Grades

1. Most societies subscribe to the premise that individuals are substantially unequal in their merits and thereby also in their claims for desirable benefits such as wealth and power. In a ‘pre-modern’¹ society, merits are decided by stable factors, such as ancestry, caste, race, and gender; inequality is permanent, and the ‘upward mobility’ of an individual is rare in both theory and practice. In a ‘modern’ society, merits are decided by unstable factors, such as achievement, enterprise, and initiative; in theory, inequality is merely accidental, and ‘upward mobility’ is offered to all individuals.

2. Yet in practice, the 20^th century has shown how ‘modern society’ reaches a stage of intense population, consumption, and environmental depletion where upward mobility is achieved only by individuals who have survived fierce competition. Otherwise, social inequality is forcefully entrenched even though it contradicts modern ideals of justice in a democracy and often produces the same divisions as ‘pre-modern’ inequality by race, gender, and so on. So some institutional means are required to make inequality appear as a normal and natural factor of the ‘social order’ and indeed as a deeper expression of social justice.

3. The leading institutional means turns out to be universal education. There, all individuals are required to be sorted out by ‘assessments’ of their ‘merits’, from childhood until the threshold of adulthood. Education is thus implicated in underwriting social order and social justice, provided that its methods can implement educational justice. And justice is best served if the responsibility for degrees of success or failure is seen to rest upon each individual learner in fair competition with the others (cf. ¶ 7, 21). In theory, everyone is offered an ‘equal opportunity’ to succeed or fail; in practice, the outcome is decided by one’s individual ‘intelligence’ and ‘aptitude’, or else by one’s ‘diligence’ and ‘obedience’.

4. The sorting process produces the widest spread of results when the demands of schooling are so difficult that a substantial portion of the learners cannot achieve them, or only partially and sporadically — like an obstacle course with hurdles designed to make runners trip and fall during the race (cf. ¶ 26). For this reason, degrees of success or failure reflect not just an abstract potential like ‘intelligence’, but concrete abilities to operate near the threshold of overload. This threshold impends when the demands of physical or mental processes surpass a person’s available resources, and performance undergoes a temporary degradation. Learners who do quite well in a relaxed, supportive environment with the freedom to revise and ‘debug’ their work might suffer overload and do quite badly in a high-pressure, isolating environment, such as a strenuous ‘final examination’. If so, the examination results would seriously underestimate their real abilities; the test situation shears off or flattens out potential peaks of success and produces artificially deflated results (cf. ¶ 38f). Yet those same results are certified to be the best indicators – and in some institutions, the only indicators – of the learners’ ‘achievement’, or indeed of their ‘potential’ for achievement. The examinations thus identify young people as ‘high achievers’ or ‘low achievers’; and once acquired, these labels are hard to change (cf. ¶ 21).

5. By this logic, education can claim to maintain ‘high standards’ by according success to some and failure to others; winners shine more brightly next to losers. The logic is reminiscent of social Darwinism with its notions of the ‘laws of nature’ and the ‘survival of the fittest’. Learners who fail would be ‘naturally unfit to survive’, and would have no claim upon education or society. Admittedly, they usually don’t just cease to ‘survive’; they live out a lifetime of programmed frustrations and denied opportunities.

6. Still, the logic of educational justice requires that success and failure be regulated in strictly educational terms. The social inequalities facing the learners in the outside world are routinely disregarded. Upon entering school, all learners are officially considered equal — the dominant visual emblem being the school uniform. Upon leaving school, they will have proved themselves unequal; and they alone are responsible if these inequalities affect their future lives.

7. We might thus appreciate the characteristic tendency of schooling to isolate or at least insulate itself from everyday life beyond its walls. Its conventions of conduct and discourse are pungently distinct from what many children experience among family or friends. Abstract theoretical knowledge about matters like ‘mathematics’ and ‘grammar’ is heavily favoured over concrete practical skills and everyday common sense; indeed, school knowledge may seem so remote from the real world that learners do not expect to use it at all in later life (cf. ¶ 25). Most importantly, the learners may at any time be subjected to means for ‘assessing their merits’. These factors of insulation imply a paradox: schooling withdraws to some distance from social life, yet purports to be best at assessing people’s ‘merits’ for the benefits of social life.

8. The primary motor for constituting and insulating education as a separate world is educational policy, striving to codify and standardise practices with ‘rules’, ‘regulations’, ‘procedures’, and so on. The ‘administration’ believes itself authorised to regulate almost everything: who shall be where and when, who shall do what and how — even what clothing they shall be wearing. Naturally, each wave of policies creates an information deficit about whether and how far the policies are being implemented. So another wave is launched to monitor the implementation; and yet another to monitor the monitoring. Soon, the activities of teaching and learning become overlaid with networks of information channels flowing in and out of the administration.

9. We seem to be confronting another paradox. The administration assumes from the outset that education cannot or must not operate without policy, yet tacitly suspects it of doing just that. So policy gets distracted from its original mandate to organise the processes of teaching and learning, and becomes preoccupied with overseeing its own applications. Yet increasing the number of policies also increases the probabilities of at least some of them not being followed, especially by staff who are already working at the highest capacity. And each policy to oversee policy intensifies the alienating impression that the administration does not trust its own staff.

10. A major challenge for policy is to guide education in officially underwriting social justice whilst unofficially underwriting social injustice. Policy seeks to paper over the discrepancies between the benefits individuals aspire to gain and the benefits the society is willing or able to dispense. So education is supposed to convince young people that limited merits entirely justify limited benefits in later life; and that education holds the ideal measuring stick to show exactly where the limits lie (cf. ¶ 7).

11. This task becomes precarious when the society withholds its benefits on a massive and painful scale. Today, the majority of ordinary citizens face a bleak future of competing in the ‘global free-market economy’ just to secure a laborious employment at low wages. Meanwhile, wealth and power are being voraciously concentrated among a tiny super-rich minority who restlessly move their residences, assets, and businesses wherever they can best be free of social responsibilities like paying income taxes and providing their employees with job security and pensions (cf. Reich 1993; Martin and Schumann 1996).

12. At such a stage, education cannot realistically be expected to convince learners that the limitations and inequalities it represents can be the outcome of justice. Swiftly rising proportions of learners resent or resist having assessments pronounced on their ‘limited merits’, and education is publicly perceived to be in the throes of a ‘crisis’.

13. Not at all surprisingly, the so-called ‘education crisis’ has fostered a fresh burst of policy-making. Administrators are determined to demonstrate that current policies are sound and would indeed ensure ‘high quality’ except that they are not being properly implemented — just what policy-makers have suspected all along (cf. ¶ 9). Programmes, departments, or whole schools and colleges are now being ‘assessed’, and ones with low scores strongly pressured to ‘improve’. Somehow, improvement is expected from increasing already heavy work-load of the staff with new tasks of record-keeping, documenting, reporting, measuring, verifying, certifying, attesting…

14. The unappetising alternative would be to acknowledge that the crisis might have been caused by overburdening education with gratuitous or unrealistic policies. We must at least concede that prevailing policies have failed to prevent the crisis, and we need to examine the possible explanations. Politicians and their spokespersons in mass media of newspapers, popular books, and so on, accuse either the learners or the teachers, or both groups, of just not doing their work properly. Instead of exploring plausible reasons why not,² the same sources advocate stern policies for forcing these errant groups back to their duties or else expelling them from the system.

15. The teachers and learners themselves could offer rather different explanations. When professional opportunities and social benefits grow scarce, these groups are naturally alarmed if they believe education does not prepare young people for the realities of their futures. And when the concrete market-value of a degree or a diploma has become doubtful, the motivation for pursuing it through an obstacle course of theoretical abstractions and high-pressure tests is difficult to sustain. Nor can one readily see salvation in new policies for enforcing old policies that are no longer adequate for a rapidly changing society. Instead, teachers and learners are experimenting with innovative methods and programmes, which of course can be decried for not following ‘established policies’ (cf. ¶ 38, 70).

16. What is in fact called for is a realistic re-assessment of the impact of current policies, especially where they may be mediating against social and educational justice. I propose here to assess one crucial area of policy, namely grading, in terms of its logic — what premises, preconditions, and consequences it entails, and how these relate to each other or to the rest of the educational enterprise.

B. Grading and testing

17. ‘Grades’ (or ‘marks’, ‘scores’, and the like) can symbolise educational justice so readily because Western culture entertains a profound respect for abstract numbers. On one level, numbers and statistics appear totally objective, neutral, and rational. On another level, the human ability to perform precise operations upon abstract numbers is thought to be a valid proof of ‘high intelligence’, a factor itself famously expressed by a number termed a ‘quotient’. After all, such operations are difficult and tedious; they afford many opportunities for error; and their theoretical and abstract nature sets them well apart from everyday knowledge and common sense (cf. ¶ 7). So numbers permeate modern education, whose ultimate product is the starkly simple ‘grade point average’ (‘GPA’), a single number held to summarise your ‘merits’.

18. But if grading by numbers is to underwrite educational justice, some fundamental premises would be entailed:

(e) the quantities can be dependably summed or averaged into a single number representing a fair and accurate assessment of one’s total education outcome.

Totally reliable and consistent means would thus be required both for converting quality into quantity (to get from a concrete performance to an abstract number) and for converting quantity into quality (to interpret the number as a description of the learner).

19. The hallmark of quantification is the decimal scaling from 1 to 100, whose reliability and exactitude also command great respect in Western culture. But applying the scale to the process of ‘grading’ may not prove so logical. The most transparent application I can imagine would be for a ‘test’ with exactly 100 separate items, each item posing the same degree of difficulty and entailing a secure distinction between ‘right and wrong answers’. For a thought-experiment, we could envision a ‘mathematics test’ with 100 problems in ‘long division’, each having a solution expressed as a positive whole number. To ensure fairness, enough time would be allotted for each problem to be addressed by each learner; and learners would be effectively prevented from obtaining answers in advance, or copying them from each other with or without collaboration. By this logic, the total test comprises 100 mini-tests; and the ‘right or wrong answer’ on any one of them would not prejudice the others.

20. Subsequent marking by the teacher would be perfectly objective and efficient: you compare number with number, and at the end you total up the matches. Then you can put the totals onto the familiar scale sectored by decimal logic: 90 to 100, 80 to 89, 70 to 79, and 60 to 69, and 0 to 59.

21. So far, no problems. But the logic begins to pinch when the quantities get interpreted as qualities, the usual ones being ‘excellent - good - average - poor - failing’; and when these qualities are complacently used to describe the learners too. No logic can readily explain the reason why a single error should convert a ‘good’ performance (at 80) into a merely ‘average’ one (at 79); or a ‘pass’ (at 60) into a ‘fail’ (at 59). Nor again the reason why the performance should define who is a ‘good student’ or a ‘poor student’. These labels merely salve our conscience about questions of justice by suggesting once more that a grade is the natural product of each individual (¶ 3, 7), especially for a ‘failing’ student who suffers the penalty repeating a whole year-long course.

22. Nor could logic readily explain the reasons why human merit should be demonstrated by one’s ability to perform ‘long division’ or similar numerical operations in a society whose members can easily obtain pocket calculators. All these reasons are rooted in the mission of the schools to quantify performance by counting ‘right and wrong answers’. Few means are better suited than tests requiring numerical operations difficult enough to produce the proportion of ‘wrong answers’ and fill out the ‘grade curve’ (cf. ¶ 35ff). The ‘maths test’ is relatively ‘closed’ in respect to what learners must do and what results they must get.

23. For the same reasons, the math test covers not merely ‘intelligence’ in the Western perception (¶ 16) but also obedience. For many operations, there is no other rational way to obtain the solution than to perform the steps enjoined by the teacher or textbook. To do ‘long division’, you must decompose it into recursive steps of multiplication and subtraction; and the number of recursions must reflect the number of decimal places in the dividend and the divisor. Common-sense estimates or trial-and-error, as they are practiced in everyday life, would be not just disobedient, but inherently useless.

24. Again for the same reasons, mathematics is an insulated and self-contained enterprise, just what education prefers (cf. ¶ 7f). It can be regarded as a special talent reserved for unusually ‘smart’ people. The rest are proffered an alibi to solace their feelings of resignation: nature has decreed that they are ‘not good in math’, and they must face this unlucky fact of life (cf. ¶ 55). If they do, they have truly internalised the limits wherewith they are circumscribed by their schooling.

25. Yet nature has not made the intellect of human beings in general ‘good in math’ if this means precisely manipulating large and abstract numbers for no practical purpose. So to judge the ‘merits’ of young people on this basis is to discriminate against them for not possessing abnormal skills that are largely irrelevant to real life. Moreover, such judgements reinforce the larger message that school knowledge is generally quite distinct from real-life knowledge, no matter what the particular subject matter might be (cf. ¶ 7). From there a small steps lead to the conjecture that the grades assigned in school are not a fair assessment of individual merit for social benefits.

26. Let us return for a moment to the analogy of a race course with difficult hurdles to be jumped (¶ 4). The talent to succeed in such a race is obviously a rare accomplishment. The runners must have undergone strenuous training directly at the threshold of overload, and suffered many setbacks and bruises. They were not deterred by the sheer difficulty and did not have to agonise over whether nature has decreed that they are ‘good at racing’. But — unlike the texts and exams in the schools — whether or not they choose to compete in races is their own decision; and their racing talents are not interpreted as a general measure of their overall merit as human beings.

27. Most basically, I cannot readily grasp the logic that an education whose assessment is concentrated on competition near overload should be a sound measure or preparation for future life. Aside from high-performance sports, most careers offer success to people who co-operate smoothly with others and contribute to large joint projects that naturally require lengthy revising and debugging. If such a project had to be prepared and submitted under the same conditions as a school examination, its quality would be critically damaged.

28. Moreover, the dominant logic of grading encourages learners to suppose that each block of abstract knowledge need be memorised only in order to pass the next test, and can be forgotten soon thereafter. The product and goal of learning is the isolated grade and not the cumulative skill; and the numbers count more than the achievements. Such may be the reason why adults have great difficulty recalling the content they worked so hard to ‘learn’ across the years of their schooling.

29. The rhythm of ‘memorise – test – forget’ is reinforced by the tactics involved in adapting the content of schooling to fit the logics of grading. Converting human knowledge into abstract theoretical material for a test with a set of right and wrong answers entails disconnecting terms, concepts, or ideas from the wider contexts that are indispensable for cumulative learning (Freire 1970). Even when learners sincerely intend to retain knowledge over a long term, they must contend with each stressful episode of writing a test near the threshold of overload and then seeing their ‘errors’ publicly tabulated, often without fully understanding what is supposed to be ‘wrong’. This tabulation in turn instils negative affects like anxiety or depression, which the learners seek to repress by forgetting the entire episode. Here a further block would be imposed upon cumulative learning.

30. I have argued so far that the underlying logic of grades and grading as conventionally practiced need not underwrite educational justice and may indeed foster injustices for which the learners are held responsible. The same dualism is still more acute for the logic of several ancillary constructions in educational policy, all of which presuppose the justice of grading.

31. The logic of grade averages would be that fluctuations among the individual grades of a learner should be balanced out, whereby the low ones bring down the high ones and vice versa. Apparently, extremes at either end are considered unseemly or doubtful, perhaps as flukes. The concept of ‘average’ gets interpreted literally in its statistical sense — the most common or frequent range — yet also figuratively as mediocrity (cf. ¶ 36).

32. This logic ignores a fundamental lack of symmetry between failure and success on a test. A failure may be caused by accidental or extraneous circumstances, and is correspondingly likely to be a real fluke. Learners may have suffered from overload as described above. Or, they may have been hindered from adequate preparation by intrusive distractions such as other examinations on the same day or acrimonious quarrels with friends or family.

33. A success, on the contrary, implicates a deliberate and concentrated effort to accomplish the requisite preparation and to resist the forces of overload and distraction. The success therefore reveals far more about learner’s actual potential, and is far less likely to be an accident or a fluke.

34. Averaging test grades can accordingly mean allowing minor accidents to cancel out major achievements. The danger is most acute when the failure is expressed in the bottom range upon the decimal scale of 100, say, between 0 and 30, since it will certainly convert into a failure the next grade and possibly the next grade after that. The danger is less if the averages get calculated after the grades have been compressed onto the scale of 0 to 4, but a single failure through a fluke will still be far more devastating than its true importance would justify.

35. The logic of grade curves would be that the results of a test or assignment should always be spread across a consistent pattern: few marks being very high or very low, more marks being somewhat high or low, and a cluster of average marks in the middle. Here, the total logic of grading becomes circular and self-confirming. The differences among learners are not documented by the actual results; the results are constructed by interpreting them so as produce the prescribed differences. High, low, and average can get manipulated and moved to fall where they fit the curve rather than where they document the learners’ real achievement.

36. The ‘grade curve’ poses a peculiar threat to educational justice. At the top end, it may deny a portion of high marks when ‘too many’ students have honestly achieved them. At the bottom end, it may seize upon students just to fill up a quota for ‘poor’ and ‘failing’ although their achievement would deserve better judgements. Even in the middle, a heavy proportion of ‘average’ grades presents a disquieting image of the teacher, whose efforts have apparently been attended with mediocre success. This image is sharply at variance with the concern indicated on staff assessment questionnaires that a teacher’s merit should depend upon communicating clearly and effectively, and upon giving special attention to the learners’ problems (cf. ¶ 70).

37. Also, the ‘grade curve’ can readily be used to revise the level of an examination after the fact whenever the results might look disproportionate without a curve. The test might be faulted for being ‘too easy’ or (far less common) ‘too hard’. Or, the scores might all get bunched up in the middle, as if a test could be ‘too medium’. Such faults can be ironed out by ‘curving’ the grades, but at the price of undermining the logic of quantification whereby the relation between a performance and a quantity is believed to be stable (cf. ¶ 18).

38. Finally, the logic of prohibiting grade inflation hinges upon the logic of the grade curve even where the latter is not officially enjoined. The term ‘inflation’ was presumably picked to suggest that grades have been artificially pumped up to levels not justified by performance. But my own discussion would suggest instead that the normal procedures of grading by high-pressure tests lead to an artificial grade deflation and rates learners below their real potential (cf. ¶ 4). If so, innovative teaching methods which assist the development of that potential will naturally produce results falling significantly over the statistical mean for a school or a department using conventional methods (cf. ¶ 15, 70). Such methods do not ‘inflate’ assessment but rather free it from deflation.

39. For these reasons, a travesty of justice would be committed if a set of supposedly ‘inflated grades’ got arbitrarily lowered by administrators who neither know the means of the grading nor have inspected the relevant documentation in tests or assignments. Such an act sacrifices to administrative expediency the central premise of quantification based upon objective marking, and substitutes an assessment with no marking at all. The next step in this logic would be for the administrators to simply assign the grades they prefer right away and dispense with the processes of teaching and learning whose reliability they misprize.

D. Situated learning and successive approximations

40. Most skills acquired outside formal education are approached by the logic known as situated learning (Lave and Wenger 1990). A skilled person collaborates with one or more unskilled persons in performing activities that will help the target skill to emerge. The activities are simplified or partial versions of the target, designed to stay well within the capacities of the learners at each step.

41. Such has long been the dominant logic in ‘pre-modern’ societies where skills are cultivated by master craftsmen, such as weavers, bricklayers, stonecutters, and blacksmiths. The normal outcome is a reliable success — not ‘average’, not ‘poor’, and not ‘failing’. A master would have been amazed and angered had he been accused of ‘low standards’ by a committee of administrators who diagnosed ‘grade inflation’ in the certifications he was awarding to his apprentices.

42. Although these skills are specialised and refined, they are not insulated from future lives in the outside world. They are, on the contrary, situated in the very fabric of life; and learners could directly experience the motivation to perform at their very best. Whenever a piece of work is not of high quality, the learners do it over until it is. They could not just accept a ‘low mark’ and move onto the next step, because the total process is cumulative, and each step has to be fully mastered in order to perform the subsequent ones. Nor are the errors counted up and the totals entered in permanent public records.

43. In many areas of modern education, a situated step-by-step learning could be extremely productive. But so far, its application has been limited mainly to some high-level professional programmes such as engineering and medicine; and even there it has had to compete for attention with ponderous contingents of abstract theoretical work.

44. In principle, the logic of situated learning would conflict with the tendencies of ordinary schooling to insulate itself (cf. ¶ 7f) and even more with the impediments it places on cumulative learning (cf. ¶ 28f). The conflict is surely most debilitating when the subject-matter happens to a non-native language. Each language is a complex interactive system whose component skills can be activated and applied during fluent performance without demanding much conscious attention. In the native language, mastery is achieved and tuned during many years of situated learning. Learners of a non-native language face the powerful challenges of constructing new skills whilst resisting unhelpful interference from native skills.

45. The new skills can only be developed through cumulative learning. A skill cannot be detached and disregarded if it has not been mastered at the appropriate time. Nor can a skill be constructed upon a disconnected episode whose results are merely ‘average’, let alone ‘poor’. The only rational approach for learning a non-native language is to make sure each skill is reasonably secured and frequently refreshed — and doing so will not produce a ‘grade curve’ (cf. ¶ 70)

46. For closely related reasons, high-pressure ‘tests’ are not appropriate or reliable indicators of achievement. Non-native language performances are keenly sensitive to minor disturbances, which may reverberate and trigger a temporary state of overload and degradation. As every experienced teacher has observed, learners may possess theoretical knowledge of the ‘pronunciation’ or ‘grammar’ of the language and yet fail to apply it in practice. Paradoxically, they may even produce errors in the very items they are consciously struggling to control, and be visibly dismayed and frustrated to hear what comes out.

47. Moreover, a language performance represents an essentially open task where the number and nature of specific actions are rarely well-defined. Imposing extensive restrictions, such as reciting prefabricated ‘dialogues’ out of textbooks, removes the activity from the domain of real-life communication. At most, we can elicit superficial imitations of native speakers in order to minimise the numbers of ‘errors’; but the numbers may only shoot up again as soon as the learners attempt independent performances.

48. The assessment of performances is therefore more problematic for learners of a non-native language than anywhere else in the curriculum. The logic of grading encourages the exclusively negative tactic of deriving a number by counting ‘errors’ and subtracting the total from an arbitrary number, usually 100. This tactic does not attempt to quantify the learners’ positive achievement. Mistakes get assigned a value and importance out of all proportion to their real cognitive significance or to the immense quantity of possible items that could be either ‘right’ or ‘wrong’.

49. If the target language happens to be English, formidable hurdles are posed just by orthography. There, the relations between letters and sounds are notoriously unreliable in themselves and probably quite disparate from those in the learners’ native language. The effects are striking when native speakers of Arabic follow its orthographic strategies in struggling to spell English words from the sounds. The problems are intensified because the transcription of Arabic sounds in the Roman alphabet has never been standardised among competing alternatives. For example, my own fourth-year students gave Romeo’s lady a whole gallery of names: ‘Julit, Julite, Juleiet, Joluet, Jollet, Joulet, joeliette, Joliyat’. The perception and notation of vowel sounds quite naturally pose the greatest difficulties.

50. However chaotic and messy, English orthography at least offers a mechanical division between ‘right and wrong answers’, and most disputes can be settled with a dictionary. You can count up the ‘errors’, even if you can’t say which are the more or less serious or important ones. But the division fades when you adopt a more holistic view of an authentic language sample, such as this ‘entrance test’ written on the topic of ‘camping in the desert’:

[1] Suddenly I saw asmall girl in the desert lonley. She was fraid and cryed, I tryed to know anything about her but I fault. Inaddition, I can’t go away and left this small girl from himself in the desert, so [… ] we start the opperation of research about anything could help us to find their perants. […] we arrived to the police centure, there I hearded some sound came from small room inthe office. I went they, and I found woman sat onthe chair, her face was sad, so I neared to him and asked him; If she lost any child when they went ina trap in the desert

Understanding what this unfortunate applicant meant to say is easy enough. A normalised version can be created with fairly minor cosmetic editing, e.g.:

[1a] Suddenly I saw a small girl alone in the desert. She was afraid and cried. I tried to find out something about her but I failed. In any case, I couldn’t go away and leave this small girl by herself in the desert, so […] we started the search operations that could help us to find her parents. […] When we arrived at the police centre, I heard some sounds coming from a small room in the office. I went in there, and I found a woman sitting on a chair. Her face was sad, so I moved near to her and asked her if she had lost a child when she went on a trip in the desert.

Of the 110 words in the original [1], some 40 have been changed in [1a], which has 120 words. But these numbers do not give a meaningful assessment of the performance. Most of the deviations from native-speaker norms reflect a modest number of problems often detectable in the written English of native speakers of Arabic. Beyond the spelling of individual words (e.g. ‘lonley, opperation’), we can see where the boundaries between syllables and between whole words are set differently in the two languages, e.g., Arabic fusing a Preposition with a following Article (e.g. ‘inthe, onthe’). In respect to Grammar, the same sample indicates the routine confusion about the choice of Articles (e.g. ‘I found woman, sat onthe chair’) Prepositions (e.g. ‘from himself, arrived to’); the Gender and Number of Pronouns (e.g. ‘him, their’ for a ‘girl’); the Tenses of Verbs (‘start, hearded’); and the joining of Relative Clauses with no Relative Pronoun (‘I found woman sat onthe chair’).

51. Applicants of college age who can write no better on a decisive occasion have obviously had little practice with active written prose in English during their schoolwork, and little useful feedback on whatever writing they did. According to the reports of my own students, most of the writing has been passive: copying from blackboards, textbooks, or exercise manuals. Moreover, the relative absence of active prose-writing in the schools is probably related to the problems it poses for grading by numbers.

52. Further problems can arise from the insulation of education against everyday life. Learners who generally expect that the theoretical knowledge in schooling will not be put to practical use (¶ 7, 25) may adopt a similar view of the information they receive about topics like ‘English grammar’. On some abstract level, they know for instance that English does not fuse the Preposition with the Article in the manner of Arabic; but in a stressful setting like a entrance test, such knowledge is not applied or not consistently, just as would be predicted from the degradation due to overload (cf. ¶ 4, 29, 46).

53. Much of evidence I am currently collecting on learners’ ‘errors’ suggests they reflect spontaneous attempts to offset low fluency in written English by struggling to recycle the architecture of spoken Arabic. The results resemble the solutions of mathematical problems performed with an idiosyncratic, improvised arithmetic. In Arabic, the placement of Articles does not correlate securely with definiteness; nor do the forms of Verbs correlate securely with temporality — the term ‘Tense’ seems less appropriate than ‘Aspect’ (Wickens 1980: 38; Holes 1995: 176f, 188). But the learners do not know how to manage Articles or Tenses in English. These matters are extremely complicated, and the explanations one gets from teachers and textbooks fall far short of a complete or reliable account. Improvement can only come from active work with large amounts of authentic English (cf. ¶ 62ff).

54. If my assessment is valid, then penalising learners for the number of their ‘errors’ of a ‘test’ and handing out a string of ‘low grades’ cannot occur in the interests of educational justice. Even if I return their tests or assessments with detailed editing and correcting, they will not make much progress without a productive grasp of how and why those particular corrections were made. They need to reach a stage when they can recognise their own problems well enough to edit and correct their own work (cf. ¶ 64-68).

55. Active work with English prose might not be stimulated by assigning a ‘homework essay’ on a theoretical topic like ‘modern technology’ or ‘the benefits of education’. When not in the classroom, the students who most urgently need to develop their skills are also the most likely to get outside help, rationalised by the alibi that nature has decreed them to be just ‘not good in English’ (cf. ¶ 24). The alibi is self-fulfilling wherever these tactics block the development of skills.

56. One alternative I have piloted is to present a video whose content will become the basis for a student essay. The latter can be written only by students who are in my class, and cannot be copied from a library books or downloaded from the Internet. In return, the students would have plentiful language input to use even if (as typically occurs) they do not understand some of it.

57. In one instance, the presentation was a BBC film from the ‘Story of English’ series, narrated by Robert MacNeil. Some of the essays were (to say the least) colourful, witness these passages:

[2] The future of English is uncertain, some people were break up into different language which was not understanded for each other.

[3] A hundred years ago when the first English speaking migenery variability, the local tripes were spoken many language involve the contact language between a Uropian.

[4] Shakespeare used English and European stories in his plays, his language from ordinary people and Aristocrats he adds the humanity.

[5] Shakespeare considers one of the greatest writers in the history of English effected on the language. Shakespeare invited many words in his nice books and plays like Mutter and Mutter. He also has his own looks for the use of music and how to used it to prophet with the play.

The problems here are at least partly explainable from inexact perception of the soundtrack, e.g. ‘people worry’ => ‘people were’; ‘missionaries’ variety’ => ‘migenery variability’; ‘Measure for Measure’ => ‘Mutter and Mutter’.

58. For purposes of comparison, we could again construct some ‘edited’ versions:

[2a] The future of English is uncertain. Some people worry that it may break up into different languages whose speakers cannot understand each other.

[3a] A hundred years ago, when people first arrived speaking the missionaries’ variety of the English, the local tribes were speaking many languages and needed a contact language for communicating with Europeans.

[4a] Shakespeare used English and European stories in his plays. He derived his language from ordinary people and aristocrats, and covered the whole of humanity.

[5a] Shakespeare is considered one of the greatest writers in the history of English in his effect upon the language. He invented many words in his valuable works and plays like Measure for Measure. He also had his own views on how music can be used to enhance a play.

Besides the more mechanical types of orthographical and grammatical problems (e.g. such as the Tenses, Aspects, and Voices of Verbs), I have edited some rhetorical and stylistic problems, such as organisation of English prose into clauses and sentences by means of punctuation. This factor too sets English apart from Arabic: ‘until fairly recently, most Arabic books’ ‘had little or no punctuation’, in place of which ‘the language uses certain other devices that fulfil the same function’ (Wickens 1980: 14f).

59. At all events, the improvised English of my learners constitutes the basis upon they can most realistically be expected to make progress. The method I would propose is inherent in the logic of situated learning: performing activities which will build up the target skills whilst staying well within the capacities of the learners (cf. ¶ 40ff).

60. The method might be characterised as successive approximations, that is, sequential stages during which the learners’ versions of the language gradually but steadily assume more features of the target version. We scan versions for materials which can be exploited or modified to enhance the degrees of approximation at a rate the learners can manage without incurring overload. ‘Errors’ are treated like ‘bugs’ in a computer programme — as routine quasi-events to improve and put right rather than as pernicious items to record as evidence for a ‘low grade’ or a ‘poor performance’. This mode of treatment frees us from the arcane labours of trying to compute how many ‘points’ can be either ‘right or wrong’. Everything is either ‘right’ or ‘not yet right but soon will be’.

61. This outlook is productively reinforced by providing for each student a computer terminal with a current word processor (in our case, Winword in Microsoft Office 2000). The software automatically highlights all unknown words as possibly misspelled and at least some obvious problems with grammar (e.g. in Subject-Verb Agreement) and punctuation (e.g., capitalisation of the first word in a sentence). Also, a modest thesaurus can be accessed to give synonyms for a highlighted word. These aids are the language-based equivalents of the pocket calculator in mathematics, and the logic of using them is much the same (cf. ¶ 22). They obviate a waste of effort on problems already resolved by technology; and they are so widely available that students can reasonably expect to have them in future careers. Besides, these matters affect all users of English, including fluent native speakers. Surely I would be grossly unfair to demand from my learners better mechanical skills than I possess myself.

62. However, matters of usage and style are far less mechanical, and here I am piloting a different technology which can greatly increase learners’ exposure to authentic English data. I have adapted WordPilot©, a corpus-based resource program developed by John Milton at the Hong Kong University of Science and Technology (available at www.compulang.com/wordpilot.htm) and directly interfaced with Winword. The user highlights an expression and clicks the option ‘Examples’ in the Winword toolbar to see a set of authentic uses. As a data base I have compiled a 4.3 million-word corpus of famous British and Irish novels, some of which are read in the ‘Literature Programme’ of my department. I am hoping that a number of students might be able to enhance their fluency by giving closer attention to language combinations in these sources.

63. For demonstration I would propose the Verb ‘consider’, which was used by a student in an inappropriate way in sample [5] ‘Shakespeare considers one of the greatest writers…’. Taking the stem ‘consider’ as a query prompt I showed the class the large set (1645) of occurrences in the corpus compared to small set (26) of the Third Person Singular of the Active, ‘considers’.³This Verb form almost always had a Direct Object in the sense of ‘judge to be’; after that, the data showed choices among with an Object Complement in a second Direct Object [6] or in a Modifier with ‘as’ [7-8] or a Circumstance. An uncommon alternative appeared in the sense of ‘show concern for’ [9].

[7] It is very plain that he considers the good fortune of the engagement as all on my side

Fifteen times as common was the form ‘considered’ for the general Past Tense as well as for the Past Participle. . The Active uses covered the sense of ‘judged to be’ [10-11] but also ‘believed’ [12], ‘contemplated the prospect’ [13-14], and ‘looked at closely’ [15-16].

[10] he assured her with much earnest gravity that he considered the money as a mere trifle,

[11] Mr. Woodhouse considered eight persons at dinner together as the utmost that his nerves could bear

[12] I suppose he had considered that these were all the governess would require for her private perusal

[13] she considered that Isabella might otherwise perhaps be going to Clifton the next day

[16] Having considered me at leisure, he said, ‘What made you ill yesterday?’

Where the thing or person mostly appeared as the Subject of a Passive, the Agent doing the ‘considering’ was normally omitted. The favoured sense were ‘judge’ [17] or ‘show concern for’ [18-19]. Especially useful is the option of presenting one’s own viewpoint as valid [19-21], possibly after ‘all things’ have been considered’ [22-23].

[18] Mrs. Bates was considered with all the regard and respect which a harmless old lady can excite.

[19]‘Ah! my dear, where health is at stake, nothing else should be considered!’

Although many of these usages do not figure in the active English of my students, the latter have proven encouragingly astute in discovering meanings from the contexts. In addition, the usages and meanings they discover from their own explorations have been found to be more easily retained for active use in their own writing.

64. A central idea in the logic of successive approximations is that a piece of writing is always open for revisions and improvements. I shall close with a demonstration of results from a recent exercise. My course in ‘Stylistics’ was mainly organised to help students become more sensitive toward English style and improve their own. For most of them, the approach was all the more novel insofar as ‘style’ has not been a common topic or issue in the lower-division teaching of English as a non-native language. Here also, the obstacle is probably the difficulties of assigning number-grades derived by counting ‘right and wrong answers’ (cf. ¶ 51). Near the start of the course, student writing was uniformly of the quality shown back in samples [2] through [5]. But after many episodes of editing and revising samples, noticeable changes emerged in the writing of those who worked most actively, while the writing of those who remained inactive or just tinkered with the language samples hardly improved at all.

65. At the end, an assessment was administered consisting of a 400-word sample taken from the Hadith on the life of ‘A’isha bin Abu Bakr, the Prophet’s youngest wife. This topic was chosen because it presents no problems for understanding a story which my students know quite well in Arabic but which they had never read in English. Also, the epic historical film ‘al risalah’ (The Messenger) by Mustafa Akkad had been used for a series of essay presentations earlier during the term, and I had provided many of my own illustrations of the dignified style such a topic deserves.

66. As in some previous exercises, I had conspicuously transformed the sample into the mode of learner English documented near the start of this and similar courses. The instructions were ‘make a short essay by revising and improving the English style’; they had two hours and were encouraged to use dictionaries and thesauri, which would also be available in real-life situations (cf. ¶ 61). But they had to work by hand, since using the terminals here would discriminate against students who are still bashful around computers.

67. In the data shown below, samples marked with X are from my own text, whilst those marked with S are taken from student papers. In the famous episode of Abu Bakr reciting verses in his fever and ‘A’isah memorising them without understanding the words, [24.1] and [25.1] were edited into [24.2] and [25.2]:

[24.1-X] ‘A’ishah go visit, unhappy to finds he lied completely sick and tired;

[25.1-X] father answer her in poetry verses, not realise, just sound her like samesame noise!

[25.2-S] He answered her in poetry verses, which were like delirium to her ears.

[26.1-X] so we can says all this demonstration how so good she remembering, which across years could preserved so many of the Prophet priceless saying.

[26.2-S] So this was a demonstration of the great power of her memory, which, with the passage of time, preserved so many priceless sayings of the Prophet.

[26.3-S] Her attitude with her father was a good illustration to show her ability of retention and recall.

[26.4-S] ‘A’isha was a famous and trusted Hadith narrator, because of her strong memory and her excellent capacity of memorising the Prophet’s sayings.

Sample [26.3] and [26.4] had been astutely placed before the actual narration of the episode, thus explaining to readers why the latter merited recounting. The ending of the essay was also well edited by some students, as when [27.1] became [27.2].

[27.1-X] you can see how much setted example ‘A’ishah in help education and mostly ofall education Muslim women in the laws and teaching of Islam which needs to followed

[27.2-S] The example of ‘A’ishah in elevating education and particularly the education of Muslim women in the laws and teachings of Islam is one which needs to be followed.

68. But, as on most tests or assessments, the results were by no means uniformly successful. When the task is as novel as this one, success could only be achieved by those who had been actively working on it over a fair period of time; inactive classwork must lead to low success. The same passage like [28.1] was skilfully edited skilfully in [28.2-3] but also unskilfully in [28.4].

[28.1-X] ‘A’isha generous person and patient, put up with other of Prophet household poorness and hungering

[28.2-S] ‘A’isha was a generous and patient kind of person. She tolerated the hunger and poorness in the Prophet’s household.

[28.3-S] The generous and strong character of ‘A’isha suffered a lot of agony and starvation.

Similarly, [29.1] became [29.2] but also [29.3]; [30.1] became [30.2] but also [30.3].

[29.1-X] often go on long times when for days on end not fire lighted in the no much furnished house for the cooking

[29.2-S] For long periods of time, no fire was lit in the humbly furnished house for cooking food.

[29.3-S] she always goon long time with the fire of cooking didn’t enlighten their house

[30.2-S] ‘A’ishah’s life extended 50 years after the departure of the Prophet.

69. Finding the appropriate level to build up the skills for using English as a non-native language is a problematic but crucial condition for success. I will always need a deal of purely diagnostic information on my students prior skills; and, under the current methods in the lower schools, the levels across a whole university class are inevitably far from uniform. Pending some reliable means for tracking these learners into different classes, I will have contingents for whom a situated approach may appear either to low or too high. Still, the great majority undeniably need to refine their writing skills and to gain competence in revising.

70. From an administrative standpoint, the problem is quite different: too much uniformity rather than not enough. As I have noted at various points, situated learning through successive approximations is strongly oriented toward success and deliberately seeks to minimise the proportion of low successes and even more of ‘failures’. The results therefore cannot conform to the logics of ‘grading’, ‘grade curves’ or ‘grade inflation’ described in this paper (cf. ¶ 15, 45, 54,)

71. But I hope to have adduced in the paper some grounds to favour the view that educational justice as well as high standards would best be served by transforming an ambience which has tended to lower performance and impose ‘grade deflation’, and nowhere so strongly as in non-native language education. There at least, to assign ‘low marks’ and move on can defeat the whole enterprise and obstruct significant progress for many learners. So to be logical, we must make every effort to support and encourage cumulative learning as the true road to ‘high standards’

Notes

1 The far more common term ‘traditional’ will in fact not do here, having been used so widely and loosely for commodities like ‘antiques’ and ‘souvenirs’

2 Utterly implausible reasons are presented by a right-wing conspiracy theory: teachers are ‘radicals’ and ‘remnants of the 1960s and 1970s counterculture, militants and racists both black and white’, who’ teach students not to respect the values that their parents have taught them’ and who ‘have one intent only — undermining the American education system’ (Florida Review 10/12/1990).

3 For comparison, ‘considered’ occurred 383 times, ‘considering’ 159, ‘consideration’ 293 and ‘considerable’ 228.

References

Beaugrande, Robert de. 1997. New Foundations for a Science of Text and Discourse. Greenwich, CT: Ablex.

Freire, Paulo. 1970. Pedagogia do oprimido. Rio de Janeiro: Editora Paz e Terra.

Holes, Clive. 1995. Modern Arabic: Structures, Functions, and Varieties. London: Longman.

Lave, Jean and Etienne Wenger. 1990. Situated Learning: Legitimate Peripheral Participation. Cambridge: Cambridge UP.

Martin, Hans-Peter and Harald Schumann. 1996. Die Globaliserungsfalle: Der Angriff auf Demokratie und Wohlstand. Reinbeck bei Hamburg: Rowohlt.

Reich, Robert B. 1993. The Work of Nations: Preparing for 21st-Century Capitalism. NY: Simon and Schuster.

Examining the Logic of Educational Justice

A. Education in the ‘modern society’

B. Grading and testing

D. Situated learning and successive approximations

Notes

References