Socrates in Plato's Apology warned the Athenians that his death at their hands would be a greater loss to them than to him.  This is a dramatic rendering of our need for iconoclasts, "gadflies" in the translations of Plato, who will challenge the ideas and practices that are at the center of our common sense.  Mill makes the same point in On Liberty when he praises even erroneous conceptions that, when tossed into the market place of ideas, only sharpen our thinking and nudge us closer to the truth.

So we should be grateful to Alfie Kohn for taking on this burden and carrying it out with intelligence, passion and charm.  His well-argued Punishment by Rewards, was my first introduction to the research on the downsides of the heavy-handed classroom applications of Skinner.  Unconditional Parenting makes similar points about child rearing.  The book under review, The Homework Myth, takes on homework, that staple of education whose necessity and benefits are largely taken for granted.

I will say at the beginning that this book is a disappointment on at least two levels.  I think there actually is a case to be made against homework as it is currently practiced.  This case would be in two parts.  First, a serious argument that children have an important right to their own childhood.  Such a right would imply a substantial burden of proof upon anyone proposing to limit natural childhood activities.  Homework, with the energy and time it subtracts from children's lives, is such a limitation.  Second, the burden of proof for homework would require a sensitive discussion of what sorts of out-of-school assignments are realistically necessary for children to progress at desired rates.  For example, very few would disapprove of requiring fifth graders to read five or six of one-hundred books from a teacher-composed list.  The present volume does neither of these two things.  In addition the methodology of Kohn's critique of the practice of requiring homework leaves much to be desired.  It is this latter that I will address here.

We could expect that a critique of homework would begin with a clear statement of what the author takes homework to be.  ". . . additional assignments to be completed at home."(3) suits the purposes of this just fine.  If the thesis is that children should not be required to produce any schoolwork or perform any school related activities during the time they are not in school, it would be interesting and clear (though I would think false).  The problem is that throughout the book Kohn supplies numerous other characterizations: first graders "…filling out worksheets at home." (6), "…busywork…" (20), "…reviewing batches of facts about some topic …" (31), "…cramming…"(32), "… drilled on the material …" (39), "… drilling …" (45), "… its lack of intrinsic interest, its inappropriate difficulty level, its having been assigned without any input from the students …" (58), "… repetitive efforts to learn specific things …" (105), "Mindless mimicry mathematics …"(109), "… drill ourselves on a certain skill …"(112), "… slogging through worksheets…" (115), "… drudgery…" (116), "… busywork…" (155), well, you get the point.  We don't know if Kohn has set up a straw man because there are no statistics on what is actually required as homework. 

If Kohn's thesis is that poorly constructed assignments are bad educational policy, that's true but not very interesting.  There is nowhere in the book any analysis of what types might be more or less likely to benefit learning.  As a consequence I am not sure what Kohn's position is.  That there should be no required extra-school assignments?  That there should be no tedious, rote extra-school assignments?  That extra-school assignments should be limited to creative activities or to activities that the students choose?  That levels and types of homework should vary by the individual characteristics of students.  The author says in one place, "… I've argued that homework (at least the way it's usually assigned and completed) does little to help kids acquire … [responsibility]." (61)  The parentheses render this ambiguous and in any case the claim is limited to teaching responsibility.  Elsewhere we find, "… there's generally very little, if any, benefit to assigning homework." (185)  And the teachers Kohn admires seem all to resemble eighth grade teacher Jim DeLuca who writes, "I rarely give my students any kind of homework.  I do not believe in homework, especially in language arts class." (114)     

Kohn does a good job arguing that the research purporting to show the benefits of homework does not succeed.  But it's important to note that such a critique is a rather simple task for the reason that this research problem is so complicated.  What would be required to show that homework increases learning?  Employing experimental controls we would need (1) a measurable criterion of academic progress, (2) random samples from a heterogeneous population to construct the experimental and control groups, (3) pretests of all groups, (4) the application of varying and measurable levels of homework to all groups over an extended period, and (5) posttests of all groups over an extended period.  Using survey methods rather than experimental controls we would need, (1) a measurable criterion of academic success, (2) a heterogeneous population that has been well studied both academically (including length and types of homework assignments) and in all other ways that might affect academic achievement (age, IQ, home life, social class, school quality, etc.), (3) selection of samples that vary only in the homework levels and types, and (4) measures of academic success.  The point is that social research trying to establish that X is, or fails to be, a cause of Y is technically and logistically very difficulty (This is a problem for the current assessment craze).  Social researchers generally settle for less than such perfection and so their results are open to simple critique.

With this in mind, Kohn cites favorably several studies that, "… do not support the notion that students who spend more time on homework have higher achievement gains than do their classmates." (27).  He cites another study that, "By contrast, the amount of time children spent reading for pleasure was strongly correlated with higher scores." (28).  But neither of these claims is inconsistent with the idea that homework provides an academic benefit.  For example, brighter children will spend less time on the same homework assignment than struggling children and will do better on the test.  Kohn is aware of this possibility but states that at least such studies show that homework is not doing the poorer student much good.  But what would have happened to the gap between the two groups had the struggling child spent only the time of the brighter child?  The fact is we don't know.  About the second citing,  of course independent readers will do better on tests than those who are not intrinsically motivate to read.  Statistically they will be the high achievers in the first instance as a result of IQ, upper class home lives, and more.  What has this to do with homework?  We don't know because the necessary controls are not present.

Kohn has written elsewhere about the limitations of standardized testing, (The Case Against Standardized Testing, 2000).  On the basis of this he disregards studies showing correlations between homework and testing success.  Scores on standardized tests are irrelevant.  About other studies that seemed to show positive results from homework, "… the measures of success basically involved memorizing and regurgitating facts." (40).  The author reports a late 1970s study of 300 New Jersey "A+ teachers" that found that these exceptional teachers, "… tended to give less homework …" (44)  Fine, but what were their schools like?, What did they teach?  Did they encounter behavior problems?  What were the conditions of their students' home lives?  Kohn notes that there are data showing correlations between homework and final grades.  But he dismisses these results on the basis that grades have little or no reliability as measures of learning, "… the basis of a grade is typically as subjective as the result is uninformative.  Any given assignment may well be given two different grades by two equally qualified teachers – and … two different grades by a single teacher who reads it at two different times." (33).  This claim about grade reliability is unsupported and in fact, if testing were of matters as rote and boring as Kohn has claimed, the testing should have no reliability problems at all (although its validity as a measure of important learning would be questionable).

There is more to this book rebutting arguments for homework that are unrelated to academic progress.  And there is more to the author's overall perspective.  Alfie Kohn has strong ideas about the educational system in the U.S. and I expect much of it is correct.  The system is overly mechanized, politicians' discussions focus on national competitiveness (the U.S. vs. Japan, Europe, etc.) rather than on student needs, instruction is not sufficiently individualized, the emphasis on standardized testing strips the curricula of challenges that develop creativity and higher order thinking skills, children's interests are systematically ignored, and more.  I look forward to his next effort.

© 2007 John Mullen

John D. Mullen is Professor of Philosophy at Dowling College in Oakdale, New York.  He has written the widely read book, Kierkegaard's Philosophy, a logic text, Hard Thinking, and co-authored with Byron M. Roth, Decision Making: Its Logic and Practice.  Most recently in 2006 he has written "Nature, Nurture and Individual Change", which appears on-line in the journal Behavior and Philosophy and argues that the issue of nature vs. nurture is irrelevant to questions of personal change.

Categories: General