The Web Site for Critical Realism The WSCR Archive

[The following paper was originally written for the Ecology of Language Acquisition Workshop, University of Amsterdam, 11-15 January 1999, and is archived here with the kind permission of the author, who retains the copyright. A copy of Critical Realism and Ecological Psychology is also available on Mark Fettes's "Linguistic Ecology" Web site.]

Critical Realism and Ecological Psychology: Foundations for a Naturalist Theory of Language Acquisition


Department of Theory and Policy Studies in Education
Ontario Institute for Studies in Education/University of Toronto

Since the onset of the modern era in Europe, some 300 years ago, theories of language have been strongly influenced by what Mikhail Bakhtin refers to as the "centripetal forces" in society (Bakhtin, 1981 [1935]). The conception of "languages" as unitary and autonomous systems was made plausible by the invention of such modern technologies as the monolingual dictionary and the logic-based grammar (Harris, 1980; 1981), and has been elaborated and reinforced by ideological practices of "facticity", i.e. normative ways of producing and interpreting textual representations of the world (Smith, 1990). Such technologies of linguistic normalization have played a key part in the evolution of modern societies, away from diversified, stratified and relatively static forms in which order derives from the ties of kinship, place and mind (Tönnies, 1957), towards fluid and turbulent forms in which order derives from the accumulation of material and cultural capital (Bourdieu, 1990; 1991).

Today modernity is widely regarded as a universal and irreversible process of social development, and theories of language premised on its ideals continue to be held up as models of scientific explanation. Yet as asserted by Bakhtin and elaborated in the postmodern sociology of Zygmunt Bauman, the focus on unity and autonomy has also worked to distract attention from the conflict, change, and uncertainty endemic to all human affairs, in language as elsewhere (Bakhtin, 1981; Bauman, 1992). Among linguists working in English, Paul Friedrich has argued a particularly compelling case for "a more relativistic view" of linguistic order, in which discreteness can make room for continuity, exact meaning for associative meaning, passive reproduction for active creativity, the generic "native speaker" for the unique individual, and "the rage for order" can concede an enduring place for chaos (Friedrich, 1985). For this, we need a theory that explains language in terms of the interaction of knowing and purposeful speakers, rather than those based on premises of structural determinism.

The purpose of this paper is to sketch such a theory and to examine its implications for theories of language acquisition. In the available time, I will have to touch rather lightly on a number of important and contentious issues. I hope that in subsequent discussions and in ongoing contacts following the workshop we will be able to explore these in more detail. My purpose in this initial presentation is to suggest the overall form that a theory of linguistic ecology must necessarily take, particularly in terms of epistemology, the relationship of language to knowledge; and ontology, the kind of thing language is.


In much modern discourse, "mind" is routinely equated with "brain" -- for instance, in the title (and contents) of Stephen Pinker's How the Mind Works (1997). A great deal of work in modern philosophy and psychology has relied upon this idea of mind as a physical organ of thought, Descartes' res cogitans, distinct from both the non-thinking, mechanically reactive body and the external, mechanically determined environment. The mind is contained in the brain, and is itself a container of knowledge: this simple metaphor has infiltrated a huge range of disciplines, so that the concept of "language acquisition", for example, suggests a process whereby an external object, "language", is taken up and held by the mind of the learner.

There are many problems with the container paradigm, of which I will mention only three. First, it has proven impossible to reconcile with philosophical realism, since there is no way to show that "meaning in the head" corresponds to "meaning in the world" (Millikan, 1984). Rather than arguing that realism is at fault, as late twentieth-century philosophers have been wont to do, it seems more plausible to suggest that our basic metaphor of mind is wrong (Ben-Ze'ev, 1995). Second, the container metaphor derogates individual knowledge. People are said either to enter life as blank slates, upon which are gradually inscribed the teachings of their class, their culture, their profession; or they are held to share a universal, genetically inscribed faculty for the construction of meaning. Both forms of knowledge are subject to the prescriptive factual practices of the apostles of legislative reason (Bauman, 1987; Smith, 1990); in this way the container metaphor helps sustain the characteristically modern power/knowledge dynamic identified by Foucault (Foucault, 1980). Third, in its incarnation as the "banking theory" of learning (Freire, 1972) this conception of mind continues to sustain a hugely inefficient and alienating educational system which works to the particular disadvantage of minority and working-class students. Together with David Corson (Corson, 1997), I would suggest that applied linguistics needs to be concerned with all of these issues.

The alternative I will develop in this paper abandons all of the classic dichotomies of Cartesian philosophy: mind--body, spirit--life, human--animal, person--world. It is an ecological or systems perspective, in that the central reality is seen to be relationships between things, the latter simply constituting the structures through which the relationships are realized. One particularly important source of insight has been Edward Reed's ecological psychology, which offers a radically naturalist, non-Cartesian account of mind and agency (Reed, 1996). According to Reed, as life has evolved, natural selection has ensured the development of mechanisms for perceiving and responding to persistently available information in the environment: this he calls "the effort after meaning". People, and frogs, and earthworms, all demonstrate such adaptive awareness of their surroundings -- awareness conceived not as a private internal state, but as a constantly shifting active relationship between the organism and its surroundings. As more complex animals have evolved, they have developed increasingly sophisticated forms of retrospective and prospective awareness: that is, the ability to recall past relationships with the world, compare them with the present, and thereby project the most likely developments in the immediate future. Such ecological knowledge is not given, but achieved by the individual-in-environment, employing and modifying adaptive strategies developed by communities and species over vast stretches of time. Mind is thus not a structure but an active integrative process of seeking for, interpreting and responding to meaning, i.e. adaptively significant forms of order in the world.


Such a redefinition of mind entails a redefinition of knowledge along the lines proposed in the "schema paradigm" (Ben-Ze'ev, 1995). Schemas (or schemata, if you prefer) can be thought of as complex hierarchical neural networks associated with particular states of awareness and action. As an organism grows and accumulates experience, it develops the capacity "to arrive at states similar to its previous states of awareness while preserving a knowledge of their past origin" (retrospective awareness, or memory), and to use this capacity to deepen its awareness of the present situation and its immediate consequences (prospective awareness). Knowledge is thus not "stored", in the sense of being put away in permanent safe keeping, but provisionally "retained", like any other active capacity that declines when not put to regular use (Ben-Ze'ev, 1995).

The schema paradigm emphasizes the holistic nature of perception and learning. Newly acquired capacities are not stored on a separate "shelf" or "warehouse", but through modifications to the cognitive system as a whole: structure and content are one and the same. Learning is thus still more consequential than is usually thought, for it changes what an organism is: what meanings it can be aware of, what values it can attain. By the same token, learning is also a more complex affair, for it involves changes in a cognitive system that is already massively interconnected and integral to a particular way of living in the world. In other words, learners (of language as of other things) are anything but blank slates: they are highly adapted complex systems in their own right, whose negotiation with the learning process will be strongly influenced by their individual styles of encountering the world.

This view of knowledge differs sharply from the idealism inherent in the Cartesian paradigm. First and foremost, it is active knowledge, stemming from the lifelong, continuous effort towards meaning and value that characterizes all living things. Secondly, it is realist knowledge, in the sense that it is derived from direct encounters with the world. This knowledge is tentative and fallible, but adaptive: under certain conditions, it is a reliable guide to awareness and action. Thirdly, it is non-propositional knowledge: not only does it not depend on language, it exists in another medium entirely, one that can be imperfectly visualized with the help of schema theory but is ultimately one and the same as life itself (Capra, 1996; Maturana & Varela, 1992). Such a theory of knowledge satisfies the tenets of Roy Bhaskar's critical realism (Bhaskar, 1989), Dorothy Smith's "insider materialism" (Smith, 1990), and Ruth Millikan's epistemological naturalism (Millikan, 1984); it holds that we are both phylogenetically and ontogenetically adapted to discovering real meanings in the real world.


Our thinking about meaning in the last few decades has been heavily influenced by the cybernetic revolution, where information is simply treated as difference (e.g. (Bateson, 1972; 1979). But the natural world is structured by interaction, at every level from the quantum to the quasar; while our machines can get by on representation, as living, embodied beings we ourselves rely on the perception of relationship. Ruth Millikan (1984) has shown how a basic capacity for awareness of relationship can ground the discovery of enduring substances and the properties that differentiate them; in her account, every such act of identification draws on and modifies a network of schematic associations, or "intensions", derived from past states of awareness. Mark Turner (1996) has extended this idea to the perception of action and change, or what he calls "story". To take one of his examples, every time we reach out to pick up a glass the actual event is different in small details. Yet through recurrent sensory and motor experience, we acquire a narrative schema that picks out the typical features of the acts of picking up a glass: the way the glass looks, the way it feels, the motor control necessary to keep it level, and so on. In this way our capacity for prospective awareness is progressively built up and refined, as we learn to avail ourselves more effectively and efficiently of meanings and values in our everyday environment.

In contrast to Turner, who views schemas as neat, clearly delineated concepts that we impose upon the world, the ecological theory of cognition regards them as fuzzy and open-ended heuristic devices that each individual develops, starting even before birth, by adjusting its perceptions and actions to yield productive encounters with the world. This general capacity we share with all living things. The importance of Turner's work lies in his account of how the human mind goes beyond such "literal "cognition. By analyzing how we interpret metaphors in language, and how we use such metaphors to gain novel insights into the world, Turner shows that a great deal of human thought involves what he calls the "projection" and "blending" of lower-order schemas, usually based in "small spatial and bodily stories" derived from our direct encounters with the world. The crucial insight is that this is not just a linguistic capacity: it is a more fundamental cognitive capacity that underlies our use of language. Turner analyzes some basic kinds of schematic projections in English, such as "events are actions" (where we use our experience of human agency to understand agent-less happenings), "actors are movers or manipulators" (where complex forms of agency are analyzed in terms of familiar physical actions), "events are spatial stories" (where change over time is compared to change in physical space), and so on. He also shows how such familiar imaginative devices as talking animals rely on our capacity to draw selectively on our schematic knowledge of the world to produce novel blends, and that such blending can account for much or all of our logical and problem-solving ability as well as the centrality of story-telling to all human societies (Turner, 1996).

Turner calls this capacity "parable". Within the framework of ecological psychology, however, it can be seen to constitute a particular kind of prospective awareness, one that humans possess to an exceptional degree. Since I suspect that schematic projection and schematic blending underlie all forms of human creativity, I will term this the capacity for imaginative awareness.


Why should imaginative awareness, on the evidence of problem-solving ability, be so little-developed among other large-brained species? I suspect that it is fundamentally a risky gambit. Literal cognition encourages the organism to remain within known confines, while imaginative awareness projects the familiar on the unknown, with potentially catastrophic results. Literal cognition rests on ecological integration, with individuals relying on the constant feedback from their environment to guide their awareness; imaginative awareness is built on novel schematic projection within the individual. Taken together, these observations suggest that imaginative awareness becomes a viable evolutionary strategy only when a means can be found for sharing such private projections.

Consider the difference between isolated individuals exploring the world by means of imaginative awareness, and a community with a collective public store of imaginative stories. In the first case, even when an individual discovers that a particular projection "fits" reality -- that is, it proves a reliably useful guide for efforts towards meaning and value -- the discovery will be of no lasting significance unless it can be passed on to the next generation. Moreover, many, many projections are likely to prove a very poor fit to reality, thereby decreasing the adaptive fit of the most imaginative individuals. It is difficult to see how the capacity for parable could ever gain a hold in such conditions. If, however, individuals are able to communicate with one another about their imaginative discoveries, then information on both fit and lack of fit can be passed on, and the collective store of useful projections will grow over time.

Language, I suggest, does exactly this -- although not unproblematically, as we shall see. But language as we know it today may well have been piggybacked on a prior capacity for mimetic communication, as argued by Marlin Donald (1991). With the first stirrings of imaginative awareness in a species that was already specialized in collective efforts towards meaning and value, instinctual and improvised sign systems could begin to evolve towards conventional ones. Such systems are truly emergent: they can only be understood in terms of the interaction of the individual with the communicative environment embodied in other individuals, and are thus reducible neither to individual psychology nor to a system independent of its users. In convention-dependent communication, individuals encounter a system inherent in the ongoing co-ordering of communicative acts around them, strive to acquire the capacity to use the system for their own purposes, and thereby contribute both to the system's reproduction and to changing it in infinitesimal or sometimes more dramatic ways. Language as we know it today is the product of perhaps two million years of evolution of such conventional sign systems, which in turn, in a positive feedback loop, have constituted an environment selecting for the development of imaginative awareness, fine motor control, and specifically linguistic vocal and auditory skills, among other uniquely human traits (Donald, 1991).


Let me now highlight the philosophical distinctions between this ecological view of language and the neo-Cartesian view that has infiltrated the mainstream of applied linguistics. As an emergent, evolving system of co-ordered public acts, language is preexistent and external to any individual agent. This means that individuals never really "acquire a language" in the sense of being able to reproduce the whole system in all its dynamic complexity. Individuals learn to reproduce elements of the system, comprising myriad natural groupings and sub-systems of public language devices, for co-ordering their own awareness and actions and those of others in ways that are useful to them. It is true that an individual's stock of capacities for interpreting and producing language devices constitutes a system of its own, but the ontology and epistemology of this private system differ strikingly from the emergent phenomenon of public language.

In learning to use the public language system, individuals integrate their schematic knowledge of "doing language" with the rest of their dynamic schematic knowledge of the world. Since no one's linguistic or bodily experience is identical, each individual has their own unique inner system of intensions (consisting of schemas and schematic networks ) to use in the interpretation and production of language devices. As Turner, Millikan and many others have recognized, any language device typically shares many intensions: this is what lies behind Turner's observation that "blending is already involved in our most unitary and literal... conception of basic physical objects, such as horse and horn, and in our most unitary and literal... conception of small spatial stories, such as horse moves and horn impales" (Turner 1996: 112). Roughly speaking, a person's intensions for "horse" comprise all those schemas relating to past encounters with horses, and with the public language token "horse," that are available for projection in the processes of imaginative awareness that make language possible. Each schema contributes to a differing extent to the schematic network that constitutes that person's concept of "horse"; when "horse" is linked to another concept such as "moves," the two networks combine in a complex fashion, reducing the contribution of some schemas (such as those related to smell or touch) and increasing the contribution of others related to horses moving. Concepts thus occupy a middle ground between experience and language, or perception and description. Although we often speak as though concepts were public devices, they are not: they are the private schematic networks underlying individuals' use of public language devices.

As proposed by Ruth Millikan (1984), all knowledge of language begins from the observation that a given language device has a meaning, i.e. that it functions, within real utterances, either to get hearers to act in certain ways purposed by speakers, or to clarify speakers' thoughts and intentions in ways that hearers find acceptable. This is enough to start the development of a small schematic network for recognizing the device, linked to a more complex schematic network for interpreting it -- that is, for using it to guide one's imaginative awareness and direct actions; and, eventually, to another small schematic network for producing it. Of course, all of these networks are not discrete but interlinked with other networks derived from all our other encounters with the world. The notions of meaning as discrete and unitary, of linguistic abilities as modular and structurally determined, of knowledge and thought as logical and propositional, are all based on the assumption that private language capacities and public language devices share the same ontology. Critical realism and ecological psychology imply that they are completely different phenomena. Most of the properties of language as a system derive from the differential reproduction of language devices in a particular community of users, with our biological makeup simply imposing a number of constraints on the kinds of sound and the length and complexity of the structures we produce and recognize. The technologies of writing and printing have in turn relaxed some of these constraints and imposed others, implying that general linguistics, and work in applied linguistics that refers to general linguistic theory, need to be much more careful than they have traditionally been in specifying the cultural and technological context of communication.


Some of the "natural groupings and sub-systems" of language devices that I referred to above are quite familiar: the groupings of different sounds to constitute a phoneme, the groupings of phonemes to constitute a word, and so on. As Millikan points out, the basic mechanism behind such grouping is the need to recognize a particular language device as "the-same-device-again", and involves no other cognitive skill than the capacity for identification that is basic to awareness of meaning. Recently, Nick Ellis has adopted the term "chunking" to describe this learning mechanism, and suggested that it can account for a good deal of the order observed in both L1 and L2 acquisition (Ellis, 1997). The ecological account of cognition seems to be compatible with Ellis's proposal, according to which "collocations and chunks of language... are slowly analyzed on a word-by-word basis to allow the determination of grammatical word class and c-selection" (Ellis 1997: 53) -- that is, a growing understanding of holophrases and contextualized words entails the development of a rich complex of schematic associations and projections which necessarily, and without the intervention of a specialized language module, generates tacit "knowledge" of semantic and syntactic categories. However, to focus on words and phrases alone is to neglect other important forms of order in the ecology of communication.

Linguistic tokens are not interpreted automatically, simply by virtue of their availability in the environment: they must be perceived as relevant to the current focus of attention. Linguistic communication therefore relies on the achievement of shared awareness among the participants. In the framework of ecological psychology, learning when and how to achieve shared awareness with another person is a vital skill that children begin to acquire soon after birth. By the age of nine to twelve months, children are able to enter into "triadic interaction frames" that include a caregiver and an object or event; to do so, they must master "what appear to be highly complex cross-modal perception and action skills" (Reed 1996: 129). As individuals grow up, they learn to enter a broader and broader range of interaction frames, including those in which the interlocutor is not physically present (telephone conversations, correspondence, and many different forms of print communication including labeling, forms, various literary genres and scientific prose). All of these frames are themselves conventional communicative systems in which a variety of signs (mimetic, linguistic and so on) can co-evolve to guide awareness and action in productive ways; any regular and lasting relationship between two or more people provides a context in which new and specific interaction frames can develop out of old generic ones.

The linguistic consequence of this social structuring of communication is the Bakhtinian speech genre (Bakhtin, 1986 [1953]) #43]. I cannot find a better way of describing this phenomenon than the words of Bakhtin himself:

We know our native language -- its lexical composition and grammatical structure -- not from dictionaries and grammars but from concrete utterances that we hear and that we ourselves reproduce in live speech communication with people around us. We assimilate forms of language only in forms of utterances and in conjunction with these forms. The forms of language and the typical forms of utterances, that is, speech genres, enter our experience and our consciousness together, and in close connection with one another. To learn to speak means to learn to construct utterances (because we speak in utterances and not in individual sentences, and, of course, not in individual words). Speech genres organize our speech in almost the same way as grammatical (syntactical) forms do. We learn to cast our speech in generic forms and, when hearing others' speech, we guess its genre from the very first words; we predict a certain length (that is, the approximate length of the speech whole) and a certain compositional structure; we foresee the end; that is, from the very beginning we have a sense of the speech whole, which is only later differentiated during the speech process. If speech genres did not exist and we had not mastered them, if we had to originate them during the speech process and construct each utterance at will for the first time, speech communication would be almost impossible (1986: 78-79).

Learning language entails learning genres, and this is as true for L2 as for L1. One of the major difficulties of L2 acquisition is that classroom genres of interaction simply do not resemble L1 genres, even when an emphasis is placed on communication -- for at the heart of genre is relationship, which binds individuals to one another and to the human community around them. As Bakhtin implies, relationships cannot be "constructed at will" -- they grow and evolve in a particular complex environment. Acquisition research thus needs to take account of the relationships that are being constructed as learning proceeds, and explore how the attendant genres constrain language interpretation and production (as in Tarone and Liu's account of "systematic interlanguage variation" -- Tarone & Liu, 1995).


In a triadic interaction frame, communication is structured not only by the relationship of the communicators, but also, semi-independently, by the object of their joint attention, i.e. the topic of their communication. The relationship and its attendant speech genres display evolutionary continuity over time: that is, each encounter builds on and modifies the frame developed in previous interactions. Exactly the same is true of the language tokens used to co-order the imaginative awareness of the interlocutors with regard to a particular topic. As I have already stressed, language use is simultaneously language reproduction, as speakers assemble language devices acquired by exposure to others' speech in order to match, as best they can, the particular state of awareness they wish to convey. Two sets of schematic networks are thus involved: one involving the speaker/writer's own schematic knowledge of the topic, including the language devices used by others to refer to it (possibly across a variety of genres); the other involving an assessment of how particular language devices will be interpreted by the hearer or reader within the particular genre at hand. Once again, Bakhtin provides a wonderfully clear description:

The consequence is that language learners not only inherit, reproduce and The topic of the speaker's speech, regardless of what this topic may be, does not become the object of speech for the first time in any given utterance; a given speaker is not the first to speak about it. The object, as it were, has already been articulated, disputed, elucidated, and evaluated in various ways. Various viewpoints, world views, and trends cross, converge, and diverge in it. ... The utterance is addressed not only to its own object, but also to others' speech about it....

But the utterance is related not only to preceding, but also to subsequent links in the chain of speech communion. When a speaker is creating an utterance, of course, these links do not exist. But from the very beginning, the utterance is constructed while taking into account possible responsive reactions, for whose sake, in essence, it is actually created (1986: 93-94).

The consequence is that language learners not only inherit, reproduce and modify interaction frames and speech genres, but imagination frames and discourses. In order to use the existing language system, they must learn to co-order (approximately) their own imaginative awareness with the language users around them, a process entailing development of the cognitive capacity that Carspecken calls "position-taking" (1996). In reproducing language devices that they have already encountered in action -- reproduction that can serve multiple pragmatic purposes and thus contribute to the individual's immediate effort towards value -- they simultaneously ensure the evolutionary continuity of particular ways of imagining and talking about the world. All language use is thus integrated within far-reaching systems of co-ordered awareness and action that co-structure the real-world environment of individuals and thereby constrain the meanings and values to which they have access. Research on language acquisition implies understanding the forms in which learners encounter such systems, and their active response in developing specific imaginative and position-taking skills expressed in discourse.


We have seen that communication is naturally structured in evolutionarily continuous genres, derived from relationships between people, and discourses, derived from relationships between people and the world about which they communicate. Now consider the specificity of these relationships. For the most part, we communicate with people who live close to us, who share some of our material conditions of life, who are aware of some of the same ecological meanings and who can contribute to our achievement of similar ecological values. But we do not communicate with any one person about everything: with this one we share certain genres and discourses, with another we share others. Such intersecting and overlapping of genres and discourses leads to the diffusion of language devices throughout the particular group of individuals constituting a community. Tönnies' classic description of Gemeinschaft (Tönnies, 1957) can be reinterpreted in these terms: that is, his three dimensions of kinship, place and mind can be seen to arise from the ongoing negotiation of genre and discourse among any group of people engaged in long-term value-oriented communication. The complete or partial closure of such a community establishes a linguistic system in which language devices are constantly being reproduced and recombined, in a situation analogous to the reproduction and recombination of the genetic material within a biological community; in both cases, differential reproduction drives the evolution of the system as an adaptive whole, even though many individual features may be less than ideally suited to their role. The comparison between linguistic and biological evolution was popular in the nineteenth century, of course, but could not be sustained within the meliorist and gradualist accounts of evolution that predominated until recently. Stephen Jay Gould's depiction of evolution as chancy, opportunistic, punctuated and non-directional (in his essays in Natural History magazine) provides a much more plausible frame for understanding the fantastic variety and intricacy of the world's languages ; indeed, Gould's theory of punctuated equilibria has recently been applied to understanding the general mechanism of language diversification (Dixon, 1997).

A theory of community provides one approach to analyzing the ecological context of language acquisition, particularly in conditions where literacy and formal schooling play little role. The fact is, however, that most of the contexts studied in L2 acquisition research are strongly influenced by the ideal form of social organization that Tönnies termed "Gesellschaft" (Tönnies, 1957), and that Bauman has retrospectively analyzed as "legislative reason" (Bauman, 1987; 1992). Modern linguistics and applied linguistics are largely the products of legislative reason, one of whose hallmarks is a preference for conceptual coherence over organic coherence: treating dictionaries and grammars, for instance, as more definitive of language than people's actual observable behaviour, and thus focusing on langue rather than parole or langage, the "ideal native speaker" rather than real people in real contexts, theories of competence and acquisition rather than theories of agency and use, and so on. This conception of language has real effects on what people do and how they learn; by focusing awareness on the forms of words and phrases rather than on the ecological context and purpose of communication, it brings new genres and discourses into being. First established in the most literate and highly schooled classes, the language behaviours characteristic of "taught mother tongue" now dominate media and schooling in modern industrial societies (Illich, 1981) and form an inescapable part of the context of language acquisition. Acquisition research should surely include the consideration of how linguistics and education themselves have affected its object of study, and extend this to exploring the social implications of acquisition theory; Illich's trenchant critique would provide an excellent starting point.


It will now be clear that the ecological theory I have presented implies a critical approach to language acquisition, and indeed to any kind of social science. There are strong epistemological grounds for believing that all naturalist science must be critical science, and that the supposed division between the natural and human sciences is itself a product of the era of legislative reason. Science works through language devices that allow different individuals to co-order their direct and imaginative awareness of the world; but these language devices (terms, descriptions, theories, discourses) are not the world itself, and thus each individual's understanding and use of them is a situated, fallible activity that carries no guarantees. Language devices work by accomplishing something in the world, but what they accomplish is not knowable in a priori fashion (Millikan, 1984), and may very well vary for different individuals and different contexts. So the investigation of this situatedness, of the ecological conditions for the interpretation and reproduction of language devices, becomes the ultimate test of "truth". Theories are said to be true as long as they productively guide the encounters of scientists with the world; but this, of course, begs the question of how "productively" should be defined, and ignores the uses that be made of those theories in contexts remote from scientific practice. There is no escaping the issue of value in any corner of science, which is why the critical stance is fundamental.

My own preference is for an emancipatory science, a linguistics that helps people understand the constraints upon and consequences of their actions and thereby assists them in choosing and acting wisely. My vision is one of a linguistics in active communication with its neighbours in the biological, social and human sciences, sharing and developing a holistic understanding of human thought, action, and ecological integration. Critical realism and ecological psychology together offer a promising foundation for this enterprise. Upon this base I have sought to build a robust, albeit programmatic, theoretical framework which offers many opportunities for criticism, testing, revision and elaboration. It may prove of particular value in linking work on the ecology of language acquisition to the development of systems approaches in other fields of the human and biological sciences (Capra, 1996), including the fields of indigenous education, language policy and planning, and interlinguistics which have provided the direct inspiration for my own work (Fettes, 1997a; 1997b; 1998; Fettes & Bolduc, 1998).


Bakhtin, M. (1981). The Dialogic Imagination: Four Essays (Caryl Emerson, Michael Holquist, Trans.). Austin: University of Texas Press.

Bakhtin, M. (1981 [1935]). Discourse in the novel. In M. Holquist (Ed.), The Dialogic Imagination: Four Essays by M.M. Bakhtin (pp. 259-422). Austin: University of Texas Press.

Bakhtin, M. (1986 [1953]). The problem of speech genres. In C. Emerson & M. Holquist (Eds.), Speech Genres and Other Late Essays (M.M. Bakhtin) (pp. 60-102). Austin: University of Texas Press.

Bateson, G. (1972). Steps to an Ecology of Mind. New York: Ballantine.

Bateson, G. (1979). Mind and Nature: A Necessary Unity. New York: E.P. Dutton.

Bauman, Z. (1987). Legislators and Interpreters. Ithaca: Cornell University Press.

Bauman, Z. (1992). Intimations of Postmodernity. London and New York: Routledge.

Ben-Ze'ev, A. (1995). Is there a problem explaining cognitive progress? In R. F. Goodman & W. R. Fisher (Eds.), Rethinking Knowledge: Reflections Across the Disciplines (pp. 41-56). Albany: State University of New York Press.

Bhaskar, R. (1989). Reclaiming Reality: A Critical Introduction to Contemporary Philosophy. London: Verso.

Bourdieu, P. (1990). Reproduction in Education, Society and Culture (2nd ed.). London: Sage.

Bourdieu, P. (1991). Language and Symbolic Power. Cambridge: Harvard University Press.

Capra, F. (1996). The Web of Life: A New Scientific Understanding of Living Systems. New York: Anchor/Doubleday.

Carspecken, P. F. (1996). Critical Ethnography in Educational Research: A Theoretical and Practical Guide. New York and London: Routledge.

Corson, D. (1997). Critical realism: An emancipatory philosophy for applied linguistics? Applied Linguistics, 18(2), 166-188.

Dixon, R. M. W. (1997). The Rise and Fall of Languages. Cambridge/New York: Cambridge University Press.

Donald, M. (1991). Origins of the Modern Mind. Cambridge: Harvard University Press.

Ellis, N. (1997). The epigenesis of language: Acquisition as a sequence learning phenomenon. In A. Ryan & A. Wray (Eds.), Evolving Models of Language (pp. 41-57). Clevedon: British Association for Applied Linguistics/Multilingual Matters.

Fettes, M. (1997a). Esperanto and language awareness. In L. van Lier & D. Corson (Eds.), Knowledge about Language. The Encyclopedia of Language and Education, Vol. 6 (pp. 151-159). Dordrecht and Boston: Kluwer.

Fettes, M. (1997b). Language planning and education. In R. Wodak & D. Corson (Eds.), Language Policy and Political Issues in Education. The Encyclopedia of Language and Education, Vol. 1 (pp. 13-22). Dordrecht and Boston: Kluwer.

Fettes, M. (1998). Indigenous education and the ecology of community. Language, Culture, and Curriculum, 11 (3), in press.

Fettes, M., & Bolduc, S. (Eds.). (1998). Al lingva demokratio/Towards Linguistic Democracy/Vers la démocratie linguistique. Rotterdam: Universala Esperanto-Asocio.

Foucault, M. (1980). Power/Knowledge: Selected Interviews and Other Writings 1972-1977. New York: Pantheon.

Freire, P. (1972). Pedagogy of the Oppressed. Harmondsworth: Penguin.

Friedrich, P. (1985). The Language Parallax. Linguistic Relativism and Poetic Indeterminacy. Austin: University of Texas Press.

Harris, R. (1980). The Language-Makers. Ithaca: Cornell University Press.

Harris, R. (1981). The Language Myth. London: Duckworth.

Illich, I. (1981). Taught mother language and vernacular tongue. In D. P. Pattanayak (Ed.), Multilingualism and Mother-Tongue Education (pp. 1-39). Delhi: Oxford University Press.

Maturana, H. R., & Varela, F. J. (1992). The Tree of Knowledge: The Biological Roots of Human Understanding (Robert Paolucci, Trans.). (Revised ed.). Boston and London: Shambhala.

Millikan, R. G. (1984). Language, Thought, and Other Biological Categories: New Foundations for Realism. Cambridge, MA: MIT Press.

Pinker, S. (1997). How the Mind Works. New York: W.W. Norton.

Reed, E. S. (1996). Encountering the World: Toward an Ecological Psychology. Oxford: Oxford University Press.

Smith, D. E. (1990). The Conceptual Practices of Power: A Feminist Sociology of Knowledge. Toronto: University of Toronto Press.

Tarone, E., & Liu, G.-Q. (1995). Situational context, variation, and second language acquisition theory. In G. Cook & B. Seidlhofer (Eds.), Principle and Practice in Applied Linguistics: Studies in Honour of H.G. Widdowson (pp. 107-124). Oxford: Oxford University Press.

Tönnies, F. (1957). Community and Society (Gemeinschaft und Gesellschaft, 1887) (Charles P. Loomis, Trans.). East Lansing: Michigan State University Press. Turner, M. (1996). The Literary Mind. Oxford: Oxford University Press.

Related Links


Copyright © 1999 Mark Fettes

The Web Site for Critical Realism The WSCR Archive