Annotated Bibliography Final Draft

University of Massachusetts Amherst ScholarWorks@UMass Amherst

English Department Faculty Publication Series English

1994

Ranking , Evaluating , Liking: Sorting Out Three Forms of Judgment. Peter Elbow University of Massachusetts - Amherst, elbow@english.umass.edu

Follow this and additional works at: https://scholarworks.umass.edu/eng_faculty_pubs

Part of the Higher Education and Teaching Commons

This Article is brought to you for free and open access by the English at ScholarWorks@UMass Amherst. It has been accepted for inclusion in English Department Faculty Publication Series by an authorized administrator of ScholarWorks@UMass Amherst. For more information, please contact scholarworks@library.umass.edu.

Recommended Citation Elbow, Peter, "Ranking, Evaluating, Liking: Sorting Out Three Forms of Judgment." (1994). College English. 12. Retrieved from https://scholarworks.umass.edu/eng_faculty_pubs/12

1

Peter Elbow

Ranking, Evaluating and Liking: Sorting Out Three Forms of Judgment

From: College English 55.2 (1993): 187-206. Reprinted in Everyone Can Write: Essays Toward a Hopeful Theory of

Writing and Teaching Writing. NY: Oxford University Press, 2000. This version lacks some formatting and copy editing

in the published versions.

This essay is my attempt to sort out the different acts we call assessment--the different

ways in which we express or frame our judgments of value. I have been working on this tangle

not just because it is interesting and important in itself but because assessment tends so much

to drive and control teaching. Much of what we do in the classroom is determined by the

assessment structures we work under.

Assessment is a large and technical area and I'm not a professional. But my main premise

or subtext in this essay is that we nonprofessionals can and should work on it because

professionals have not reached definitive conclusions about the problem of how to assess

writing (or anything else, I'd say). Also, decisions about assessment are often made by people

even less professional than we, namely legislators. Pat Belanoff and I realized that the field of

assessment is open when we saw the harmful effects of a writing proficiency exam at Stony

Brook and worked out a collaborative portfolio assessment system in its place (Belanoff and

Elbow; Elbow and Belanoff). Professionals keep changing their minds about large scale testing

and assessment. And as for classroom grading, psychometricians provide little support or

defense of it.

The Problems with Ranking and the Benefits of Evaluating

By ranking I mean the act of summing up one's judgment of a performance or person into

a single, holistic number or score. We rank every time we give a grade or holistic score.

Ranking implies a single scale or continuum or dimension along which all performances are

hung.

By evaluating I mean the act of expressing one's judgment of a performance or person by

pointing out the strengths and weaknesses of different features or dimensions. We evaluate

every time we write a comment on a paper or have an conversation about its value. Evaluation

implies the recognition of different criteria or dimensions--and by implication different contexts

and audiences for the same performance. Evaluation requires going beyond a first response

that may be nothing but a kind of ranking ("I like it" or "This is better than that"), and instead

looking carefully enough at the performance or person to make distinctions between parts or

features or criteria.

It's obvious, thus, that I am troubled by ranking. But I will resist any temptation to argue

that we can get rid of all ranking--or even should. Instead I will try to show how we can have

less ranking and more evaluation in its place.

2

I see three distinct problems with ranking: it is inaccurate or unreliable; it gives no

substantive feedback; and it is harmful to the atmosphere for teaching and learning.

(1) First the unreliability. To rank reliably means to give a fair number, to find the single

quantitative score that readers will agree on. But readers don't agree.

This is not news--this unavailability of agreement. We have long seen it on many fronts.

For example, research in evaluation has shown many times that if we give a paper to a set of

readers, those readers tend to give it the full range of grades (Diederich). I've recently come

across new research to this effect--new to me because it was published in 1912. The

investigators carefully showed how high school English teachers gave different grades to the

same paper. In response to criticism that this was a local problem in English, they went on the

next year to discover an even greater variation among grades given by high school geometry

teachers and history teachers to papers in their subjects. (See the summary of Daniel Starch

and Edward Elliott's 1913 School Review articles in Kirschenbaum, Simon and Napier 258-59.)

We know the same thing from literary criticism and theory. If the best critics can't agree

about what a text means, how can we be surprised that they disagree even more about the

quality or value of texts. And we know that nothing in literary or philosophical theory gives us

any agreed-upon rules for settling such disputes.

Students have shown us the same inconsistency with their own controlled experiments of

handing the same paper to different teachers and getting different grades. This helps explain

why we hate it so when students ask us their favorite question, "What do you want for an A?":

it rubs our noses in the unreliability of our grades.

Of course champions of holistic scoring argue that they get can get agreement among

readers--and they often do (White). But they get that agreement by "training" the readers

before and during the scoring sessions. What "training" means is getting those scorers to stop

reading the way they normally read--getting them to stop using the conflicting criteria and

standards they normally use outside the scoring sessions. (In an impressive and powerful book,

Barbara Herrnstein Smith argues that whenever we have widespread inter-reader reliability, we

have reason to suspect that difference has been suppressed and homogeneity imposed--almost

always at the expense of certain groups.) In short, the reliability in holistic scoring is not a

measure of how texts are valued by real readers in natural settings, but only of how they are

valued in artificial settings with imposed agreements.

Defenders of holistic scoring might reply (as one anonymous reviewer did), that holistic

scores are not perfect or absolutely objective readings but just "judgments that most readers

will agree are the appropriate ones given the purpose of the assessment and the system of

communication." But I have been in and even conducted enough holistic scoring sessions to

know that even that degree of agreement doesn't occur unless "purpose" and

"appropriateness" are defined to mean acceptance of the single set of standards imposed on

that session. We know too much about the differences among readers and the highly variable

nature of the reading process. Supposing we get readings only from academics, or only from

people in English, or only from respected critics, or only from respected writing programs, or

only from feminists, or only from sound readers of my tribe (white, male, middle class, full

3

professors between the ages of fifty and sixty). We still don't get agreement. We can

sometimes get agreement among readers from some subset, a particular community that has

developed a strong set of common values, perhaps one English department or one writing

program. But what is the value of such a rare agreement? It tells us nothing about how

readers from other English departments or writing programs will judge--much less how readers

from other domains will judge.

(From the opposite ideological direction, some skeptics might object to my skeptical train

of thought: "So what else is new?" they might reply. "Of course my grades are biased,

'interested' or 'situated'--always partial to my interests or the values of my community or

culture. There's no other possibility." But how can people consent to give grades if they feel

that way? A single teacher's grade for a student is liable to have substantial consequences--for

example on eligibility for a scholarship or a job or entrance into professional school. In grading,

surely we must not take anything less than genuine fairness as our goal.)

It won't be long before we see these issues argued in a court of law, when a student who

has been disqualified from playing on a team or rejected from a professional school sues,

charging that the basis for his plight--teacher grades--is not reliable. I wonder if lawyers will be

able to make our grades stick.

(2) Ranking or grading is woefully uncommunicative. Grades and holistic scores are

nothing but points on a continuum from "yea" to "boo"--with no information or clues about the

criteria behind these noises. They are 100 percent evaluation and 0 percent description or

information. They quantify the degree of approval or disapproval in readers but tell nothing at

all about what the readers actually approve or disapprove of. They say nothing that couldn't be

said with gold stars or black marks or smiley-faces. Of course our first reactions are often

nothing but global holistic feelings of approval or disapproval, but we need a system for

communicating our judgments that nudges us to move beyond these holistic feelings and to

articulate the basis of our feeling--a process that often leads us to change our feeling. (Holistic

scoring sessions sometimes use rubrics that explain the criteria--though these are rarely passed

along to students--and even in these situations, the rubrics fail to fit many papers.) As C. S.

Lewis says, "People are obviously far more anxious to express their approval and disapproval of

things than to describe them" (7).

(3) Ranking leads students to get so hung up on these oversimple quantitative verdicts

that they care more about scores than about learning--more about the grade we put on the

paper than about the comment we have written on it. Have you noticed how grading often

forces us to write comments to justify our grades?--and how these are often not the comment

we would make if we were just trying to help the student write better? ("Just try writing

several favorable comments on a paper and then giving it a grade of D" [Diederich 21].)

Grades and holistic scores give too much encouragement to those students who score

high--making them too apt to think they are already fine--and too little encouragement to those

students who do badly. Unsuccessful students often come to doubt their intelligence. But

oddly enough, many "A" students also end up doubting their true ability and feeling like frauds-

-because they have sold out on their own judgment and simply given teachers whatever yields

an A. They have too often been rewarded for what they don't really believe in. (Notice that

4

there's more cheating by students who get high grades than by those who get low ones. There

would be less incentive to cheat if there were no ranking.)

We might be tempted to put up with the inaccuracy or unfairness of grades if they gave

good diagnostic feedback or helped the learning climate; or we might put up with the damage

they do to the learning climate if they gave a fair or reliable measure of how skilled or

knowledgeable students are. But since they fail dismally on both counts, we are faced with the

striking question of why grading has persisted so long.

There must be many reasons. It is obviously easier and quicker to express a global feeling

with a single number than to figure out what the strengths and weaknesses are and what one's

criteria are. (Though I'm heartened to discover, as I pursue this issue, how troubled teachers

are by grading and how difficult they find it.) But perhaps more important, we see around us a

deep hunger to rank--to create pecking orders: to see who we can look down on and who we

must look up to, or in the military metaphor, who we can kick and who we must salute.

Psychologists tell us that this taste for pecking orders or ranking is associated with the

authoritarian personality. We see this hunger graphically in the case of IQ scores. It is plain

that IQ scoring does not represent a commitment to looking carefully at peoples' intelligence;

when we do that, we see different and frequently uncorrelated kinds or dimensions of

intelligence (Gardner). The persistent use of IQ scores represents the hunger to have a number

so that everyone can have a rank. ("Ten!" mutter the guys when they see a pretty woman.)

Because ranking or grading has caused so much discomfort to so many students and

teachers, I think we see a lot of confusion about the process. It is hard to think clearly about

something that has given so many of us such anxiety and distress. The most notable confusion I

notice is the tendency to think that if we renounce ranking or grading, we are renouncing the

very possibility of judgment and discrimination--that we are embracing the idea that there is no

way to distinguish or talk about the difference between what works well and what works badly.

So the most important point, then, is that I am not arguing against judgment or

evaluation. I'm just arguing against that crude, oversimple way of representing judgment--

distorting it, really--into a single number, which means ranking people and performances along

a single continuum.

In fact I am arguing for evaluation. Evaluation means looking hard and thoughtfully at a piece

of writing in order to make distinctions as to the quality of different features or dimensions.

For example, the process of evaluation permits us to make the following kinds of statements

about a piece of writing:

-The thinking and ideas seemed interesting and creative.

-The overall structure or sequence seemed confusing.

-The writing was perfectly clear at the level of individual sentences and even paragraphs.

-There is an odd, angry tone of voice that seems unrelated or inappropriate to what the writer

was saying.

-Yet this same voice is strong and memorable and makes one listen if irritated.

5

-There are a fair number of mistakes in grammar or spelling: more than "a sprinkling" but less

than "riddled with."

To rank, on the other hand, is to be forced to translate those discriminations into a single

number. What grade or holistic score do these judgments add up to? It's likely, by the way,

that more readers would agree with those separate, "analytic" statements than would agree on

a holistic score.

I've conducted many assessment sessions where we were not trying to impose a set of

standards but rather to find out how experienced teachers read and evaluate, and I've had

many opportunities to see that good readers give grades or scores right down through the

range of possibilities. Of course good readers sometimes agree--especially on papers that are

strikingly good or bad or conventional, but I think I see difference more frequently than

agreement when readers really speak up.

The process of evaluation, because it invites us to articulate our criteria and to make

distinctions among parts or features or dimensions of a performance, thereby invites us further

to acknowledge the main fact about evaluation: that different readers have different priorities,

values, and standards.

The conclusion I am drawing, then, in this first train of thought is that we should do less

ranking and more evaluation. Instead of using grades or holistic scores--single number verdicts

that try to sum up complex performances along only one scale--we should give some kind of

written or spoken evaluation that discriminates among criteria and dimensions of the writing--

and if possible that takes account of the complex context for writing: who the writer is, what

the writer's audience and goals are, who we are as reader and how we read, and how we might

differ in our reading from other readers the writer might be addressing.

But how can we put this principle into practice? The pressure for ranking seems

implacable. Evaluation takes more time, effort, and money. It seems as though we couldn't get

along without scores on writing exams. Most teachers are obliged to give grades at the end of

each course. And many students--given that they have become conditioned or even addicted

to ranking over the years and must continue to inhabit a ranking in most of their courses--will

object if we don't put grades on papers. Some students, in the absence of that crude gold star

or black mark, may not try hard enough (though how hard is "enough"--and is it really our job

to stimulate motivation artificially with grades--and is grading the best source of motivation?).

It is important to note that there are certain schools and colleges that do not use single-

number grades or scores, and they function successfully. I taught for nine years at Evergreen

State College, which uses only written evaluations. This system works fine, even down to

getting students accepted into high quality graduate and professional schools.

Nevertheless we have an intractable dilemma: that grading is unfair and

counterproductive but that students and institutions tend to want grades. In the face of this

dilemma there is a need for creativity and pragmatism. Here are some ways in which I and

others use less ranking and more evaluation in teaching--and they suggest some adjustments in

how we score larger scale assessments. What follows is an assortment of experimental

compromises--sometimes crude, seldom ideal or utopian--but they help.

6

(a) Portfolios. Just because conventional institutions oblige us to turn in a single

quantitative course grade at the end of every marking period, it doesn't follow that we need to

grade individual papers. Course grades are more trustworthy and less damaging because they

are based on so many performances over so many weeks. By avoiding frequent ranking or

grading, we make it somewhat less likely for students to become addicted to oversimple

numerical rankings--to think that evaluation always translates into a simple number--in short,

to mistake ranking for evaluation. (I'm not trying to defend conventional course grades since

they are still uncommunicative and they still feed the hunger for ranking.) Portfolios permit me

to refrain from grading individual papers and limit myself to writerly evaluative comments--and

help students see this as a positive rather than a negative thing, a chance to be graded on a

body of their best work that can be judged more fairly. Portfolios have many other advantages

as well. They are particularly valuable as occasions for asking students to write extensive and

thoughtful explorations of their own strengths and weaknesses.

A midsemester portfolio is usually an informal affair, but it is a good occasion for giving

anxious students a ballpark estimate of how well they are doing in the course so far. I find it

helpful to tell students that I'm perfectly willing to tell them my best estimate of their course

grade--but only if they come to me in conference and only during the second half of the

semester. This serves somewhat to quiet their anxiety while they go through seven weeks of

drying out from grades. By midsemester, most of them have come to enjoy not getting those

numbers and thus being able to think better about more writerly comments from me and their

classmates.

Portfolios are now used extensively and productively in larger assessments, and there is

constant experimentation with new applications (Belanoff and Dickson; Portfolio Assessment

Newsletter; Portfolio News).

(b) Another useful option is to make a strategic retreat from a wholly negative position.

That is, I sometimes do a bit of ranking even on individual papers, using two "bottom-line"

grades: H and U for "Honors" and "Unsatisfactory." I tell students that these translate to about

A or A- and D or F. This practice may seem theoretically inconsistent with all the arguments I've

just made, but (at the moment, anyway) I justify it for the following reasons.

First, I sympathize with a part of the students' anxiety about not getting grades: their fear

that they might be failing and not know about it--or doing an excellent job and not get any

recognition. Second, I'm not giving many grades; only a small proportion of papers get these

'H's or 'U's. The system creates a "non-bottom-line" or "non-quantified" atmosphere. Third,

these holistic judgments about best and worst do not seem as arbitrary and questionable as

most grades. There is usually a bit more agreement among readers about the best and worst

papers. What seems most dubious is the process of trying to rank that whole middle-range of

papers--papers that have a mixture of better and worse qualities so that the numerical grade

depends enormously on a reader's priorities or mood or temperament. My willingness to give

these few grades goes a long way toward helping my students forgo most bottom-line grading.

I'm not trying to pretend that these minimal "grades" are truly reliable. But they

represent a very small amount of ranking. Yes, someone could insist that I'm really ranking

every single paper (and indeed if it seemed politically necessary, I could put an OK or S [for

7

satisfactory] on all those middle range papers and brag, "Yes, I grade everything.") But the fact

is that I am doing much less sorting since I don't have to sort them into five or even twelve

piles. Thus there is a huge reduction in the total amount of unreliability I produce.

(It might seem that if I use only these few minimal grades I have no good way for figuring

out a final grade for the course--since that requires a more fine-grained set of ranks. But I don't

find that to be the case. For I also give these same minimal grades to the many other important

parts of my course such as attendance, meeting deadlines, peer responding, and journal

writing. If I want a mathematically computed grade on a scale of six or A through E, I can easily

compute it when I have such a large number of grades to work from--even though they are only

along an odd three point scale.)

This same practice of crude or minimal ranking is big help on larger assessments outside

classrooms, and needs to be applied to the process of assessment in general. There are two

important principles to emphasize. On the one hand we must be prudent or accommodating

enough to admit that despite all the arguments against ranking, there are situations when we

need that bottom-line verdict along one scale: which student has not done satisfactory work

and should be denied credit for the course? which student gets the scholarship? which

candidate to hire or fire? We often operate with scarce resources. But on the other hand we

must be bold enough to insist that we do far more ranking than is really needed. We can get

along not only with fewer occasions for assessment but also with fewer gradations in scoring. If

we decide what the real bottom-line is on a given occasion--perhaps just "failing" or perhaps

"honors" too--then the reading of papers or portfolios is enormously quick and cheap. It leaves

time and money for evaluation--perhaps for analytic scoring or some comment.

At Stony Brook we worked out a portfolio system where multiple readers had only to

make a binary decision: acceptable or not. Then individual teachers could decide the actual

course grade and give comments for their own students--so long as those students passed in

the eyes of an independent rater (Elbow and Belanoff; Belanoff and Elbow). The best way to

begin to wean our society from its addiction to ranking may be to permit a tiny bit of it (which

also means less unreliability)--rather than trying to go "cold turkey."

(c) Sometimes I use an analytic grid for evaluating and commenting on student papers. Here's

an example:

Strong OK Weak

CONTENT, INSIGHTS, THINKING, GRAPPLING WITH TOPIC

GENUINE REVISION, SUBSTANTIVE CHANGES, NOT JUST EDITING

ORGANIZATION, STRUCTURE, GUIDING THE READER

LANGUAGE: SYNTAX, SENTENCES, WORDING, VOICE

MECHANICS: SPELLING, GRAMMAR, PUNCTUATION, ROOFREADING

OVERALL [Note: this is not a sum of the other scores.]

8

I often vary the criteria in my grid (e.g. "connecting with readers" or "investment") depending

on the assignment or the point in the semester.

Grids are a way I can satisfy the students' hunger for ranking but still not give in to

conventional grades on individual papers. Sometimes I provide nothing but a grid (especially on

final drafts), and this is a very quick way to provide a response. Or on midprocess drafts I

sometimes use a grid in addition to a comment: a more readerly comment that often doesn't

so much tell them what's wrong or right or how to improve things but rather tries to give them

an account of what is happening to me as I read their words. I think this kind of comment is

really the most useful thing of all for students, but it frustrates some students for a while. The

grid can help these students feel less anxious and thus pay better attention to my comment.

I find grids extremely helpful at the end of the semester for telling students their

strengths and weaknesses in the course--or what they've done well and not so well. Besides

categories like the ones above, I use categories like these: "skill in giving feedback to others,"

"ability to meet deadlines," "effort," and "improvement." This practice makes my final grade

much more communicative.

(d) I also help make up for the absence of ranking--gold stars and black marks--by having

students share their writing with each other a great deal both orally and through frequent

publication in class magazines. Also, where possible, I try to get students to give or send writing

to audiences outside the class. At the University of Massachusetts, freshmen pay a ten dollar

lab fee for the writing course, and every teacher publishes four or five class magazines of final

drafts a semester. The effects are striking. Sharing, peer feedback, and publication give the

best reward and motivation for writing, namely, getting your words out to many readers.

(e) I sometimes use a kind of modified contract grading. That is, at the start of the course

I pass out a long list of all the things that I most want students to do--the concrete activities

that I think most lead to learning--and I promise students that if they do them all they are

guaranteed a certain final grade. Currently, I say it's a B--it could be lower or higher. My list

includes these items: not missing more than a week's worth of classes; not having more than

one late major assignment; substantive revising on all major revisions; good copy editing on all

major revisions; good effort on peer feedback work; keeping up the journal; and substantial

effort and investment on each draft.

I like the way this system changes the "bottom-line" for a course: the intersection where

my authority crosses their self-interest. I can tell them, "You have to work very hard in this

course, but you can stop worrying about grades." The crux is no longer that commodity I've

always hated and never trusted: a numerical ranking of the quality of their writing along a

single continuum. Instead the crux becomes what I care about most: the concrete behaviors

that I most want students to engage in because they produce the more learning and help me

teach better. Admittedly, effort and investment are not concrete observable behaviors, but

they are no harder to judge than overall quality of writing. And since I care about effort and

investment, I don't mind the few arguments I get into about them; they seem fruitful. ("Let's

try and figure out why it looked to me as though you didn't put any effort in here.") In contrast,

I hate discussions about grades on a paper and find such arguments fruitless. Besides, I'm not

9

making fine distinctions about effort and investment--just letting a bell go off when they fall

palpably low.

It's crucial to note that I am not fighting evaluation with this system. I am just fighting

ranking or grading. I still write evaluative comments and often use an evaluative grid to tell my

students what I see as strengths and weaknesses in their papers. My goal is not to get rid of

evaluation but in fact to emphasize it, enhance it. I'm trying to get students to listen better to

my evaluations--by uncoupling them from a grade. In effect, I'm doing this because I'm so fed

up with students following or obeying my evaluations too blindly--making whatever changes my

comments suggest but doing it for the sake of a grade; not really taking the time to make up

their own minds about whether they think my judgments or suggestions really make sense to

them. The worst part of grades is that they make students obey us without carefully thinking

about the merits of what we say. I love the situation this system so often puts students in: I

make a criticism or suggestion about their paper, but it doesn't matter to their grade whether

they go along with me or not (so long as they genuinely revise in some fashion). They have to

think; to decide.

Admittedly this system is crude and impure. Some of the really skilled students who are

used to getting A's and desperate to get one in this course remain unhelpfully hung up about

getting those 'H's on their papers. But a good number of these students discover that they

can't get them, and they soon settle down to accepting a B and having less anxiety and more of

a learning voyage.

The Limitations of Evaluation and the Benefits of Evaluation-free Zones

Everything I've said so far has been in praise of evaluation as a substitute for ranking. But

I need to turn a corner here and speak about the limits or problems of evaluation. Evaluating

may be better than ranking, but it still carries some of the same problems. That is, even though

I've praised evaluation for inviting us to acknowledge that readers and contexts are different,

nevertheless the very word evaluation tends to imply fairness or reliability or getting beyond

personal or subjective preferences. Also, of course, evaluation takes a lot more time and work.

To rank you just have to put down a number; holistic scoring of exams is cheaper than analytic

scoring.

Most important of all, evaluation harms the climate for learning and teaching--or rather

too much evaluation has this effect. That is, if we evaluate everything students write, they tend

to remain tangled up in the assumption that their whole job in school is to give teachers "what

they want." Constant evaluation makes students worry more about psyching out the teacher

than about what they are really learning. Students fall into to a kind of defensive or on-guard

stance toward the teacher: a desire to hide what they don't understand and try to impress.

This stance gets in the way of learning. (Think of the patient trying to hide symptoms from the

doctor.) Most of all, constant evaluation by someone in authority makes students reluctant to

take the risks that are needed for good learning--to try out hunches and trust their own

judgment. Face it: if our goal is to get students to exercise their own judgment, that means

10

exercising an immature and undeveloped judgment and making choices that are obviously

wrong to us.

We see around us a widespread hunger to be evaluated that is often just as strong as the

hunger to rank. Countless conditions make many of us walk around in the world wanting to ask

others (especially those in authority), "How am I doing, did I do OK?" I don't think the hunger

to be evaluated is as harmful as the hunger to rank, but it can get in the way of learning. For I

find that the greatest and most powerful breakthroughs in learning occur when I can get myself

and others to put aside this nagging, self-doubting question ("How am I doing? How am I

doing?")--and instead to take some chances, trust our instincts or hungers. When everything is

evaluated, everything counts. Often the most powerful arena for deep learning is a kind of

"time out" zone from the pressures of normal evaluated reality: make-believe, play, dreams--in

effect, the Shakespearian forest.

In my attempts to get away from too much evaluation (not from all evaluation, just from

too much of it), I have drifted into a set of teaching practices which now feel to me like the best

part of my teaching. I realize now what I've been unconsciously doing for a number of years:

creating "evaluation-free zones."

(a) The paradigm evaluation-free zone is the ten minute, nonstop freewrite. When I get

students to freewrite, I am using my authority to create unusual conditions in order to

contradict or interrupt our pervasive habit of always evaluating our writing. What is essential

here are the two central features of freewriting: that it be private (thus I don't collect it or have

students share it with anyone else); and that it be nonstop (thus there isn't time for planning,

and control is usually diminished). Students quickly catch on and enter into the spirit. At the

end of the course, they often tell me that freewriting is the most useful thing I've taught them

(see Belanoff, Elbow, and Fontaine).

(b) A larger evaluation-free zone is the single unevaluated assignment--what people

sometimes call the "quickwrite" or sketch. This is a piece of writing that I ask students to do--

either in class or for homework--without any or much revising. It is meant to be low stakes

writing. There is a bit of pressure, nevertheless, since I usually ask them to share it with others

and I usually collect it and read it. But I don't write any comments at all--except perhaps to put

straight lines along some passages I like or to write a phrase of appreciation at the end. And I

ask students to refrain from giving evaluative feedback to each other--and instead just to say

"thank you" or mention a couple of phrases or ideas that stick in mind. (However, this writing-

without-feedback can be a good occasion for students to discuss the topic they have written

about--and thus serve as an excellent kick-off for discussions of what I am teaching.)

(c) These experiments have led me to my next and largest evaluation-free zone--what I

sometimes call a "jump start" for my whole course. For the last few semesters I've been

devoting the first three weeks entirely to the two evaluation-free activities I've just described:

freewriting (and also more leisurely private writing in a journal) and quickwrites or sketches.

Since the stakes are low and I'm not asking for much revising, I ask for much more writing

homework per week than usual. And every day we write in class: various exercises or games.

The emphasis is on getting rolling, getting fluent, taking risks. And every day all students read

11

out loud something they've written--sometimes a short passage even to the whole class. So

despite the absence of feedback, it is a very audience-filled and sociable three weeks.

At first I only dared do this for two weeks, but when I discovered how fast the writing

improves, how good it is for building community, and what a pleasure this period is for me, I

went to three weeks. I'm curious to try an experiment with teaching a whole course this way. I

wonder, that is, whether all that evaluation we work so hard to give really does any more good

than the constant writing and sharing (Zak).

I need to pause here to address an obvious rejoinder: "But withholding evaluation is not

normal!" Indeed, it is not normal--certainly not normal in school. We normally tend to

emphasize evaluations--even bottom-line ranking kinds of evaluations. But I resist the

argument that if it's not normal we shouldn't do it.

The best argument for evaluation-free zones is from experience. If you try them, I

suspect you'll discover that they are satisfying and bring out good writing. Students have a

better time writing these unevaluated pieces; they enjoy hearing and appreciating these pieces

when they don't have to evaluate. And I have a much better time when I engage in this

astonishing activity: reading student work when I don't have to evaluate and respond. And yet

the writing improves. I see students investing and risking more, writing more fluently, and

using livelier, more interesting voices. This writing gives me and them a higher standard of

clarity and voice for when we move on to more careful and revised writing tasks--tasks that

involve more intellectual pushing and that sometimes make their writing go tangled or sodden.

The Benefits and Feasibility of Liking

Liking and disliking seem like unpromising topics in an exploration of assessment. They

seem to represent the worst kind of subjectivity, the merest accident of personal taste. But I've

recently come to think that the phenomenon of liking is perhaps the most important evaluative

response for writers and teachers to think about. In effect, I'm turning another corner in my

argument. In the first section I argued against ranking--with evaluating being the solution.

Next I argued not against evaluating--but for no-evaluation zones in addition to evaluating.

Now I will argue neither against evaluating nor against non-evaluation zones, but for something

very different in addition, or perhaps underneath, as a foundation: liking.

Let me start with the germ story. I was in a workshop and we were going around the

circle with everyone telling a piece of good news about their writing in the last six months. It

got to Wendy Bishop, a good poet (who has also written two good books about the teaching of

writing), and she said, "In the last six months, I've learned to like everything I write." Our jaws

dropped; we were startled--in a way scandalized. But I've been chewing on her words ever

since, and they have led me into a retelling of the story of how people learn to write better.

The old story goes like this: We write something. We read it over and we say, "This is

terrible. I hate it. I've got to work on it and improve it." And we do, and it gets better, and this

happens again and again, and before long we have become a wonderful writer. But that's not

12

really what happens. Yes, we vow to work on it--but we don't. And next time we have the

impulse to write, we're just a bit less likely to start.

What really happens when people learn to write better is more like this: We write

something. We read it over and we say, "This terrible . . . . But I like it. Damn it, I'm going to get

it good enough so that others will like it too." And this time we don't just put it in a drawer, we

actually work hard on it. And we try it out on other people too--not just to get feedback and

advice but, perhaps more important, to find someone else who will like it.

Notice the two stories here--two hypotheses. (a) "First you improve the faults and then

you like it." (b) "First you like it and then you improve faults." The second story may sound odd

when stated so baldly, but really it's common sense. Only if we like something will we get

involved enough to work and struggle with it. Only if we like what we write will we write again

and again by choice--which is the only way we get better.

This hypothesis sheds light on the process of how people get to be published writers.

Conventional wisdom assumes a Darwinian model: poor writers are unread; then they get

better; as a result, they get a wider audience; finally they turn into Norman Mailer. But now I'd

say the process is more complicated. People who get better and get published really tend to be

driven by how much they care about their writing. Yes, they have a small audience at first--

after all, they're not very good. But they try reader after reader until finally they can find

people who like and appreciate their writing. I certainly did this. If someone doesn't like her

writing enough to be pushy and hungry about finding a few people who also like it, she

probably won't get better.

It may sound so far as though all the effort and drive comes from the lonely driven writer-

-and sometimes it does (Norman Mailer is no joke). But, often enough, readers play the

crucially active role in this story of how writers get better. That is, the way writers learn to like

their writing is by the grace of having a reader or two who likes it--even though it's not good.

Having at least a few appreciative readers is probably indispensable to getting better.

When I apply this story to our situation as teachers I come up with this interesting

hypothesis: good writing teachers like student writing (and like students). I think I see this

born out--and it is really nothing but common sense. Teachers who hate student writing and

hate students are grouchy all the time. How could we stand our work and do a decent job if we

hated their writing. Good teachers see what is only potentially good, they get a kick out of

mere possibility--and they encourage it. When I manage to do this, I teach well.

Thus, I've begun to notice a turning point in my courses--two or three weeks into the

semester: "Am I going to like these folks or is this going to be a battle, a struggle?" When I like

them everything seems to go better--and it seems to me they learn more by the end. When I

don't and we stay tangled up in struggle, we all suffer--and they seem to learn less.

So what am I saying? That we should like bad writing? How can we see all the

weaknesses and criticize student writing if we just like it? But here's the interesting point: if I

like someone's writing it's easier to criticize it.

13

I first noticed this when I was trying to gather essays for the book on freewriting that Pat

Belanoff and Sheryl Fontaine and I edited. I would read an essay someone had written, I would

want it for the book, but I had some serious criticism. I'd get excited and write, "I really like

this, and I hope we can use it in our book, but you've got to get rid of this and change that, and I

got really mad at this other thing." I usually find it hard to criticize, but I began to notice that I

was a much more critical and pushy reader when I liked something. It's even fun to criticize in

those conditions.

It's the same with student writing. If I like a piece, I don't have to pussyfoot around with

my criticism. It's when I don't like their writing that I find myself tiptoeing: trying to soften my

criticism, trying to find something nice to say--and usually sounding fake, often unclear. I see

the same thing with my own writing. If I like it, I can criticize it better. I have faith that there'll

still be something good left, even if I train my full critical guns on it.

In short--and to highlight how this section relates to the other two sections of this essay--

liking is not same as ranking or evaluating. Naturally, people get them mixed up: when they

like something, they assume it's good; when they hate it, they assume it's bad. But it's helpful

to uncouple the two domains and realize that it makes perfectly good sense to say, "This is

terrible, but I like it." Or, "This is good, but I hate it." In short, I am not arguing here against

criticizing or evaluating. I'm merely arguing for liking.

Let me sum up my clump of hypotheses so far:

-It's not improvement that leads to liking, but rather liking that leads to improvement.

-It's the mark of good writers to like their writing.

-Liking is not same as evaluating. We can often criticize something better when we like it.

-We learn to like our writing when we have a respected reader who likes it.

-Therefore, it's the mark of good teachers to like students and their writing.

If this set of hypotheses is true, what practical consequences follow from it? How can we

be better at liking? It feels as though we have no choice--as though liking and not-liking just

happen to us. I don't really understand this business. I'd love to hear discussion about the

mystery of liking--the phenomenology of liking. I sense it's some kind of putting oneself out--or

holding oneself open--but I can't see it clearly. I have a hunch, however, that we're not so

helpless about liking as we tend to feel.

For in fact I can suggest some practical concrete activities that I have found fairly reliable

at increasing the chances of liking student writing:

(a) I ask for lots of private writing and merely shared writing, that is, writing that I don't

read at all, and writing that I read but don't comment on. This makes me more cheerful

because it's so much easier. Students get better without me. Having to evaluate writing--

especially bad writing--makes me more likely to hate it. This throws light on grading: it's hard

to like something if we know we have to give it a D.

(b) I have students share lots of writing with each other--and after a while respond to

each other. It's easier to like their writing when I don't feel myself as the only reader and judge.

14

And so it helps to build community in general: it takes pressure off me. Thus I try to use peer

groups not only for feedback, but for other activities too, such as collaborative writing,

brainstorming, putting class magazines together, and working out other decisions.

(c) I increase the chances of my liking their writing when I get better at finding what is

good--or potentially good--and learn to praise it. This is a skill. It requires a good eye, a good

nose. We tend--especially in the academic world--to assume that a good eye or fine

discrimination means criticizing. Academics are sometimes proud of their tendency to be

bothered by what is bad. Thus I find I am sometimes looked down on as dumb and

undiscriminating: "He likes bad writing. He must have no taste, no discrimination." But I've

finally become angry rather than defensive. It's an act of discrimination to see what's good in

bad writing. Maybe, in fact, this is the secret of the mystery of liking: to be able to see

potential goodness underneath badness.

Put it this way. We tend to stereotype liking as a "soft" and sentimental activity. Mr.

Rogers is our model. Fine. There's nothing wrong with softness and sentiment--and I love Mr.

Rogers. But liking can also be hard-assed. Let me suggest an alternative to Mr. Rogers: B. F.

Skinner. Skinner taught pigeons to play ping-pong. How did he do it? Not by moaning, "Pigeon

standards are falling. The pigeons they send us these days are no good. When I was a pigeon . .

. ." He did it by a careful, disciplined method that involved close analytic observation. He put

pigeons on a ping-pong table with a ball, and every time a pigeon turned his head 30 degrees

toward the ball, he gave a reward (see my "Danger of Softness").

_What would this approach require in the teaching of writing? It's very simple . . . but not

easy. Imagine that we want to teach students an ability they badly lack, for example how to

organize their writing or how to make their sentences clearer. Skinner's insight is that we get

nowhere in this task by just telling them how much they lack this skill: "It's disorganized.

Organize it!" "It's unclear. Make it clear!" Notice how much more practical and helpful it is to

move from this kind of advice: "Do something different from what you're doing here" to this

kind: "Do more of what you're doing there."

No, what we must learn to do is to read closely and carefully enough to show the student

little bits of proto-organization or sort of clarity in what they've already written. We don't have

to pretend the writing is wonderful. We could even say, "This is a terrible paper and the worst

part about it is the lack of organization. But I will teach you how to organize. Look here at this

little organizational move you made in this sentence. Read it outloud and try to feel how it

pulls together this stuff here and distinguishes it from that stuff there. Try to remember what it

felt like writing that sentence--creating that piece of organization. Do it some more."

When academics criticize behaviorism as crude it often means that they aren't willing to

do the close careful reading of student writing that is required. They'd rather give a cursory

reading and turn up their nose and give a low grade and complain about falling standards. No

one has undermined behaviorism's main principle of learning: that reward produces learning

more effectively than punishment.

(d) I improve my chances of liking student writing when I take steps to get to know them

a bit as people. I do this partly through the assignments I give. That is, I always ask them to

15

write a letter or two to me and to each other (for example about their history with writing). I

base at least a couple of assignments on their own experiences, memories, or histories. And I

make sure some of the assignments are free choice pieces--which also helps me know them.

In addition, I make sure to have at least three conferences with each _student each

semester--the first one very early. I often call off some class es in order to keep conferences

from being too onerous (insisting nevertheless that students meet with their partner or small

group when class is called off). Some teachers have mini-conferences with students during

class--while students are engaged in writing or peer group meetings. I've found that when I

deal only with my classes as a whole--as a large group--I sometimes experience them as a herd

or lump--as stereotyped "adolescents"; I fail to experience them as individuals. For me,

personally, this is disastrous since it often leads me to experience them as that scary tribe that I

felt rejected by when I was an eighteen-year-old--and thus, at times, as "the enemy." But when

I sit down with them face to face, they are not so stereotyped or alien or threatening--they are

just eighteen-year-olds.

Getting a glimpse of them as individual people is particularly helpful in cases where their

writing is not just bad, but somehow offensive--perhaps violent or cruelly racist or homophobic

or sexist--or frighteningly vacuous. When I know them just a bit I can often see behind their

awful attitude to the person and the life situation that spawned it, and not hate their writing so

much. When I know students I can see that they are smart behind that dumb behavior; they

are doing the best they can behind that bad behavior. Conditions are keeping them from acting

decently; something is holding them back.

(e) It's odd, but the more I let myself show the easier it is to like them and their writing. I

need to share some of my own writing--show some of my own feelings. I need to write the

letter to them that they write to me--about my past experiences and what I want and don't

want to happen.

(f) It helps to work on my own writing--and work on learning to like it. Teachers who are

most critical and sour about student writing are often having trouble with their own writing.

They are bitter or unforgiving or hurting toward their own work. (I think I've noticed that failed

PhDs are often the most severe and difficult with students.) When we are stuck or sour in our

own writing, what helps us most is to find spaces free from evaluation such as those provided

by freewriting and journal writing. Also, activities like reading out loud and finding a supportive

reader or two. I would insist, then, that if only for the sake of our teaching, we need to learn to

be charitable and to like our own writing.

A final word. I fear that this sermon about liking might seem an invitation to guilt. There

is enough pressure on us as teachers that we don't need someone coming along and calling us

inadequate if we don't like our students and their writing. That is, even though I think I am

right to make this foray into the realm of feeling, I also acknowledge that it is dangerous--and

paradoxical. It strikes me that we also need to have permission to hate the dirty bastards and

their stupid writing.

After all, the conditions under which they go to school bring out some awful behavior on

their part, and the conditions under which we teach sometimes make it difficult for us to like

16

them and their writing. Writing wasn't meant to be read in stacks of twenty-five, fifty, or

seventy-five. And we are handicapped as teachers when students are in our classes against

their will. (Thus high school teachers have the worst problem here, since their students tend to

be the most sour and resentful about school.)

Indeed, one of the best aids to liking students and their writing is to be somewhat

charitable toward ourselves about the opposite feelings that we inevitably have. I used to think

it was terrible for teachers to tell those sarcastic stories and hostile jokes about their students:

"teacher room talk." But now I've come to think that people who spend their lives teaching

need an arena to let off this unhappy steam. And certainly it's better to vent this sarcasm and

hostility with our buddies than on the students themselves. The question, then, becomes this:

do we help this behavior function as a venting so that we can move past it and not be trapped

in our inevitable resentment of students? Or do we tell these stories and jokes as a way of

staying stuck in the hurt, hostile, or bitter feelings--year after year--as so many sad teachers do?

In short I'm not trying to invite guilt, I'm trying to invite hope. I'm trying to suggest that if

we do a sophisticated analysis of the difference between liking and evaluating, we will see that

it's possible (if not always easy) to like students and their writing--without having to give up our

intelligence, sophistication, or judgment.

* * *

Let me sum up the points I'm trying to make about ranking, evaluating, and liking:

-Let's do as little ranking and grading as we can. They are never fair and they undermine

learning and teaching.

-Let's use evaluation instead--a more careful, more discriminating, fairer mode of

assessment.

-But because evaluating is harder than ranking, and because too much evaluating also

undermines learning, let's establish small but important evaluation-free zones.

-And underneath it all--suffusing the whole evaluative enterprise--let's learn to be better

likers: liking our own and our students' writing, and realizing that liking need not get in

the way of clear-eyed evaluation.

WORKS CITED

Diederich, Paul. Measuring Growth in English. Urbana: NCTE, 1974.

Belanoff, Pat and Peter Elbow. "Using Portfolios to Increase Collaboration and Community in a

Writing Program." WPA: Journal of Writing Program Administration 9.3 (Spring 1986):

27-40. (Also in Portfolios: Process and Product. Ed. Pat Belanoff and Marcia Dickson.

Portsmouth NH: Boynton/Cook Heinemann. 1991.)

17

Belanoff, Pat, Peter Elbow and Sheryl Fontaine eds. Nothing Begins with N: New Investigations

of Freewriting. Southern Illinois UP, 1991.

Bishop, Wendy. Something Old, Something New: College Writing Teachers and Classroom

Change. Carbondale: Southern Illinois UP, 1990.

---. Released into Language: Options for Teaching Creative Writing. Urbana: NCTE, 1990.

Elbow, Peter. "The Danger of Softness." What is English. New York: MLA, 1990: 197-210.

Elbow, Peter and Pat Belanoff. "State University of New York: Portfolio-Based Evaluation

Program." New Methods in College Writing Programs: Theory into Practice. Eds. Paul

Connolly and Teresa Vilardi. New York: MLA, 1986: 95-105. (Also in Portfolios: Process

and Product. Ed. Pat Belanoff and Marcia Dickson. Portsmouth NH: Boynton/Cook

Heinemann. 1991.)

Gardner, Howard. Frames of Mind: The Theory of Multiple Intelligences. New York: Basic,

1983.

Kirschenbaum, Howard, Simon Sidney, and Rodney Napier. Wad-Ja-Get? The Grading Game in

American Education. New York: Hart Publishing, 1971.

Lewis, C. S. Studies in Words. London: Cambridge UP, 2nd ed, 1967.

Portfolio Assessment Newsletter. Five Centerpointe Drive, Suite 100, Lake Oswego, Oregon

97035.

Portfolio News. c/o San Dieguito Union High School District, 710 Encinitas Boulevard, Encinitas,

CA 92024.

Smith, Barbara Herrnstein. Contingencies of Value: Alternative Perspectives for Critical Theory.

Cambridge: Harvard UP, 1988.

White, Edward M. Teaching and Assessing Writing. San Francisco: Jossey-Bass, 1985.

Zak, Frances. "Exclusively Positive Responses to Student Writing." Journal of Basic Writing 9.2

(1990): 40-53.