Language Learning and Teaching

     We cannot say, of course, that meaningfully learned material is never forgotten. But in the case of such learning, forgetting takes place in a much more intentional and purposeful manner because it is a continuation of the very process of subsumption by which one learns; forgetting is really a second or "obliterative" stage of subsumption, characterized as "memorial reduction to the least common denominator" (Ausubel 1963:218). Because it is more economical and less burdensome to retain a single inclusive concept than to remember a large number of more specific items, the importance of a specific item tends to be incorporated into the generalized meaning of the larger item. In this obliterative stage of subsumption, the specific items become progressively less identifiable as entities in their own right until they are finally no longer available and are said to be forgotten (see Figure 4.2).

     It is this second stage of subsumption that operates through what I have called "cognitive pruning" procedures (Brown 1972). Pruning is the elimination of unnecessary clutter and a clearing of the way for more material to enter the cognitive field, in the same way that pruning a tree ultimately allows greater and fuller growth. Using the building-block analogy, one might say that, at the outset, a structure made of blocks is seen as a few individual blocks, but as "nucleation" begins to give the structure a perceived shape, some of the single blocks achieve less and less identity in their own right and become subsumed into the larger structure. Finally, the single blocks are lost to perception, or pruned out, to use the metaphor, and the total structure is perceived as a single whole without clearly defined parts.

   An example of such pruning may be found in a child's learning of the concept of "hot"—that is, excessive heat capable of burning. A small child's first exposure to such heat may be either direct contact with or verbally mediated exposure to hot coffee, a pan of boiling water, a stove, an iron, a candle. That first exposure may be readily recalled for some time as the child maintains a meaningful association between a parent's hot coffee and hurting. After a number of exposures to things that are very hot, the child begins to form a concept of "hotness" by clustering experiences together and forming a generalization. In so doing the bits and pieces of experience that actually built the concept are slowly forgotten—pruned— in favor of the general concept that, in the years that follow, enables the child to extrapolate to future experiences and to avoid burning fingers on hot objects.

   An important aspect of the pruning stage of learning is that subsumptive forgetting, or pruning, is not haphazard or chance—it is systematic. Thus by promoting optimal pruning procedures, we have a potential learning situation that will produce retention beyond that normally expected under more traditional theories of forgetting.

   Research on language attrition has focused on a variety of possible causes for the loss of second language skills (see Weltens & Cohen 1989; Weltens 1987; Lambert & Freed 1982). Some of the more common reasons center on the strength and conditions of initial learning, on the kind of use that a second language has been put to, and on the motivational factors contributing to forgetting. Robert Gardner (1982) contended that in some contexts a lack of an "integrative" orientation (but see caveats in Chapter 6) toward the target culture could contribute to forgetting.

   Native language forgetting occurs in some cases of subtractive bilingualism (members of a minority group learn the language of the majority group, and the latter group downgrades speakers of the minority language). Some researchers have suggested that "neurolinguistic blocking" and left-/right-brain functioning could contribute to forgetting (Obler 1982). And it appears that long-term forgetting can apply to certain linguistic features I (lexical, phonological, syntactic, and so on) and not to others (Andersen I 1982). Finally, Olshtain (1989) suggested that some aspects of attrition can I be explained as a reversal of the acquisition process.

   Research on language attrition usually focuses on long-term loss and not on those minute-by-minute or day-by-day losses of material that learners experience as they cope with large quantities of new material in the course of a semester or year of classroom language learning. It is this classroom context that poses the more immediate problem for the language teacher. Ausubel's solution to that problem would lie in the initial learning process: systematic, meaningful subsumption of material at the outset in order to enhance the retention process.

   Ausubel's theory of learning has important implications for second language learning and teaching. The importance of meaning in language and of meaningful contexts for linguistic communication has been discussed in the first three chapters. Too much rote activity, at the expense of meaningful communication in language classes, could stifle the learning process.

   Subsumption theory provides a strong theoretical basis for the rejection of conditioning models of practice and repetition in language teaching. In a meaningful process like second language learning, mindless repetition, imitation, and other rote practices in the language classroom have no place. The Audiolingual Method, which emerged as a widely used and accepted method of foreign language teaching, was based almost exclusively on a behavioristic theory of conditioning that relied heavily on rote learning. The mechanical "stamping in" of the language through saturation with little reference to meaning is seriously challenged by subsumption theory. Rote learning can be effective on a short-term basis, but for any long-term retention it fails because of the tremendous buildup of interference. In those cases in which efficient long-term retention is attained in rote-learning situations like those often found in the Audiolingual Method, maybe by sheer dogged determination, the learner has somehow subsumed the material meaningfully in spite of the method!

   The notion that forgetting is systematic also has important implications for language learning and teaching. In the early stages of language learning, certain devices (definitions, paradigms, illustrations, or rules) are often used to facilitate subsumption. These devices can be made initially meaningful by assigning or "manufacturing" meaningfulness. But in the process of making language automatic, the devices serve only as interim entities, meaningful at a low level of subsumption, and then they are systematically pruned out at later stages of language learning. We might thus better achieve the goal of communicative competence by removing unnecessary barriers to automaticity. A definition or a paraphrase, for example, might be initially facilitative, but as its need is minimized by larger and more global conceptualizations, it is pruned.

   While we are all fully aware of the decreasing dependence upon such devices in language learning, Ausubel's theory of learning may help to give explanatory adequacy to the notion. Language teachers might consider urging students to "forget" these interim, mechanical items as they make progress in a language and instead to focus more on the communicative use (comprehension or production) of language.

ROGERS'S HUMANISTIC PSYCHOLOGY

Carl Rogers is not traditionally thought of as a "learning" psychologist, yet he and his colleagues and followers have had a significant impact on our present understanding of learning, particularly learning in an educational or pedagogical context. Rogers's humanistic psychology has more of an affective focus than a cognitive one, and so it may be said to fall into the perspective of a constructivist view of learning. Certainly, Rogers and Vygotsky (1978) share some views in common in their highlighting of the social and interactive nature of learning.

     Rogers devoted most of his professional life to clinical work in an attempt to be of therapeutic help to individuals. In his classic work Client-Centered Therapy (1951), Rogers carefully analyzed human behavior in general, including the learning process, by means of the presentation of nineteen formal principles of human behavior. All nineteen principles were concerned with learning from a "phenomenological" perspective, a perspective that is in sharp contrast to that of Skinner. Rogers studied the "whole person" as a physical and cognitive, but primarily emotional, being. His formal principles focused on the development of an individual's self-concept and of his or her personal sense of reality, those internal forces that cause a person to act. Rogers felt that inherent in principles of behavior is the ability of human beings to adapt and to grow in the direction that enhances their existence. Given a nonthreatening environment, a person will form a picture of reality that is indeed congruent with reality and will grow and learn. "Fully functioning persons," according to Rogers, live at peace with all of their feelings and reactions; they are able to reach their full potential.

     Rogers's position has important implications for education (see Curran 1972; Rogers 1983). The focus is away from "teaching" and toward "learning." The goal of education is the facilitation of change and learning. Learning how to learn is more important than being taught something from the "superior" vantage point of a teacher who unilaterally decides what shall be taught. Many of our present systems of education, in prescribing curricular goals and dictating what shall be learned, deny persons both freedom and dignity. What is needed, according to Rogers, is for teachers to become facilitators of learning through the establishment of interpersonal relationships with learners. Teachers, to be facilitators, must first be real and genuine, discarding masks of superiority and omniscience. Second, teachers need to have genuine trust, acceptance, and a prizing of the other person—the student—as a worthy, valuable individual. And third, teachers need to communicate openly and empathetically with their students and vice versa. Teachers with these characteristics will not only understand themselves better but will also be effective teachers, who, having set the optimal stage and context for learning, will succeed in the goals of education.

     We can see in Carl Rogers’s humanism quite a departure from the scientific analysis of Skinnerian psychology and even from Ausubel's rationalistic theory. Rogers is not as concerned about the actual cognitive process of learning because, he feels, if the context for learning is properly created, then human beings will, in fact, learn everything they need to.

     Rogers's theory is not without its flaws. The educator may be tempted to take the nondirective approach too far, to the point that valuable time is lost in the process of allowing students to "discover" facts and principles for themselves. Also, a nonthreatening environment might become so non-threatening that the facilitative tension needed for learning is absent. There is ample research documenting the positive effects of competitiveness in a classroom, as long as that competitiveness does not damage self-esteem and hinder motivation to learn (see Bailey 1983).

     One much talked-about educational theorist in the Rogersian tradition is the well-known Brazilian educator Paolo Freire, whose seminal work, Pedagogy of the Oppressed (1970), has inspired many a teacher to consider the importance of the empowerment of students in classrooms. Freire vigorously objected to traditional "banking" concepts of education in which teachers think of their task as one of "filling" students "by making deposits of information which [they] consider to constitute true knowledge-deposits which are detached from reality" (1970: 62). Instead, Freire has continued to argue, students should be allowed to negotiate learning outcomes, to cooperate with teachers and other learners in a process of discovery, to engage in critical thinking, and to relate everything they do in school to their reality outside the classroom. While such "liberationist" views of education must be approached with some caution (Clarke 1990), learners may nevertheless be empowered to achieve solutions to real problems in the real world.

     The work of Rogers (1983), Freire (1970), and other educators of a similar frame of mind has contributed significantly in recent years to a redefinition of the educational process. In adapting Rogers's ideas to language teaching and learning, we need to see to it that learners understand themselves and communicate this self to others freely and nondefensively. Teachers as facilitators must therefore provide the nurturing context for learners to construct their meanings in interaction with others. When teachers rather programmatically feed students quantities of knowledge, which they subsequently devour, they may foster a climate of defensive learning in which learners try to protect themselves from failure, from criticism, from competition with fellow students, and possibly from punishment. Classroom activities and materials in language learning should therefore utilize meaningful contexts of genuine communication with students engaged together in the process of becoming "persons."

     The various perspectives on learning that have been outlined in this section are schematically represented in Table 4.1. 

Table 4.1   Theories of learning 

TYPES OF LEARNING

Theories of learning of course do not capture all of the possible elements of general principles of human learning. In addition to the four learning theories just considered are various taxonomies of types of human learning and other mental processes universal to all. The educational psychologist Robert Gagne (1965), for example, ably demonstrated the importance of identifying a number of types of learning that all human beings use. Types of learning vary according to the context and subject matter to be learned, but a complex task such as language learning involves every one of Gagne's types of learning—from simple signal learning to problem solving. Gagne (1965: 58-59) identified eight types of learning: 

1. Signal learning. The individual learns to make a general diffuse response to a signal. This is the classical conditioned response of Pavlov.
2. Stimulus-response learning. The learner acquires a precise response to a discriminated stimulus. What is learned is a connection or, in Skinnerian terms, a discriminated operant, sometimes called an instrumental response.
3. Chaining. What is acquired is a chain of two or more stimulus-response connections.The conditions for such learning have also been described by Skinner.
4. Verbal association. Verbal association is the learning of chains that are verbal. Basically, the conditions resemble those for other (motor) chains. However, the presence of language in the human being makes this a special type of chaining because internal links may be selected from the individual's previously learned reper toire of language.
5. Multiple discrimination. The individual learns to make a number of different identifying responses to many different stimuli, which may resemble each other in physical appearance to a greater or lesser degree. Although the learning of each stimulus-response connection is a simple occurrence, the connections tend to inter fere with one another.
6. Concept learning. The learner acquires the ability to make a common response to a class of stimuli even though the individual members of that class may differ widely from each other. The learner is able to make a response that identifies an entire class of objects or events.
7. Principle learning. In simplest terms, a principle is a chain of two or more concepts. It functions to organize behavior and experience. In Ausubel's terminology, a principle is a "subsumer"—a cluster of related concepts.
8. Problem solving. Problem solving is a kind of learning that requires the internal events usually referred to as "thinking. "Previously acquired concepts and principles are combined in a conscious focus on an unresolved or ambiguous set of events.

  It is apparent from just a cursory definition of these eight types of learning that some types are better explained by certain theories than others. For example, the first five types seem to fit easily into a behavioristic framework, while the last three are better explained by Ausubel's or Rogers's theories of learning. Since all eight types of learning are relevant to second language learning, the implication is that certain "lower"-level aspects of second language learning may be more adequately treated by behavioristic approaches and methods, while certain "higher"-order types of learning are more effectively taught by methods derived from a cognitive approach to learning.

   The second language learning process can be further efficiently categorized and sequenced in cognitive terms by means of the eight types of learning. 

1. Signal learning in general occurs in the total language process: human beings make a general response of some kind (emotional, cognitive, verbal, or nonverbal) to language.
2. Stimulus-response learning is evident in the acquisition of the sound system of a foreign language in which, through a process of conditioning and trial and error, the learner makes closer and closer approximations to nativelike pronunciation. Simple lexical items are, in one sense, acquired by stimulus-response connections; in another sense they are related to higher-order types of learning.
3. Chaining is evident in the acquisition of phonological sequences and syntactic patterns—the stringing together of several responses—although we should not be misled into believing that verbal chains are necessarily linear. Generative linguists (like McNeill, as we saw in Chapter 2) have wisely shown that sentence structure is hierarchical.
4. The fourth type of learning involves Gagne's distinction between verbal and nonverbal chains, and is not really therefore a separate type of language learning.
5. Multiple discriminations are necessary particularly in second language learning where, for example, a word has to take on several meanings, or a rule in the native language is reshaped to fit a second language context.
6. Concept learning includes the notion that language and cognition are inextricably interrelated, also that rules themselves - rules of syntax, rules of conversation—are linguistic concepts that have to be acquired.
7. Principle learning is the extension of concept learning to the formation of a linguistic system, in which rules are not isolated in rote memory, but conjoined and subsumed in a total system.
8. Finally, problem solving is clearly evident in second language learning as the learner is continually faced with sets of events that are truly problems to be solved—problems every bit as difficult as algebra problems or other "intellectual" problems. Solutions to the problems involve the creative interaction of all eight types of learning as the learner sifts and weighs previous information and knowledge in order to correctly deter-mine the meaning of a word, the interpretation of an utterance, the rule that governs a common class of linguistic items, or a conversationally appropriate response.

  It is not difficult, upon some reflection, to discern the importance of varied types of learning in the second language acquisition process (see Larsen-Freeman 1991). Teachers and researchers have all too often dismissed certain theories of learning as irrelevant or useless because of the misperception that language learning consists of only one type of learning. "Language is concept learning," say some; "Language is a conditioning process," say others. Both are correct in that part of language learning consists of each of the above. But both are incorrect to assume that all of language learning can be so simply classified. Methods of teaching, in recognizing different levels of learning, need to be consonant with whichever aspect of language is being taught at a particular time while also recognizing the interrelatedness of all levels of language learning.

TRANSFER, INTERFERENCE, AND OVERGENERALIZATION