95. ISBN (US) 0 87840 144 X

(paperback).

Johnson, M. 2004: A philosophy of second language acquisition. New Haven, CT: Yale

University. ix + 207 pp. US$40.00. ISBN (US) 0 300 10026 4 (paperback).

Jordan, G. 2004: Theory construction in second language acquisition. In the series

Language learning and language teaching: Vol. 8. Amsterdam / Philadelphia: John

Benjamins. xviii + 294 pp. €36.00. ISBN (EU) 90 272 1706 8 (paperback). US$42.95.

ISBN (US) 1 58811 482 1 (paperback).

These three books converge, from three different directions, on the matter of what

should constitute a theory of second language (L2) acquisition, and how to go about
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
4

constructing such a theory. Together they frame their shared object of interest to form a

triangle of three unequal sides with no one book equidistant from the other two. And

although none of the three authors cites either of the other two books, each one writes

with the positions of the others in view, so that together they afford three perspectives on

how to theorize L2 acquisition, each cognizant of the others. The resulting three-sided,

three-angled configuration defined by three points is not inevitable, since it would be

possible to juxtapose any number of other contributions to this long-running debate, as

there have been many, which have held many relationships with respect to each other. i

But the reflections of Johnson, Block, and Jordan on L2 theories and theory-making

indicate something of the range of the debate. Coincidentally, all three authors organize

their expositions around triads of various kinds, so that the books comprise a threesome

constructed of multiple sets of triplets.

I Johnson (2004): Three sides

Marysia Johnson starts her argument with what she defines as three parties to a

“hierarchy of power and control of knowledge in SLA” (p. 2), namely

theoreticians/researchers, teachers/testers, and learners. She objects to giving priority to

the contributions of theoreticians over those of teachers over those of students, and calls

for “a new model . . . in which all participants have equal status, privileges, and rights”

(ibid). But this re-distribution of power cannot be realized within the existing L2

research tradition, which Johnson characterizes as “linear” (p. 3), and invested in a

“conduit metaphor of knowledge transfer” and “a false belief in the existence of a

unidimensional reality” (p. 4). Johnson rejects what she calls the mainstream “cognitive-
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
5

computational tradition” (p. 11) of L2 research, with its “strong cognitive and

experimental bias” (p. 5), its commitment to abstract linguistic competence over real-life

linguistic performance, and the high value it places on the quantitative research methods

of the natural sciences. Johnson’s goal is to analyze the varied shortcomings of this

tradition, and to argue for the superiority of a sociocultural theory of L2 learning derived

from the writings of Vygotsky and Bakhtin. Her version of sociocultural theory locates

L2 acquisition “not in the human mind but in locally bound dialogical interactions” (p. 4).

It values attending to individuals’ diversified experiences instead of group norms (p. 16).

This is because sociocultural theory accepts the existence of a “hyperdimensional social

reality [where] many voices need to be acquired and accepted” (p. 5). In this sense

Johnson portrays it as open to the contributions of all three parties concerned with

understanding L2 acquisition: theorists, teachers, and learners.

Having motivated sociocultural theory on these synchronic grounds, Johnson

situates it historically as the third of three successive traditions of L2 research. Chapters

1 through 5—about half of the book—address the first two, behaviorism and the

cognitive-computational tradition. As is conventional, Johnson joins behaviorism to

Bloomfieldian structuralism and Fries’ and Lado’s contrastive analysis, then juxtaposes

contrastive analysis to the works of Corder, Selinker, and Dulay and Burt, taken as

products of early generative attacks on structuralist assumptions about language and

language learning. ii Although the application of behaviorism in theories of language

learning is no longer viable, Johnson sees its legacy living on in the experimental

methods of cognitive-computational research. She claims that those methods follow from

a positivist philosophy of science, in that they extract data from objectified subjects,
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
6

manipulating and measuring subjects’ responses according to pre-established criteria (pp.

10–11). Non-conforming responses are discarded or at best marginalized. Moreover,

Johnson criticizes research that analogizes L2 learning to the operation of a computer.

To employ language like “data,” “input,” “intake,” “output,” “processing,” or “storage”

in the context of L2 learning is to adopt an asocial, mechanistic, orientation where the

learner is “a loner in an artificially-created social context . . . described in terms of stable

features defined a priori” (p. 85). Johnson attributes these assumptions to Kevin Gregg

(“one of the staunchest proponents of Chomsky’s linguistic theory” p. 37) and equally to

Robert Bley-Vroman, whom she depicts as critical of generativism; to Stephen Krashen;

to Michael Long and others inspired by his Interaction Hypothesis; to Bill VanPatten in

his research on input processing; to Susan Gass and Larry Selinker and their Input and

Interaction model; and to scholars of other affiliations like Vivian Cook and Rod Ellis.

She also finds fault in research that explicitly acknowledges a role for social and

performance-based components of L2 acquisition (Canale & Swain, 1980; Bachman,

1990), on the grounds that it still “present[s] an idealized, homogenized [,] . . . artificial

and abstract” (p. 86) notion of human communication. To Johnson, all of this work

constitutes a cognitive-computational tradition that “advocates the search for

generalizability, the power of statistical procedures, the uniformity of human mental

processes” (p. 14). In doing so, it “projects an image of a human being as a giant

computer, self-sufficient and alone in the material world” (p. 15).

As an alternative to this nightmarish scenario, Johnson offers the work of Lev

Vygotsky (1896–1934) and Mikhail Bakhtin (1895–1975). Chapter 6 introduces

Vygotsky’s life and writings, including his key assertions that cognition originates in
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
7

(and maintains a dialectal relationship with) social experience; that language, as “private

speech,” mediates between cognition and social experience; that the difference between

an individual’s present and potential capacities can be accessed within a “zone of

proximal development”; and, in the Activity Theory developed by Vygotsky’s students,

that analyzing the components of human activities (motives, goals, operations, tools,

interactions, etc.) provides insight into the development of consciousness. Chapter 7

turns to Bakhtin’s counter-Saussurean emphasis on utterances over abstract sentences,

and his claim that the association of typical utterances with specific contexts yields

myriad patterns of language use, which he calls “speech genres.” Speech genres are

necessarily modeled on verbal exchange, so that in acquiring a language, learners

appropriate the voices associated with speech genres, adopting them as inner dialogue.

Bakhtin argued that instead of looking for underlying commonalities, study of language

should investigate this “dialogized heteroglossia” as the basis of cognition.

Chapter 8 summarizes L2 research that has employed Vygotsky’s and Bakhtin’s

ideas. Some examples: Ajaafreh and Lantolf (1994) use the notion of a zone of proximal

development to re-conceptualize L2 fossilization, and to assess the effectiveness of

specific error-correction techniques. Sullivan (2000) concludes that the implicit priority

placed on equality, freedom, and individual choice in pair-work exercises makes

communicative language teaching problematic in the sociocultural world of an L2

classroom in Viet Nam. Gillette (1998) uses the terms of Activity Theory in her

exposition of how students’ varying attitudes toward L2 acquisition affect how they go

about learning, which in turn affects the outcome of their efforts. Johnson also reports

conflicting evidence about whether the use of private speech increases or decreases with
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
8

L2 proficiency, and work that analyzes advanced learners’ reflections on their

development of a new sense of self within an L2 culture.

Chapter 9 reiterates Johnson’s call to build a new, dialogically-based, model of L2

acquisition, attuned to the social, not cognitive, foundations of language learning. That

model would be attentive to multiple local speech situations rather than language

universals, and would employ methods that focus on the specific experiences of

individuals, not (to use an expression from Chapter 1, p. 16) “normalized and

homogenized” group means. Johnson concludes with some recommendations for

improving L2 teaching and testing by the light of the sociocultural theory of L2

acquisition she promotes.

While it is clear that Johnson’s objective is to raise the profile of an approach to

L2 acquisition based on Vygotsky’s and Bakhtin’s writings, a reader has to work to

understand how she positions sociocultural theory within the field. Johnson declares in

several places that sociocultural theory has the power to “unite” the study of L2

acquisition: unite the existing “divergent views of SLA” (p. 1), unite “theory, research,

teaching, and testing” (p. 17), unite “the two divergent traditions: the cognitive and the

social” (p. 188; also, p. 45), and even “merge together” L2 learners’ “external and

internal realities” (pp. 170–1). But under the new, united, regime that Johnson envisions,

it is not obvious what role cognitive-computational research might have, since she

globally rejects its epistemology, goals, methods, and results. Alongside language about

“unifying” the field of L2 acquisition, Johnson sometimes calls for “replacement” of the

cognitive-computational model or of its components (pp. 169, 179). And in what might

be an unguarded moment, she remarks that adoption of Vygotskian sociocultural theory

 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
9

“would require that we abandon . . . the existence of a general language ability [and] . . .

eradicate the assertion that SLA progresses along a predetermined mental path” (p. 172;

emphasis in the original). Therefore despite Johnson’s assertions about uniting L2

theory, what she has in mind may really be to “cull” it of the cognitive-computational

approach, to use the loaded term of Long (1993, pp. 225 et seqq.) while reversing the

direction of Long’s critique.

Johnson’s remark about “abandoning” and “eradicating” is telling. Sociocultural

theory values listening to many voices, but notwithstanding her assertions about uniting

L2 theory, she seems singularly intolerant of theoretical heteroglossia. Moreover, for an

advocate of a “dialogic” approach, Johnson is oddly unprepared for dialogue. It is

essential to dialogue that one make a sincere and patient effort to listen to one’s

conversational partner. But although Johnson champions listening to the voices of

theorists, teachers, and learners, she doesn’t seem to hear what “cognitive-computational”

research has to say.

This is revealed in several ways. First, it is a ground rule of conversation that one

calls one’s partners by whatever name they choose. Therefore it is salient that Johnson

resorts to inventing the label “cognitive-computational” to name an approach she

opposes. Those who take that approach—which by Johnson’s lights include Gregg,

Bley-Vroman, Krashen, VanPatten, and Swain, inter alia—don’t use that label for

themselves. This is in part because they perceive important differences among

themselves that no single such cover term would honor. It is also because whatever

common denominator might be located in the work of alleged cognitive-

computationalists, that would constitute a rather paltry basis for defining group
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
10

membership, perhaps akin to identifying cognitive-computationalists as those who value

some variety of empirically-based research, a characterization which would likely apply

to certain socioculturally-oriented scholars as well.

Second, it is similarly axiomatic that conversation requires one to try hard to

understand whatever one’s partner values, and why, no matter how alien it may seem.

Johnson displays little such effort. For example, although Chapter 8 includes step-by-

step accounts of more than a dozen studies that draw on sociocultural theory, in Chapter 3

she passes over generative L2 research with the remark that “Flynn (1987), for example,

claims that adult L2 learners have full access to [Universal Grammar]. White (1989),

however, believes that L2 learners only have access to the parameters that have been

activated in their first language” (p. 41). Johnson’s use of verbs is noteworthy: Flynn

“claims”; White “believes”; earlier and later in the same passage, Gregg “sees . . . claims

. . . recommends”; Bley-Vroman “considers . . . proposes”; Felix “agrees.” What is

arresting here is the virtual absence of references to twenty-five years of data. Instead,

Johnson represents this stream of research as if it were carried out by the exchange of

speech acts: claim and counterclaim; statement of belief; consideration, proposal,

recommendation; agreement, disagreement. But it isn’t. Generative L2 theorists, along

with most work in the cognitive-computational tradition, perceive their research as driven

forward by the dogged gathering and interpretation of strategically defined empirical

data. Johnson makes clear her distaste for the methods of cognitive-computational

research, but insofar as that distaste prevents her from trying to understand the results of

this research, and why those who gather it value it so highly, so far real dialogue cannot

take place.
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
11

One might add that Johnson reports data collected in some of the research

presented in Chapter 8, and interprets those data as support for socioculturally-oriented

inquiry into L2 acquisition (granted that the role of data in this tradition is more often to

illustrate than to confirm or refute a hypothesis). Johnson does not discount data across-

the-board as immaterial to theory construction; she only neglects to bring forward data

relevant to views she opposes. By downplaying what cognitive-computationalists present

as their signature contribution, Johnson doesn’t seem to try to understand this work on its

own terms. iii

What is more, Johnson sometimes doesn’t seem to be listening to herself. The

unclarity of whether she wants to unite cognitive and social approaches, or replace the

former with the latter, is one inconsistency in her own voice. Another example lies in the

gap between her assertion on p. 18 that in behaviorism “language learning (whether first

or second) was considered to adhere to the same principles,” and the text she cites on p.

23 from Charles Fries—whom she associates with behaviorism—that “Learning a second

language . . . constitutes a very different task from learning the first language.” Johnson

is not the first to fail to hear the dissonance of these two claims, since the identification of

Fries with behaviorism is fully conventional (Thomas, 2004). But it is disappointing that

Johnson, as an outsider to cognitive-computational tradition who prizes individual

differences that others gloss over, cannot break through to perceive the “heteroglossia”

underlying these two remarks.

On the other hand, Johnson’s book makes for some good reading. One virtue is

that she writes clearly, managing to produce a helpful introduction to sociocultural theory

in L2 acquisition that evades the notorious unreadability of many postmodernist tracts

 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
12

(Gregg, 2000). Another is that some of the work she brings attention to is genuinely

rewarding. For instance, I found Sullivan’s (2000) analysis of ‘Playfulness as mediation

in communicative language teaching in a Vietnamese classroom’ persuasive: pair-work

probably can’t succeed as intended in this L2 learning context, for exactly the reasons

Sullivan indicates. As in much of the research Johnson presents, Sullivan observes a

phenomenon, and then analyzes it using sociocultural categories and terms. But the fact

that sociocultural categories and terms can be used in this way does not render them

inevitable. Nor does anything in Sullivan’s lively and thoughtful description convince

me that this kind of analysis of what went on in a particular classroom should replace, or

could “eradicate,” research into (say) developmental sequences in the acquisition of L2

argument structure.

In their methodologies and assumptions about what is of greatest value in L2

acquisition, the sociocultural and cognitive-computational traditions are, in Beretta’s

(1991) word, oppositional. But—at least as presented by Johnson—in other ways they

seem complementary, in the sense that they don’t have enough in common to be treated

as rivals. Most obviously, this is because one is concerned with the social dimensions of

L2 learning, and the other with cognition. iv In addition, the cognitive-computational

tradition invests first and foremost in explaining acquisition. Johnson sometimes declares

that sociocultural research aims to explain something, for example, when she writes that

it would “focus on identifying, describing, and explaining all possible speech genres” (p.

173). But the work she cites mostly “investigates”: “investigat[es] the effects of the

various speech genres on the learner’s second language ability” (p. 173); “investigat[es]

the processes that lead to becoming an active participant in locally bond social contexts”
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
13

(p. 179). In fact, because sociocultural research abstains from generalizing or

universalizing, it is not clear how it could fashion an explanation that transcended

accounts of individuals’ experiences. For these reasons, sociocultural and “cognitive-

computational” research do not assume the same burdens. If theories necessarily attempt

explanation—as Jordan (2004), among others, believes they should—then in a technical

sense sociocultural research cannot constitute a theory. Rather, it investigates the social

and interactional domain of acquisition within what Johnson aptly labels in her title a

“philosophy” of second language acquisition.

II Block (2004): Three angles

If Johnson, Block, and Jordan triangulate the issue of theorizing L2 acquisition, the

shortest side of the triangle lies between Johnson and Block. David Block shares

Johnson’s zeal for “more interdisciplinary and socially-informed (or socially sensitive)”

research, expressions that together appear 29 times in the 8 pages of Block’s Preface and

Chapter 1. His book synthesizes support for that “social turn,” about which he feels

discussion to date has not been sufficiently constructive. Block differs from Johnson in

that he insists that his aim is to “circumvent exclusionary stances” (p. 7) so that the social

turn he looks forward to will change the boundaries of existing L2 theorizing but not

replace it.

Chapter 2, “A short history of SLA,” displays an important characteristic of the

book, its heavy reliance on secondary sources. The first part of Chapter 2 is mostly built

out of textbooks and digests of the field, notably Gass and Selinker (2001), with

references to two dozen other surveys or synthetic overviews. Because Block’s sources
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
14

largely agree about what constitutes the history of L2 acquisition, he concludes that

therefore they are accurate. One might be cautious about conducting business this way,

especially since some of these texts do what Block does, that is, cite each other as

authoritative. Moreover, comparison of how the history of L2 acquisition is treated in the

textbooks and digests that Block relies on reveals that they converge on a relatively small

sample of primary sources. This suggests that what counts as historical background has

been conventionalized, making it easy to mistake what has been written about what

happened for what really happened.

In another instance of reaping what others have sown, Block finishes the chapter

by paraphrasing different scholars’ views of what a theory of L2 acquisition must account

for. Block concludes that the Input-Interaction-Output model he attributes most centrally

to Susan Gass (Gass, 1997; Gass & Selinker, 2001; Long, 1996) is the “the closest thing

that we have to a ‘big’ theory to date” (p. 26), and “the most tangible result of 30 years of

. . . intensive research into how individuals learn second languages” (p. 30). Block

reproduces as the gist of the IIO model a figure appearing in embryonic version in Gass

1988 (p. 200), then developed in Gass 1997 (p. 3) and Gass and Selinker 2001 (p. 401).

The figure summarizes proposals about how “Apperceived Input” relates to

“Comprehended Intake” and eventually to “Output,” realized as a kind of flow chart

consisting of labeled boxes. The boxes are connected by arrows representing claims

about the relationships of factors bearing on L2 acquisition such as “Affect,” “L1

knowledge,” and “Hypothesis testing.”

Block places the IIO model at the center of his critique of L2 acquisition theory.

He then organizes that critique around discussion of the meanings of the three terms
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
15

“second,” “language,” and “acquisition.” Block analyzes what he perceives as

disciplinary narrowness and social insensitivity in the field in general, and in the IIO

model in particular, from these three angles.

From the first angle, Block objects to the “second” in “second language

acquisition,” on several grounds. “Second” downplays the multilingual complexity of

many learners’ real experiences, because exposure to an L2 can destabilize L1

knowledge, and because speakers often move fluidly among far more than two codes

within an idiosyncratically-bounded “mass of linguistic competence” (p. 42). IIO-

oriented studies suppress these complexities to take the “S” in “SLA” at face value.

Block also questions the appropriateness of contrasting “second” versus “foreign” versus

“naturalistic” language learning environments, citing evidence that the local context may

belie conventional notions of how input to learners differs in these three environments.

Block concludes that the critical determinant of success or failure is “how the individual

learner negotiates and carves out an identity in the target language” (p. 55). He concedes,

however, that many researchers de-prioritize these matters as a “clutter of variability” (p.

56), concluding pessimistically that the “S” in “SLA” will probably not yield to his

favored expression, “additional language acquisition.”

Taking a second angle, in Chapter 4 Block advocates revising our sense of the

“L” in “SLA.” In particular, he finds fault in the IIO-propagated notions of “task” and

“negotiation for meaning.” Because language use is not limited to information exchange,

L2 pedagogy that relies on picture-description or problem-solving tasks ill prepares

learners for the range of real discourse. What is more, exercises designed to engage

students in negotiation for meaning artificially downplay the social context of language
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
16

use, wherein negotiation of solidarity, of face, and of identity complexify why and how

real people talk. Like Johnson, Block argues that by idealizing language, much study of

L2 acquisition marginalizes important social factors. But Block is not trying to do away

with research which primes the linguistic over the social sense of “language,” only trying

to call attention to what a linguistic orientation misses. As in Chapter 3, he admits that

scholars may legitimately define their work outside of social factors (pp. 84, 86, 90).

Block’s assessment of the “A” in “SLA” recapitulates some of Johnson’s critique

of mechanical information-processing models, artificial experimental methods, and

aggregate data, and depicts Vygotsky’s and Bakhtin’s work as an alternative. v Block

suggests that acquisition be re-conceptualized as a process of participation, or of

becoming, wherein learners’ affects and attitudes are studied as keys to the process of

entering into an L2 “community of practice” (p. 113). He illustrates his argument by

analyzing an interview with a Catalan-speaking learner of English, whose experiences in

two different “communities of practice”—in a foreign-language classroom, and as a

visitor to London—show that her social sensibilities shaped her capacity to take

advantage of the language learning opportunities she encountered. As usual, Block ends

the chapter by evincing doubt that L2 theory will re-define “acquisition” so as to accept

the centrality of social and attitudinal factors.

The book closes with a catalog of what textbooks or overviews of the field have

predicted about the future of L2 acquisition. Block iterates his reservations about

whether a “social turn” will take place, although he counts his book as evidence that if

that turn were to take place, it would substantially improve our understanding of how

people acquire an L2.

 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
17

Block thus shifts, and expands, the definitions of “S,” “L,” and “A,” to include

social and interactional issues in each case. Compared to Johnson, Block offers rather

more of an insider’s view of “mainstream” L2 acquisition theory, in all of its asocial,

essentializing, mechanistic, glory; and he emphasizes that he wants to supplement, not

displace, the status quo, even as he resigns himself to the continued marginalization of

sociocultural theory. That undercurrent of doubt flowing beneath the surface of the text

makes Block seem to anticipate that the upshot of his work will be, at best, something

more like a feint than a wholehearted social turn. Perhaps this only constitutes a display

of polite pessimism with regard to the success of his face-threatening proposals. Or

perhaps Block himself is not fully committed to his claim to have demonstrated the value

of “more interdisciplinary and socially-informed” study of L2 acquisition.

For this reader, neither Johnson nor Block succeeds in making a case that L2

theory must be redefined to incorporate sociocultural issues. Sociocultural research

draws attention to intriguing facts about L2 acquisition that otherwise might not be

brought to light, but those facts do not constitute a theory that challenges the validity of

mainstream L2 acquisition research. For example, in the chapter on the “L” in “SLA,”

Block gives an extended commentary on Mackey, Gass, and McDonough’s (2000)

analysis of how learners interpret feedback in stimulated recollections of conversations

with native speakers. He speculates that Mackey et al.’s data may reveal more than the

researchers were prepared to hear, arguing that they inadequately investigated the impact

on their data of gender, language affiliation, and negotiation of identity (p. 82). Block

writes “I can think of other things that might be going on” (p. 86), a remark that seems to

sum up his approach. Block circumnavigates research on “SLA”, stopping three times to
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
18

call attention to “other things that might be going on.” No doubt there are lots of other

things going on: we can probably take for granted that, in the words of Wagner and

Gardner (2004, p. 15), second language speakers “engage in quite exquisite

[conversational] activities.” Quite exquisite as Block’s other things are, they don’t

constitute a challenge to Mackey et al.’s analysis, nor do they overcome Michael Long’s

skepticism that “a richer understanding of [learner identities], or . . . social context,

makes a difference, and a difference not just to the way this or that tiny stretch of

discourse is interpreted, but a difference to our understanding of acquisition” (Long 1998,

quoted by Block, pp. 7, 136). At base, Block himself seems to recognize this fact, in that

he repeatedly communicates doubt that researchers working outside sociocultural theory

will be persuaded of its value. Thus he gives leeway to non-socioculturally-informed

research to proceed at will. vi The “other things that [Block thinks] might be going on”

are in a complementary, not oppositional, relation to work that theorizes the syntax,

morphology, lexis, and phonology of second language acquisition.

III Jordan (2004): Three points

If Johnson’s and Block’s books connect the short side of a triangle, Geoff Jordan

writes from a position considerably farther away compared to the distance that separates

the other two. In doing so, Jordan lends perspective to Johnson’s and Block’s arguments

by bringing into view some of the context surrounding them.

Part 1, comprising Chapters 1 through 5, provides a defense of what Jordan

identifies as a rationalist approach to theory construction. vii Following a review of the

terms and points of controversy that previous debate about theory construction in L2
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
19

acquisition has brought forward, Jordan launches into a spirited exposition of western

philosophy of science, from Cartesian rationalism and Baconian empiricism, through

Hume and the Vienna Circle, on to Popper, Kuhn, Feyerabend, Lakatos, and Laudan. viii

Jordan emphasizes the creative tension between two methods in the development of

scientific theories, characterized as Baconian “research-then-theory” induction versus

Cartesian “theory-then-testing” deduction. Along with Popper (but not without

acknowledging Popper’s critics) Jordan concludes that, since we cannot prove theories to

be true, only show that they escape disconfirmation, therefore “the deductive method is

the true method of science, and the role of observation and experimentation is to test our

hypotheses” (p. 47).

Next Jordan connects the fertile late twentieth-century debate in the philosophy of

science to the rise of relativism. Relativism challenges rationalism (and the realist

epistemology rationalism assumes), on the grounds that there is no objective reality that

science can investigate and hence no independent standards for evaluating opposing

theories. Jordan covers developments in the sociology of science influenced by

relativism; the radical postmodernist critique of science; and the more tractable relativism

of constructivists, among whom he identifies Vygotsky. He accepts postmodernists’ and

constructivists’ political claims, that an entrenched élite protects its disproportionate

power, resulting in injustices of many kinds. But he objects that when it comes to

building a theory of L2 acquisition, relativism has nothing useful to offer in place of

rationalism. To Jordan, the most reliable knowledge about the world comes from

developing an explanatory theory according to rules of logic, “scrutinizing [it] so as to

discover flaws in terminology or reasoning” (p. 81), then avidly assessing and reassessing
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
20

that theory’s capacity to explain phenomena observed in the environment. This is the

core content of rationalism, and to Jordan theory construction in L2 acquisition requires a

rationalist basis. Rationalism is not, however, to be confused with science. What counts

as science has a broader scope and less precise boundaries; scientists may test theories

rationally, but arrive at them through various means.

A long Chapter 5 is the heart of Jordan’s book. He summarizes the case for

rationalism, then evaluates four existing views of what makes an adequate theory of L2

acquisition. Jordan first criticizes relativists for failing to distinguish between two

separate complaints: against disciplinary narrowness (about which Jordan feels relativists

should be free to make their case), and against prioritizing rationalism as the key to L2

theory construction (a complaint Jordan flatly rejects). To Jordan, insofar as relativists

investigate L2 acquisition atheoretically, or claim that no theory is intrinsically superior

to any other, so far they do not contribute substantively to the discipline, even though

their calls to increase interdisciplinarity and attend more to the local social context of L2

learning may have independent merit. ix Moving on to the writings of Kevin Gregg,

Michael Long, and Barry McLaughlin, Jordan represents their diverse ideas about the

contents of L2 theories as much more constructive, although he still finds room for

improvement in each case.

All this sets the stage for the “Guidelines” Jordan presents at the end of Chapter 5

(pp. 114–118), a tri-partite set of principles for evaluating candidate theories of L2

acquisition. First, Jordan formally states six assumptions: the “minimally realist

epistemology” that an external world exists and can be studied; that research cannot be

separated from theory; that theories explain phenomena; that research attempts to solve
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
21

problems; that a unique scientific method cannot be formalized; that we need many

theories, not a single paradigm. Second, Jordan specifies five criteria for evaluating

theories. Theories should: be coherent, cohesive, and clear; have empirical content; be

fruitful; be broad; be simple. Third, Jordan lists six practices and characteristics to be

avoided, as indicative of “pseudoscience”: too-casual approach to evidence; lack of

falsifiability; failure to explain; attempts to derive writers’ “real” meanings by

interpreting their language; refusal to acknowledge criticism; and predilection for obscure

prose. x Jordan presents his Guidelines as a tool for discerning what works and what

doesn’t among attempts to theorize L2 acquisition. In his opinion, scholars who accept

the Guidelines form a research community whose business it is to create more, and more

daring and varied, theories and then to submit those theories to rigorous critique

according to rationalist principles.

In Part 2 of his book, Jordan wields the tool of his Guidelines to assess existing

attempts at L2 theory construction. He separates those attempts into three groups:

generativist-inspired theorizing about Universal Grammar in L2 acquisition; “Approaches

to SLA that offend the Guidelines”; and “Signs of progress,” theories that come closer to

meeting Jordan’s critieria.

Jordan’s treatment of generative research in L2 acquisition stretches over two

chapters. Chapter 6 introduces Chomsky’s work (in curiously elementary terms, beneath

the level of sophistication presupposed by the rest of the text) and reviews points raised

by a sample of three of Chomsky’s critics (Jean Piaget, Geoffrey Sampson, and Elizabeth

Bates). Chapter 7 assesses generative L2 theory. Although Jordan makes it clear that he

considers Chomsky a thoroughgoing realist, wholly committed to rationalist

 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
22

methodology, he declares that “[L2 theorizing that is based on] UG does not measure up

well at all to the . . . Guidelines” (p. 151). Most of the chapter is taken up reporting what

others have written pro and con generative L2 theory. But looking ahead to the end of

the book where Jordan returns to the issue, his key objections are three:

(1) UG is of little use in describing the knowledge involved in SLA, since most

[of it] fall[s] outside the UG domain

(2) UG is of no use in explaining the SLA process, since it is a property theory

and thus has nothing to say about any process

(3) the poverty of the stimulus argument has no force in relation to constructing a

theory of SLA since the L2 learner already has a representational system in

place (p. 255)

Point (1) is probably most important to Jordan. He considers generative L2

theory far too narrow, as a consequence of Chomsky’s stance:

Chomsky’s strict demarcation between science and non-science effectively rules

out the study of E-language. Chomsky pays a high price for such a rigorously

scientific theory; [he is forced to adopt] an extremely limited view of what

language is and consequently his theory neither describes nor explains many of

the phenomena that interest linguists, and far less . . . the phenomena of SLA. (p.

156)

Therefore:

Those in the field of SLA who take the [UG] approach . . . can be seen as either

lucky to have a cogent framework to guide their work, or unlucky to be restricted

to such a tiny domain. (p. 255)

 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
23

In singling out generative L2 theory as the first, and only constituent member, of

the three categories of theories he assesses, Jordan emphasizes that generative grammar

has a virtue competitor theories find hard to match: it provides the basis for an adequate

property theory (that is, a theory of what it is that is acquired). But overall Jordan

conservatively assesses generative L2 theory’s capacity to contribute to rationalist

research, because of generative grammar’s limited range. Jordan’s other reservations

have to do with generativists’ lack of a transition theory (a theory of how acquisition

takes place), and his conviction that the poverty of the stimulus does not hold for L2

learners. About the latter point, Jordan adverts briefly to Carroll’s (2001) rejection of the

poverty of the stimulus in L2, then breezily claims that transfer allows adults to acquire

L2 knowledge empirically (pp. 255–256). This matter deserves much more thorough

treatment than Jordan gives it.

Putting generative theory behind him, Jordan moves on in Chapter 8 to proposals

that he judges to be even less in compliance with the Guidelines. These include

postmodernist approaches (dismissed on the basis of his earlier analysis of their

incompatibility with rationalism); contrastive analysis; the ethnography of

communication; Krashen’s Monitor Model; variable competence; acculturalization /

pidiginization; and research on aptitude and motivation in L2 learning. Jordan indicates

how, in his opinion, each of these approaches “offends the Guidelines” in its own ways.

Chapter 9 proceeds on to what Jordan presents as greener pastures, research that

more closely observes (at least some of) the Guidelines. This third category includes

error analysis; morpheme order and other studies of staged development; processing-

based research variously developed by McLaughlin, Schmidt, Long, and Pienemann;

 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
24

Towell and Hawkins’ model; Bates and MacWhinney’s Competition model; and

emergentism. This is not to say that everything treated in Chapter 9 compares favorably

with everything treated in Chapter 8, because in both cases Jordan makes free with

criticism and—occasionally—praise.

The final Chapter 10 veers a bit out of control as Jordan tries to synthesize his

argument, but can’t resist commenting on a few more ideas about L2 acquisition theories

that happened not to fit in earlier. However, the book ends with a neat turn of the screw:

I believe that Popper is essentially right. Problems are the stuff of theories; we

should articulate what the problems are that our theory is going to address, and

then we should fly any kite we like. When we come to evaluate our theory, then

we need to use rational criteria that rest on realist epistemological assumptions.

(p. 265)

Overall, Jordan’s application in Part 2 of the argument he built up in Part 1 comes

across as something of an anticlimax. That is because, first, Part 2 relies heavily on

secondary sources—many of the same standard textbooks and digests that Block relies

on—so that Jordan’s commentaries don’t always seem to derive from firsthand exposure

to the work he is assessing. This doubtless rendered the task of writing the book more

manageable, but it likely limited Jordan’s creativity: a pity. Second, in Chapters 8 and 9

Jordan doesn’t always go deeply enough into specific proposals about L2 acquisition to

justify his evaluations or explore their ramifications. xi Even in the case of generative L2

theory, to which he devotes 42 pages, Jordan leaves hanging the key issue of the poverty

of the stimulus. In Chapter 8, he dismisses six “offensive” approaches (seven, if one

 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
25

counts his recapitulation of the argument against relativism) in 34 opinionated,

informative, but thin, pages.

IV Conclusion

Stepping back to bring all three books into view, it is worth pointing out that what

Jordan most objects to is not the milder relativism of Block or even Johnson. Both the

latter would probably assent to the existence of an objective external world, and agree

with Jordan that research should attempt to solve problems and that theories should be

fruitful and broad. Nevertheless, neither Block nor Johnson writes as if expecting that a

theory must explain phenomena or be falsifiable. The absence of those core

characteristics is enough to identify Block’s and Johnson’s among approaches that in

Jordan’s view “offend the Guidelines.”

However, it is a larger question whether that only means that Block and Johnson

can’t claim membership among rationalist theorizers of L2 acquisition, or whether that

means their work has no legitimate claim to one’s attention. The architecture of Jordan’s

book suggests how this kind of question could be used to good advantage in graduate

education. One might assign students in a seminar on L2 theory to read and critique

Jordan’s Part 1 as an extended argument for the value of rationalism in L2 theory

construction. Individual students could then be made responsible in depth for one or

more of the “offensive” or “sign-of-progress” theories (or for other proposals about L2

acquisition: Block’s book, Johnson’s book). The students’ first task would be to

assemble and master a bibliography of the relevant primary sources. Eventually, they

would take turns presenting to each other the contents of one or more proposals, and
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
26

justifying their evaluation according to Jordan’s Guidelines, or according to alternative

criteria whose value they can demonstrate. Students could very well exit such a seminar

fortified against the “confusion and misunderstanding” about L2 theory construction that

Jordan laments (p. 3), and that all parties to this debate have a stake in abating.
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
27

Acknowledgement

I am grateful to Kevin R, Gregg and to James P. Lantolf for reading and commenting on

earlier drafts of this text. Neither is responsible for any opinions it expresses, or errors it

contains.
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
28

References

Ajaafreh, A. and Lantolf, J. P. 1994: Negative feedback as regulation and second

language learning in the zone of proximal development. Modern Language

Journal 78, 465–83.

Bachman, L. 1990: Fundamental considerations in language testing. Oxford: Oxford

University.

Beretta, A. 1991: Theory construction in second language acquisition: complementarity

and opposition. Studies in Second Language Acquisition 13, 451–512.

Block, D. 1996: Not so fast: some thoughts on theory culling, relativism, accepted

findings, and the heart and soul of SLA. Applied Linguistics 17, 53–83.

Canale, M. and Swain, M. 1980: Theoretical bases of communicative approaches to

second language teaching and testing. Applied Linguistics 1, 1–47.

Carroll, S. E. 2001: Input and evidence: the raw material of second language

acquisition. Amsterdam / Philadelphia: John Benjamins.

Casti, J. L. 1989: Paradigms lost. London: Abacus.

Gass, S. M. 1988: Integrating research areas: a framework for second language studies.

Applied Linguistics 9, 198–217.

Gass, S. M. 1997: Input, interaction, and the second language learner. Mahwah, NJ:

Lawrence Erlbaum.

Gass, S. M. and Selinker, L. 2001: Second language acquisition: an introductory

course (2nd ed.). Mahwah, NJ: Lawrence Erlbaum.

 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
29

Gillette, B. 1998: The role of learner goals in L2 success. In Lantolf, J. P., and Appel,

G., editors, Vygotskian approaches to second language research. Norwood, NJ:

Ablex, 195–213.

Gregg, K. R. 2000: A theory for every occasion: postmodernism and SLA. Second

Language Research 16, 383–99.

Gregg, K. R. 2002: A garden ripe for weeding: a reply to Lantolf. Second Language

Research 18, 79–81.

Gregg, K. R. 2003: SLA theory: construction and assessment. In Doughty, C. J., and

Long, M. H., editors, The handbook of second language acquisition. Malden,

MA: Blackwell, 831–65.

Lantolf, J. P. and Appel, G. 1994: Theoretical framework: an introduction to

Vygotskian perspectives on second language research. In Lantolf, J. and Appel,

G., editors, Vygotskian approaches to second language research. Norword, NJ:

Ablex, 1–32.

Lantolf, J. P. 1996: SLA theory building: “letting all the flowers bloom!” Language

Learning 46, 713–749.

Lantolf, J. P. 2000: Introducing sociocultural theory. In Lantolf, J., editor,

Sociocultural theory in second language learning. Oxford: Oxford University, 1–

26.

Lantolf, J. P. 2002: Commentary from the flower garden: responding to Gregg, 2000.

Second Language Research 18, 72–78.

Long, M. H. 1993: Assessment strategies for second language acquisition theories.

Applied Linguistics 14, 225–249.

 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
30

Long, M. H. 1996: The role of the linguistic environment in second language

acquisition. In Ritchie, W. C., and Bhatia, T. K., editors, Handbook of second

language acquisition. San Diego, CA: Academic Press, 413–68.

Long, M. H. 1998: SLA: breaking the siege. University of Hawai’i Working Papers in

ESL 17, 79–129.

Mackey, A., Gass, S, and McDonough, K. 2000: How do learners perceive interactional

feedback? Studies in Second Language Acquisition 22, 471–97.

Sullivan, P. 2000: Playfulness as mediation in communicative language teaching in a

Vietnamese language classroom. In Lantolf, J., editor, Sociocultural theory in

second language learning. Oxford: Oxford University, 115–31.

Thomas, M. 2004: The ‘Ebeling Principle’ in the historiography of generative second

language acquisition theory. Manuscript submitted for publication.

Wagner, J. and Gardner, R. 2004: Introduction. In Gardner, R. and Wagner, J.

(editors), Second language conversations. London: Continuum, 1–17.

 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
31

Footnotes

i
See the special issues of Applied Linguistics Vol. 14 No. 3 (1993), and The Modern

Language Journal Vol. 81, No. 3 (1997), with subsequent rejoinder and surrejoinder

(continuing into Applied Linguistics Vol. 15 No. 3 and Vol. 17 No. 1, and into The

Modern Language Journal Vol. 82, No. 1). Also see the exchange between Lantolf

(1996, 2002) and Gregg (2000, 2002). All three texts under review cite some of this

work. Block and Jordan do so extensively.

ii
Thomas (2004) analyzes various problems in this standard presentation of the

relationships of behaviorism, American structuralism, contrastive analysis, creative

construction, and early generative grammar. It should be noted that Block and Jordan

posit essentially the same problematic relationships.

iii
In another example of failure to attend to what matters to one’s opponents, Johnson

only mentions in passing two key themes taken up by many in theorizing L2

acquisition: the creativity of everyday language use, and the poverty of the stimulus.
iv
Although Johnson discusses Vygotsky’s writings on human mental processes, the

sociocultural theory she advocates emphasizes that L2 ability “is not situated in the

learner’s mind but . . . in sociocultural and institutional settings and in a variety of

discursive practice to which the learner has been exposed” (p. 172). This contrasts

with Lantolf and Appel’s (1994) and Lantolf’s (2000) reading of Vygotskian

psychology, which foregrounds his theory of mind.

v
For an introduction to Vygotsky and Bakhtin, go to Lantolf (2000), not Block. Block’s

version is too truncated, as when he explicates Activity Theory by chaining together

 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
32

the sequence of words “Need / Objective / Motive / Goal / Action / Condition /

Operations,” interspersing each word with an arrow pointing to the word on the right

(p. 102). This communicates little about what relations Activity Theory actually posits

between these terms.

vi
This may represent a softening of Block’s stance, compared with the more

confrontational tone and less accommodating position Block (1996) staked out.
vii
Kevin Gregg (personal communication) pointed out that Jordan sometimes uses the

term “rationalism” where “realism” would seem called for, for example in opposition

to relativism or constructivism. Jordan expands the sense of “rationalism” so that it

not only contrasts with empiricism and positivism as a research methodology that

prioritizes deduction over induction, but also labels an orientation that accepts the

existence of an independent world liable to scientific inquiry. Jordan mentions realism

as an initial assumption of rationalism (p. 115). See Gregg (2003) for background.
viii
Don’t skip Jordan’s footnotes to Chapter 2, in which he has buried entertaining first-

hand anecdotes about the contentious late twentieth-century philosophy of science

pantheon, thundering at each other like gods on top of Olympus.

ix
Jordan (p. 98) includes Block and James Lantolf among relativists he criticizes in these

terms, a characterization that seems rather too broad.

x
Jordan attributes to Casti (1989) the first five of the six hallmarks of pseudoscience.
xi
One annoyance is Jordan’s persistent failure to distinguish references to Rod Ellis

versus Nick C. Ellis, despite the fact that both scholars get more than passing attention.

Consulting the index doesn’t help, as it lists a single entry for “Ellis, R., N” (p. 287).
 Post-print version of an article published in Second Language Research 21(4): 393-414.
doi:10.1191/0267658305sr258ra
33

Under that superscript, at least one cited span of pages, pp. 242–244, conflates

references to both men.