Source: Visible Language, vol. 30, pp. 284-313 (1996).

The Apprenticeship Approach to Writing Instruction

Abstract: This essay begins by reviewing the nature of apprenticeship in non-writing contexts. It then describes, distinguishes, and illustrates the apprenticeship approach to writing instruction. After surveying evidence that apprenticeship provides a more promising model of writing instruction than any other contemporary approach, this essay highlights a few practical applications of this model to writing instruction. This essay concludes that apprenticeship comes closer than other contemporary models to providing an overarching paradigm of writing instruction. The apprenticeship model is consistent with much of what we know about both language and learning; it promises to make writing instruction more enjoyable and fruitful to both learners and teachers; it resolves such perennial controversies as the place of literature, explicit teaching, grammar, and self-awareness in the composition classroom; and it assimilates the best features of traditional and process instruction while avoiding most of their pitfalls.

"Now don't you see the difference? It wasn't anything but a wind reef. The wind does that."

"So I see. But it is exactly like a bluff reef. How am I ever going to tell them apart?"

This essay argues that the composition classroom should resemble a pottery workshop. The practitioner of the apprenticeship approach to writing instruction motivates her students and sets for them appropriate and engaging tasks. She fosters naturalness and places writing in a meaningful and supportive social system. She attempts to build up her students' self-confidence in themselves as human beings and as writers. She knows that practice makes perfect, and so her students learn by writing and editing. If she can, she sometimes converts her classroom into a computer-assisted writing workshop. Her students learn by editing their text on their own, and then through feedback from her and from fellow students. She teaches her students to revise texts by relying on common sense and intuition, not an inaccessible and confusing multitude of explicit rules and strategies. She guides them towards improvements of their texts in the same way that a master-potter guides his apprentices. She employs conceptual change and guided discovery strategies. Her students learn by watching accomplished writers performing their crafts, and by reading well-written literature and prose. While her students are actually writing, she does not inundate them with corrections, abstractions, and strategies, but fits her advice to the circumstances, to what she feels her students can assimilate at that particular point. She uses explicit instruction sparingly, appropriately, and in context. She tries to avoid out-of-context excessive teaching of rules and strategies. She urges her students to focus their attention on the task at hand, not on themselves, the writing process, or their audience.

This essay urges composition instructors to assign a lesser role to explication and self-consciousness, and a greater role to strategies which are consistent with, or follow from, the apprenticeship approach. In doing so, this essay largely draws upon confirmatory theoretical and experimental evidence. A more detailed discussion of seemingly contradictory theories, empirical findings, and practices (e.g., Collins, Brown, & Newman, 1987; Flower, 1989; Rosenshine & Meister, 1994; Williams & Colomb, 1993), would require a book-length manuscript and cannot be undertaken here.

Composition instructors "do not have a unified theory to guide [their] work" (Lindemann 1993, p. 316). As a result, there is little agreement in the professional literature, and in actual practice, about fundamental questions: Should we, for example, teach grammar (Hartwell, 1985)? Has literature a place in the composition classroom (Lindemann, 1993; Tate, 1993)? In modeling essays and fiction, should we try to explicitly distill the rhetorical devices which make writing powerful (Axelrod & Cooper, 1994; Hacker, 1994), or should we focus on the content of those pieces and let students tacitly imbibe the stylistic features of good writing (Smith, 1988)? Should we encourage writers to spend much time reflecting on the process of writing, or should we keep their attention riveted on the emerging text (Applebee, 1986)? Should we teach good writing in general, or should we teach students to write like lawyers, doctors, or advertisers (Freedman, 1993)?

Many such controversies could be resolved by answering the question: Which instructional paradigm can best guide writing instruction? Before attempting to cast apprenticeship in that role, we need to be clear about the nature of apprenticeship in non-writing contexts.

Apprenticeship¹

Many skills are only partially amenable to explicit instruction. According to Michael Polanyi (1962):

A skillful performance is achieved by the observance of a set of rules which are not known as such to the person following them. For example...the swimmer keeps himself afloat by...[regulating] his respiration; he keeps his buoyancy at an increased level by refraining from emptying his lungs when breathing out and by inflating them more than usual when breathing in: yet this is not generally known to swimmers.... Again [does]...the rule observed by the cyclist...tell us exactly how to ride a bicycle? No. You obviously cannot adjust the curvature of your bicycle's path in proportion to the ratio of your unbalance over the square of your speed...Rules of art can be useful, but they do not determine the practice of an art...

An art which cannot be specified in detail cannot be transmitted by prescription, since no prescription for it exists. It can be passed on only by example from master to apprentice....It follows that an art which has fallen into disuse for the period of a generation is altogether lost....It is pathetic to watch the endless efforts--equipped with microscopy and chemistry, with mathematics and electronics--to reproduce a single violin of the kind the half-literate Stradivarius turned out as a matter of routine more than 200 years ago. To learn by example is to submit to authority....By watching the master...the apprentice unconsciously picks up the rules of the art, including those which are not explicitly known to the master himself....

In deciding a case today the [British] Courts will follow the example of other courts which have decided similar cases in the past, for in these actions they see embodied the rules of the law. This procedure recognizes the principle of all traditionalism that practical wisdom is more truly embodied in action than expressed in rules of action....The judge's action is considered more authentic than what he said he was doing....

Despite popular conceptions of science as a realm of explicit rules and computer-like precision, many commentators feel that the tacit component is the distinguishing feature of actual scientific investigations. Speaking from first-hand experiences, Polanyi (1962) holds that scientific research is "an unspecifiable art" (p. 53). Philosopher of science Paul Feyerabend (1993) goes farther. In his view, doing creative science in different settings does not involve one unspecifiable art, but many:

The events, procedures and results that constitute the sciences have no common structure; there are no elements that occur in every scientific investigation but are missing elsewhere. Concrete developments...have distinct features and we can often explain why and how these features led to success. But not every discovery can be accounted for in the same manner, and procedures that paid off in the past may create havoc when imposed on the future. Successful research does not obey general standards; it relies now on one trick, now on another; the moves that advance it and the standards that define what counts as an advance are not always known to the movers....A theory of science that devises standards and structural elements for all scientific activities...may impress outsiders--but it is much too crude an instrument for the people on the spot, that is, for scientists facing some concrete research problem....There is only one principle that can be defended under all circumstances and in all stages of human development. It is the principle: anything goes (pp. 1, 18-19).

As we have seen, mastery of an art presupposes naturalness. A good violin teacher tries to make his students as unselfconscious as he can; he avoids any advice that would put them at risk of permanently riveting their attention from the music they are playing to bystanders or to their audience, fingers, bow, or shoulder. His advice is of the kind: "Wouldn't that gavotte sound better if played a bit faster?" Or "This F was a bit too flat, try it again!" It is rarely of the kind: "Play in front of a mirror!" Nor: "Write an essay about your fingering technique!" His students are told to listen to the masters, watch them, and practice alone or in company. They are rarely given take-home assignments like this:

Flower, 1989, pp. 209-210)

Feedback in context is often crucial to the acquisition of an art: masters serve not only as models but also as coaches. The most useful advice is often given explicitly in response to the learner's actual performance. The advice may appeal to such things as the apprentice's aesthetic judgment or instincts. At times, it invokes an explicit rule, especially when this rule can be validated and brought to life by the novice's experiences. This occasional usefulness of explicit rules could be illustrated by citing Mark Twain again, written recollections of other apprentices, or countless literary episodes, but let me instead relate two homelier childhood experiences of my own. By imitating others, I taught myself the crawl stroke, but I couldn't swim as far or as fast as some of my companions. The turning point came two summers later, and only when a knowledgeable observer told me: "blow bubbles when your face is in the water." Similarly, I played harmonica for months, picked many tunes by ear, but these merely made the neighborhood's crows sound like gifted musicians. Again, the eureka moment came shortly after a better player gave me a rule: "harmonica players use their tongues."

No doubt there are vast differences between learning to play harmonica, ride a bicycle, or carry out a creative genetics experiment; nonetheless, such activities do have common, overlapping, features. At this point, we need to tentatively distill a few features of apprenticeship:

1. As in most learning situations, motivation, perceived relevance, feelings, attitudes towards the learning environment, social context, self-confidence, prior learning experiences, aptitudes, and learning style, affect the rate of progress and final mastery of the craft (Freedman, 1993, p. 247; Pintrich et al., 1993).

2. For the most part, decontextualized expositions about performance of a craft are of limited, if any, value. No one can learn to coordinate a car's clutch, gas pedal, and gearshift from lectures. One masters a craft gradually, by observing and imitating others, by comparing one's performance to those of others, by learning from one's mistakes, and by being coached.

3. One learns through performing a craft in a natural setting; a would-be driver must spend time behind the wheel.

4. Much learning can occur passively, by intently watching others perform.

5. Excessive self-consciousness must be avoided at all costs. We can only do one of three things: rivet our attention on the task at hand, on our actions, or on our audience. "If a pianist shifts his attention from the piece he is playing to the observation of what he is doing with his fingers while playing it, he gets confused" (Polanyi, 1962, p. 53). Similarly, some people have an excellent sense of rhythm but are too concerned with their own movements, or with the impression they make on others, to dance well.

6. Explicit coaching--when doled out sparingly, in context, to meet the learner's developmental needs--often plays a crucial role in the acquisition of an craft.

Not everything can be taught through adherence to these seven, tentative, features of apprenticeship. Some skills benefit from, or even require, abstract instruction which is separate in time or place from practical experience. Thus, although many children master tic-tac-toe by playing the game, almost everyone can become a tic-tac-toe grandmaster through explicit instruction and vigilant application of rules.

As another example, consider the knottier case of backgammon. Although backgammon mastery requires a long apprenticeship, certain features of this game can be best acquired through decontextualized instruction and metareflections.

Let us imagine a single game being played for $1. Under certain conditions, a player may propose doubling the stakes. Also, sometimes a player knows the exact odds of winning a game. Let us say that Black is permitted to double, knows that her odds of winning are 72%, and proceeds to double. White can decline the double, forfeit the game and one dollar, and hope for better luck next game. Or White can accept the double, proceed to play for $2, knowing that, by the rules of the game, Black cannot re-double the stakes. Obviously, White is caught between a rock and hard place. Should he decline and cut his losses, or should he keep hope alive by agreeing to double the stakes?

Neither common sense nor experience are reliable guides to such dilemmas. I have observed the behavior of over one hundred people under similar conditions, and found that, as in so many similar cases (see Edwards, 1968; Kahneman et al., 1982), our statistical intuition is a broken reed. A few people do eventually adopt the correct rule (always accepting a double when there is a 72% chance of losing the game) either through a priori reasoning or through watching, or playing with, an expert. Many play a lifetime yet fail to consistently execute the correct move.

In backgammon, then, correct reactions to doubles can be best learned through out-of-context instruction (for example, from a book). So, in this game, apprenticeship is best supplemented with abstract study of a few statistical rules and their concrete application. Moreover, until these rules are assimilated, players must self-consciously monitor their performance.

To sum up. Human beings can learn limited skills like tic-tac-toe in an abstract fashion. Likewise, abstract instruction is essential when people must acquire counterintuitive rules which are not readily refuted by experience. But there is no evidence that humans can learn complex tasks through decontextualized exposure to explicit rules and strategies, or through frequently stepping back and reflecting upon their performance. There is no substitute for apprenticeship, and no royal road to mastering a craft. .

Writing, according to the apprenticeship view, is a craft. A writer need not be conscious of rules embodied in well-written texts, nor of his strategies for generating and editing such texts. A writer should, according to this view, focus his attention on the task at hand, not on his audience or on her way of accomplishing this task. A liberal dose of abstractions and self-consciousness may be useful on the therapist's couch, but it is just as damaging to writers as it is to speakers, pianists, scientists, antebellum Mississippi River pilots, or contemporary quartermasters. To be sure, good writers possess metacognitive skills, but these skills can be best acquired tacitly through apprenticeship, through reading and writing in context, and not through abstract instruction. In more general terms, good writing instructors abide by something like the seven features of apprenticeship mentioned earlier.

Some of these practices are based on the notion that the rules that "you don't know won't help you" (Williams & Colomb, 1993). Likewise, another theorist is puzzled by the irony "that anyone in higher education would bother to question the usefulness of conscious, reflective thought" (Flower, 1989, p. 185). At the core of these approaches there is this highly persuasive argument: Well-written texts embody rules of grammar, style, and syntax. Experts employ well-defined strategies when they produce such texts. Once these rules and strategies have been unearthed by meticulous research, they can be gainfully passed on to novices. And, since such rules and strategies cannot be automatically assimilated, learners must now and then stand back and divert their attention from the writing task to their own writing strategies. By and by, this process becomes second nature and a novice is transformed into an expert.

The following quotes give the flavor of the explicative and self-conscious strategies of writing instruction. College freshmen are told: "Master the essentials of the sentence as an aid to clear thinking and effective writing." "Adjective clauses or phrases are nonrestrictive when they describe (rather than limit the meaning of) the noun or pronoun they modify; set off by commas, they are nonessential parenthetical elements that may be omitted (Hodges, 1990, pp. 2, 134-5; see also Zuber & Reed, 1993, on the importance of handbooks in writing instruction). A typical freshman composition program combines "regular instruction with a parallel set of assignments that both 'invite' and help the student reflect upon his or her own thinking and learning." For instance, students are asked at one point: "How did you handle this process [of defining a problem]? What writing/thinking strategies do you use?" (Flower, 1989, pp. 210, 208). "Methods to encourage reflection might consist of recording students as they think out loud and then replaying the tape for comparison with the thinking of experts and other students (Collins, Brown, & Newman, 1989, p. 483; see also Elbow, 1973, p. 144; Rose, 1984, pp. 88-9).

A widely-used handbook dedicates twenty-seven pages to the comma. This discussion begins by explaining why commas are needed. To satisfy these needs, ten rules have evolved. This handbook then describes, explains, and illustrates these rules. It then clarifies and illustrates important exceptions to these rules, and provides exercises for the application of these newly-acquired rules.

The first of this rules states: "Use a comma before a coordinating conjunction joining independent clauses." This is followed by definitions of "independent clauses" and "coordinating conjunctions." Students are then advised that a comma "tells readers that one independent clause has come to a close and that another is about to begin." This is followed by an exception, stated as a rule and then illustrated. Readers are then cautioned not to use a comma to "separate coordinate word groups that are not independent clauses." This last statement is then concretely exemplified.

After similar treatises on the remaining nine rules, readers are told that "many common misuses of the comma result from an incomplete understanding" of these self-same ten rules. "Writers often form misconceptions about" these rules, either extending them inappropriately or misinterpreting them. Such misconceptions can lead to errors. There are also other misuses of the comma. These errors and misuses require, in their turn, an extended discussion of fourteen points. The first in that new series is: "Do not use a comma between compound elements that are not independent clauses," a point which is then expounded and illustrated. Following a similar discussion of the next thirteen points, the discussion concludes with a touch of the process approach: readers are advised to look at themselves as writers and to consider some common causes and cures for any difficulty they may have with commas.

taken from Hacker (Andre's), pp. 350-376

Apprenticeship advocates readily concede the verisimilitude of all these rules. They admire the painstaking research which has gone into explicating this astonishing variety of points and counterpoints. When faced with a concrete writing problem, or when they must send their manuscript to a particularly hard-nosed editor, they may even look up a couple of concrete illustrations in such handbooks.

In short, apprenticeship practitioners think such rules admirable teaching tools--for classes in linguistics (supposing linguists have no better use for their time). As these practitioners see it, such fine rules are virtually useless for writers. It will never occur to these practitioners to tell their students that the use of comma is governed by rules, let alone to explicate these rules. In the first place, they may question the naive assumption that rules--either explicit or implicit--are involved in the writing process (cf. Bereiter, 1991). In the second place, even if they concede this point, and even if they can explicate these rules, they do their best to forget them, or at least, as in the case of philosophically-trained scientists, to compartmentalize this knowledge from their writing and teaching. They believe that such rules are no handier to a writer than the rule "that curvature of the path must be in proportion to the ratio of imbalance over the square of speed" is to a cyclist. Their students are innocent of such things as rules and exceptions on how to use and not to use commas. Their students learn to use commas from reading the works of master-writers, from reading aloud their own works, from intuitively correcting typical comma errors in their own papers or in exercise books, and from expert guidance in concrete writing situations (see below).

In sum, in sharp contrast to the apprenticeship approach to writing instruction, the explicative and self-conscious approaches agree that writing is fundamentally unlike playing piano. Writing, they imply, is simple enough, or counterintuitive enough, to require (1)abstract explication of rules and strategies, and/or (2)moving back and forth from creating a text to reflections on the writer, writing process, and audience. Given their popularity, one might expect that these approaches are based upon something a bit more solid than common sense and unsubstantiated theories. We shall now see that they are not.

Many scholars have noted that language is too complex to be deliberately and consciously learned one rule at a time. This argument has been made for the acquisition of grammar, spelling, phonics, writing style, and vocabulary (Krashen, 1993, p. 14)

Not only can we recognize 50,000 words on sight--and also, of course, by sound--we can usually make sense of all these words. Where have all the meanings come from? Fifty thousand trips to the dictionary? Fifty thousand lessons? We have learned all the conventions of language through using language, by speaking it, reading it, and making sense of it. What we know about language is largely implicit, just like our knowledge of cats and dogs. So little of our knowledge of language is actually taught, we underestimate how much of language we have learned (Smith, 1988, p. 182).

The speech of young children, and the untutored writing of people of all ages, clearly show that people know many rules implicitly (or not at all). Even competent grammarians help themselves to rules they don't fully know:

I have asked members of a number of different groups...to give me the rule for ordering adjectives of nationality, age, and number in English. The response is always the same: 'We don't know the rule.' Yet when I ask these groups to perform an active language task, they show productive control over the rule they have denied knowing....So [the grammar in our heads] is eminently usable knowledge--the way we make our life through language--but it is not accessible knowledge; in a profound sense, we do not know that we have it (Hartwell, 1985, p. 111; see also Ellis, 1990, p. 185; Smith, 1988, p. 182)

After decades of effort by linguists, later joined by AI specialists working on language understanding systems, no one has yet succeeded in working out a complete and valid set of grammatical rules for any language. If the rules are actually in our mind and if we all know them, is this failure to uncover the rules not remarkable? (Bereiter, 1991, p. 10)

Some great writers have never taken a single composition course while some mediocre writers have taken many. Mark Lester believes that "there simply appears to be no correlation between a writer's study of language and his ability to write" (Lester, cited in Hartwell, 1985, p. 115). More recently, Krashen presented "abundant evidence that literacy development can occur without formal instruction" (Krashen, 1993; p. 15, additional evidence is given in Freedman, 1993, pp. 226-7)

Writer's block lends additional evidence for the apprenticeship approach. In fact, the etiology of writer's block is reminiscent of stage fright. Blocked writers, like rule-bound, self-conscious dancers, are immobilized by rules:

One of the most dramatic differences between...high- and low-blockers is...the presence or absence of rigid rules. The teaching...of writing has too often and too zealously been reduced by English professionals and the larger culture alike to the teaching and envoking of rules....And the problem occurs not simply with grammar and style. The composing process itself is often reduced and simplified in textbooks because it is too complex a process to be presented in its multifaceted richness (Rose, 1984, pp. 89-90).

Freewriting

In a directive reminiscent of free association in psychoanalysis, freewriters are told: "Never stop to...think about what you are doing" (Elbow, 1973, p. 1). Some research supports the seemingly "surrealist position, that often enough a student's freewriting, compared to rehearsed writing, turns out more coherent, better organized, or more 'fluid'" (Haswell, 1991).

The explanations offered for the effectiveness of freewriting and related approaches are again reminiscent of psychotherapy: Invisible writing "is intended as a way of reducing self-consciousness and relieving a writer of some of the constraints that might distract his thinking while he composes" (Blau, 1991, p. 291). "Freewriting does seem to...keep the generating activities clearly separated from the analytical or editorial....[through] "what Elbow terms 'transparency,' of being able to escape the inhibiting self-consciousness that can interfere while material is being generated--or immediately afterward" (Mullin, 1991, p. 146). "Free and invisible writing...are precisely the sort of therapy that is called for to assist basic writers in overcoming their insistently premature and counterproductive focus on their readers and on the surface features of their discourse" (Blau, 1991, p. 296).

Thus, freewriting rightly expends much elbow grease against acquired inhibitions. It goes without saying, however, that lifelong apprentices--those lucky enough to have never been overburdened with rules, strategies, and self-consciousness--can direct their energies to more productive and rewarding tasks.

As we have seen, apprentices focus on the task at hand, not on spectators. The available evidence supports this feature of apprenticeship. Thus, Perl (1983) found that preoccupation with audience tends to impair writing. Often, writers who "focus on what they think others want them to write...do not establish a living connection between themselves and their topic" (p. 49).

Grammar instruction has been fashionable for ages. After all, doesn't it make sense to teach writing by drilling into learners the rules of language? However, the apprenticeship approach predicts the wastefulness of grammar instruction, and, as may be expected by now, the apprenticeship approach is right:

We need to attempt some massive dislocation of our traditional thinking, to shuck off our hyperliterate perception of the value of formal rules, and to regain the confidence in the tacit power of unconscious knowledge...Hoyt, in a 1906 experiment, found...no "relationship between a knowledge of technical grammar and the ability to use English and to interpret language."...In a 1959 article...John J. DeBoer noted [that] "the results have been consistently negative so far as the value of grammar in the improvement of language expression is concerned." In 1960 Ingrid M. Strom, reviewing more than fifty experimental studies, came to a similarly strong and unqualified conclusion....For me the grammar issue was settled with [the unqualified conclusion of a 1963 survey]: "the teaching of formal grammar has a negligible or...even a harmful effect on improvement in writing" (Hartwell, 1985, pp. 121, 126, 105). A series of studies...confirm that grammar instruction has no impact on reading and writing (Krashen, 1993, p. 22).

Practitioners of decontextualized and self-conscious instruction put themselves, and their students, through much dour work, forgetting all the while that there is a far more effective--and fun--way of understudying writing: reading. A survey of the literature led Stotsky (1983) to this remarkable conclusion:

Studies that sought to improve writing by providing reading experiences in place of grammar study or additional writing practice found that these experiences were as beneficial as, or more beneficial than, grammar study or extra writing practice (p. 637).

rm4"

oj off

Others go even farther:

It is only through reading that anyone can learn to write. The only possible way to learn all the conventions of spelling, punctuation, capitalization, paragraphing, even grammar, and style, is through reading. Authors teach readers about writing (Smith, 1988, p. 177). Reading is the only way . . . we become good readers, develop a good writing style, an adequate vocabulary, advanced grammar, and the only way we become good spellers (Krashen, 1993, p. 23).

According to this view, writing competence "is acquired subconsciously; readers are unaware they are acquiring writing competence while they are reading" (Krashen, cited in Freedman, 1993, p. 230). Moreover, besides modeling good writing, reading serves another vital function. Writers need

to project themselves into the role of the reader....to attempt to become readers and to imagine what someone other than themselves will need before the writer's particular piece of writing can become intelligible and compelling. To do so, writers must have the experience of being readers. They cannot call up a felt sense of a reader unless they themselves have experienced what it means to be lost in a piece of writing or to be excited by it (Perl, 1983, p. 50).

It would appear then that reading for pleasure is a necessary, and almost-sufficient, condition for a good writing style (but see, e.g., Crowhurst, 1991). This fact forcefully argues for the apprenticeship approach to writing.

An Argument by Analogy: Learning to Read

Certain crafts closely resemble writing. Showing that such kindred crafts can be best acquired through apprenticeship would strengthen the case for apprenticeship in writing. To provide that additional support, we shall now direct our attention, in that order, to learning to read, learning to speak a first language, learning to speak foreign languages, and learning artificial languages.

Many researchers and practitioners are convinced that reading can be best learned through apprenticeship:

[The learner is] an active and already partly competent sharer in the task of learning to read. Here the model is apprenticeship to a craftsman....the learner first undertakes the simplest parts of the job, then gradually more complex ones, increasing the share he can cope with and all the time working alongside, under the control of and with the help of, the craftsman. The apprentice does not sit passively with his mouth open; he works actively with the tools of his trade in his hand....After all, how do little children learn everything they do before school; to speak, to play, even to walk or eat? (Waterland, 1985, p. 6)....Children should be allowed to behave like apprentices--to be active partners in the task with the adult leading, not driving (p. 9)....In many ways the acquisition of written language is comparable with that of spoken language....the perfect example of pure apprenticeship....Reading cannot be taught in a formal sequenced way any more than speech can be (pp. 10-11)...I began to...turn theory into practice...Within a term, so successful was the children's response that colleagues in the rest of the school began using the apprenticeship approach with their classes (p. 17)....This response from children is the greatest success of the apprenticeship approach to reading and its greatest justification. Teachers who are now committed to what is still, in many eyes, a revolutionary idea have a responsibility to share experiences...I believe that the potential for this approach to learning has hardly begun to be explored....What about approaching writing like this? (pp. 47-48; see also Hyde & Bizar, 1989, pp. 59-60; Smith, 1988, pp. 198-9, 210, 303).

No one, to my knowledge, disputes the fact that speaking is acquired tacitly and unselfconsciously. This bolsters the apprenticeship model of writing instruction. After all, historically, writing may have originated as a means of recording speech. Also, writing still closely resembles speaking, and some of the greatest masters of the written word (e.g., Bertrand Russell) often dictated their books.

Foreign writers (e.g., Conrad, Koestler, Nabokov, Polanyi, Popper, Rand in English; Iskander in Russian; Bialik in Hebrew) sometimes reach the pinnacle of their profession. They often fail to master the sounds of their adopted language, but their writing is accent-free. For the most part, their mastery of the written word is not achieved through explicit, self-conscious, instruction. Thus, Freedman (1993) cites evidence that "nearly all second language learning entails...the subconscious inferring of the rules of language use on the basis of comprehensible examples of the target language during the process of authentic language tasks" (p. 230).

As we have seen, although apprenticeship involves for the most part tacit learning, it does contain an explicit component: unlike decontextualized abstract rules, explicitness can be of immense value--when it is used sparingly, when it accompanies and echoes practical experiences, and when it is tailored to the developmental needs of the learner. Some research suggests that foreign language instruction at its best relies on a similar mix of authentic tacit and explicit components (Ellis, 1990).

An Argument by Analogy: Artificial Language

Psychologists have studied the ability of subjects to learn artificial languages, usually constructed of nonsense syllables or letter strings....mere exposure to grammatical sentences produced tacit learning...[but] subjects given the "rules of the language" do much less well in acquiring the features than do subjects not given the rules. Indeed, even telling subjects that they are to induce the rules of an artificial language degrades performance (Hartwell, 1985, p. 117).

There is, thus, overwhelming evidence in favor of the apprenticeship approach to writing instruction. I have already explicated a few tentative rules of apprenticeship, but these only provide a broad philosophical framework. Tacit mastery of this framework is essential to good teaching, but it does not yet tell us how this framework can be applied in settings as disparate as dancing, carrying out a scientific experiment, and writing.

This framework, to begin with, points to the importance of motivational, social, and affective factors--all too often set aside in the traditional classroom. It tells us that writing can be best mastered by students who feel that it is relevant to their own lives and concerns. It suggests that certain practices, despite their popularity, seeming reasonableness, and many defenders, must be used only sparingly. These overused practices include out-of-context excessive teaching of rules and strategies and paralyzing novices with over-attention to themselves or to their audience. Explicit instruction and reflections must be used sparingly, appropriately, and in context. Instead, writers should focus almost exclusively on generating and editing a text.

In the apprenticeship classroom, grammar, abstract rules and strategies, and metareflections are kept at arm's length. Apprentices focus on what they want to say, not on how they are going to say it. They revise their texts by relying on common sense and intuition, not on an inaccessible and confusing multitude of explicit rules and strategies. When apprentices commit glaring errors, instructors or peers bring such errors to their attention, but again, the appeal in such cases is to common sense, to usage, to aesthetics, or to parsimony, not to formal rules. Apprentices don't waste their time studying rules, and exceptions thereof, for the proper use of punctuation marks. They are neither taught, nor tested on, the spelling of such words as "concede" and "quibble." They are guided towards improvement of their texts, and the texts of others, in the same way that a sculpture apprentice is taught to chisel a statue. Whenever possible, they are coached in small groups and in one-on-one conferences.

When computer labs are available, and classes are small (twenty students or less), the problem of delayed feedback can be readily overcome through computer-assisted writing workshops. These workshops mimic more closely apprenticeship in non-writing contexts, speed production and editing of texts, strengthen wordprocessing skills, and provide writers with immediate feedback from fellow writers and from a floating instructor, thereby allowing writers to rapidly produce successive drafts of their assignments.

Overcoming Misconceptions.

According to Mike Rose (1984), "many of our students have developed narrow or distorted conceptions of the writing process....our students' misconceptions profoundly affect their growth as writers" (pp. 87-8). This problem afflicts all writing classrooms, but it is particularly acute for practitioners of the apprenticeship approach. Before coming to our classroom, writers have often been urged to apply an unmanageable horde of abstract rules, to consciously employ expert strategies, and to incessantly reflect on their own writing. They may feel uncomfortable in a writing class which spends much time on the reading of contemporary short stories and little time on grammar. For the foreseeable future, then, practitioners of the apprenticeship model must clarify their educational philosophy to their students. They must also take the persistence of misconceptions into account, and rely on conceptual change strategies to overcome them (Posner et al., 1982)

Situated Learning

Old-time apprenticeship is successful, in part, because apprentices often wish to acquire a craft. Their economic future, standing in the community, and professional pride may hinge on successful completion of their training. They need not be cajoled into acquiring the craft; in Japan, for instance, learning traditional arts is often described as "stealing the master's secrets" (Singleton, 1989).

Writing novices should likewise view writing, and writing instruction, as useful, relevant, and empowering. The common practice of providing a real audience helps to achieve this goal. Here, students' writings are read aloud in class, shared in small groups, or placed in class publications. Instruction can also make use of writing tasks students encounter in real life; students can write and edit letters to family and friends, papers for other classes, or work-related writings. Another approach involves team-teaching arrangements between a writing instructor and an instructor in another discipline. For instance, in a combined biology/writing class, learners write about biology and receive, in this context of an actual writing task, instruction and feedback from both instructors.

Rule-Free Grammar Instruction

Instead of learning an unmanageable plethora of abstract rules, apprentices often edit ungrammatical or inelegant sentences. Throughout the process, no appeal is made to explicit rules. Some of the sentences are verbose, and apprentices discover on their own, or recognize, the superiority of a more concise version. Dangling modifiers are explained by analogies to situations where such modifiers make no sense or by appealing to apprentices' linguistic intuition. Other key problems are similarly worked out on the intuitive level, without mentioning once such terms as "subjects" or "prepositions." In a portion of the final test, apprentices can be asked to edit a "short story" which is an amalgamation of most of the errors they have encountered throughout the term.

Instead of dry, abstract, and largely irrelevant recitation of rules, this problem-solving activity is productive and enjoyable. This activity can, as well, be easily converted into a competitive game. Indeed, one of the best ways, in our experience, to gain converts to the apprenticeship approach of writing instruction is precisely through this engaging rule-free activity of learning to recognize stylistic errors.

As we have seen, "writing style is not consciously learned, but is largely absorbed, or subconsciously acquired, from reading" (Krashen, 1993, p. 73). The implications of this for composition apprenticeship are profound. A writing class ought to be, in part, a reading class. And the goal is not only to have students read for our class, but to foster a habit of lifelong reading. Needless to add, reading instruction at its best follows the apprenticeship approach (e.g., Smith, 1988; Waterland, 1985).

Besides providing invaluable models of good writing (Smith, 1988), and besides providing invaluable insights into readers' needs (Perl, 1983), the readings we choose can be used as springboards for the students' own writing. For example, students can summarize a reading or respond to it.

To develop a keener appetite for reading well-written essays and literary works, the readings accompanying a writing class must be captivating (see Pintrich et al., 1993). They must include pieces the instructor is excited about, for students will be more impressed by the instructor's genuine enthusiasm than by anything else she does or says. The ideal collection of readings must serve as an excellent prototype of written language. It must be flexible---allowing instructors to learn from experience and add or delete material from term to term.

Besides exposure to well-crafted products and performances, an apprentice needs to know how these products and performances are generated and improved. Similarly, besides reading fine texts, writing students benefit from seeing accomplished writers at work, and from these writers' retrospective reflections. Students may benefit from examining successive revisions of short stories, essays, or poems. They may benefit from autobiographical portraits in which writers describe critical turning points in the acquisition of their craft. They may benefit from seeing their instructor at work. For instance, by using a laptop computer and an overhead projector, a writing instructor can provide an immediate model of the process of writing and editing. As before, the goal here is neither to distill rules nor metareflections, but to provide a living model of a craftsman at work.

As we have seen, the apprenticeship approach does not banish explicit teaching and metareflections from the composition classroom. Explicit rules are sometimes vital in an authentic context (Ellis, 1990; Freedman, 1993), or when they make writers aware of rules which they have failed to tacitly acquire. Advanced writers, in particular, may

benefit from some explicit knowledge of rules and strategies of their craft, and from a modicum of self-reflection.

For instance, the injunction to avoid the passive voice in one's writing is too abstract, too devoid of context, too plagued by exceptions "for the people on the spot, that is, for writers facing some concrete writing problem" (to paraphrase Feyerabend). But when given as editorial feedback on an actual text, and when the change can be readily implemented and shown to improve that text, this injunction is useful.

This essay has argued that controversies about such issues as explicitness, genres, grammar, and metacognitions can be resolved by returning to apprenticeship as the overarching paradigm of writing instruction. One can imagine countless reasons for not embracing apprenticeship, including vested interests, the difficulty of seeing time-honored practices in a new light, the apprenticeship model's counterintuitive emphasis on contextual learning, and the effectiveness of alternative models. Nonetheless, the apprenticeship model offers a unified theory of writing instruction. It is consistent with much of what we know about both language and learning. It promises to make writing instruction more enjoyable and fruitful to both learners and teachers. It assimilates the best features of traditional and process instruction while avoiding most of their pitfalls. For these reasons, it may one day prevail.

rm4"

Acknowledgments

I thank Prof. Shreedhar P. Lohani, two anonymous reviewers, and Prof. Sharon Helmer Poggenpohl, for helping me clarify and sharpen my argument. I am grateful to Dr. Lohani for believing that apprenticeship can, and should, be applied on a large scale. The writing of this essay was supported, in part, by a Senior Fulbright Scholarship.

2. This is a paraphrase of an actual assignment in a freshman composition program.

3. Besides common sense, this assumption is traceable to uncritical acceptance of fashionable trends in cognitive psychology: "A major direction in current cognitive research is to attempt to formulate explicitly the strategies and skills underlying expert practice, to make them a legitimate focus of teaching in schools and other learning environments (Collins, Brown, & Newman, 1989, p. 480).

4. Bereiter valiantly sets out to save as much of his old paradigm as he can by insisting that rule-bound cognitive psychology and connectionism can be united. But such a reconciliation is as likely to bear fruit as was Tycho Brahe's attempt to unite geocentricism and heliocentricism.

5. For more concrete, ready-to-use, examples, please consult my manual (available upon request).

Collins, A., Brown, J. S., & Newman, S. E. (1987). Cognitive apprenticeship: teaching the crafts of reading, writing, and mathematics. Center for the Study of Reading, Technical Report No. 403.