17
EDUCATIONAL THEORY Summer 1987, Vol. 37, No. 3 0 1987 by the Board of Trustees of the University of Illinois Teacher Accountability and the Causal Theory of Teaching By David l ? Ericson and Frederick S. Ellett, Jr. In this most recent epoch of educational reform, the constant and insistent theme is that of accountability. Confronted with evidence that the schools once more are failing to educate the young, state after state is pressing for ways to curb perceived educational irresponsibility. In assigning blame, there has been finger-pointing in many directions. But like the fabled roads of the Roman Empire, they all seem ultimately to lead to one place: the teaching profession. What is it about the teaching profession that makes it such a convenient target when things seem to go badly in education? Given that our civic and cultural institutions are under siege, the structure of our family life apparently threatened, the structures of authority undermined, and the gap between rich and poor possibly widened, one might be tempted to locate the source of our educational ills in other places than the teaching profession. Instead, most legislatures have focused almost entirely on the teaching profession and related curriculum requirements. (Of course, teacher laxity has been held responsible for problems with the latter, too.) We believe there is a rather simple explanation for this focus upon teacher reforms as the key to ending our educational ills. The explanation highlights a notion embedded in the commonsense understanding of teaching and even in some prominent philo- sophical analyses of teaching. The notion is that whatever else teaching involves, it necessarily involves the attempt to bring about learning. As part of a “making something happen” profession, the teaching “act” and “enterprise” appear necessarily to be caught up in a causal endeavor. Despite the differences in their recent exchange, Robert Ennis, on the one hand, and Jim Macmillan and Jim Garrison, on the other, all hold this causal thesis on teaching.’ With the exception of Thomas F. Green, to whom we shall return, most of the prominent philosophical analyses of teaching maintain, in variants, the causal thesis. Clearly, too, the causal understanding of teaching motivates nearly all empirical educational research on teaching. And, finally, it is precisely this causal thesis that leads legislatures and school districts to focus on the teaching profession in connection with educational accountability. For if the teaching/learning situation is causal in nature, it seems easy to draw one of the following conclusions. If students are not learning, then teachers are failing or else they have not been teaching at all. In either case, it becomes possible to hold teachers responsible for their failure to educate the young. Indeed, it seems that the causal thesis of teaching directly implies that teachers should be held strictly accountable for their students’ Correspondence: Both authors, Graduate School of Education, University of California, Los Angeles, Los Angeles, CA 90024. 1. Robert H. Ennis, “On Causality,” Educational Researcher 2, no. 6 (1973): 4-11, and “Is Answering Questions Teaching?” Educational Theory 35, no. 4 (1986): 343-48; C. J. B. Macmillan and James W. Garrison, “An Erotetic Concept of Teaching,” Educational Tbeory 33, nos. 3-4 (1983): 157-67, “Erotetics Revisited,” Educational Tbeory 35, no. 4 (1986): 355-62, and “An Erotetic Notion of Causality” (Paper delivered at the Forty-second Annual Meeting of the Philosophy of Education Society, Montreal, Canada, 13 April 1986. James W. Garrison delivered a version of this paper under the title “Intentional Causation in Educational Research” at the Annual Meeting of the American Educational Research Association, San Francisco, 20 April 1986). 277 VOLUME 37, NUMBER 3

Teacher Accountability and the Causal Theory of Teaching

Embed Size (px)

Citation preview

Page 1: Teacher Accountability and the Causal Theory of Teaching

EDUCATIONAL THEORY Summer 1987, Vol. 37, No. 3 0 1987 by the Board of Trustees of the University of Illinois

Teacher Accountability and the Causal Theory of Teaching

By David l? Ericson and Frederick S. Ellett, J r .

In this most recent epoch of educational reform, the constant and insistent theme is that of accountability. Confronted with evidence that the schools once more are failing to educate the young, state after state is pressing for ways to curb perceived educational irresponsibility. In assigning blame, there has been finger-pointing in many directions. But like the fabled roads of the Roman Empire, they all seem ultimately to lead to one place: the teaching profession.

What is it about the teaching profession that makes it such a convenient target when things seem to go badly in education? Given that our civic and cultural institutions are under siege, the structure of our family life apparently threatened, the structures of authority undermined, and the gap between rich and poor possibly widened, one might be tempted to locate the source of our educational ills in other places than the teaching profession. Instead, most legislatures have focused almost entirely on the teaching profession and related curriculum requirements. (Of course, teacher laxity has been held responsible for problems with the latter, too.)

We believe there is a rather simple explanation for this focus upon teacher reforms as the key to ending our educational ills. The explanation highlights a notion embedded in the commonsense understanding of teaching and even in some prominent philo- sophical analyses of teaching. The notion is that whatever else teaching involves, it necessarily involves the attempt to bring about learning. As part of a “making something happen” profession, the teaching “act” and “enterprise” appear necessarily to be caught up in a causal endeavor. Despite the differences in their recent exchange, Robert Ennis, on the one hand, and Jim Macmillan and Jim Garrison, on the other, all hold this causal thesis on teaching.’ With the exception of Thomas F. Green, to whom we shall return, most of the prominent philosophical analyses of teaching maintain, in variants, the causal thesis. Clearly, too, the causal understanding of teaching motivates nearly all empirical educational research on teaching. And, finally, it is precisely this causal thesis that leads legislatures and school districts to focus on the teaching profession in connection with educational accountability. For if the teaching/learning situation is causal in nature, it seems easy to draw one of the following conclusions. If students are not learning, then teachers are failing or else they have not been teaching at all. In either case, it becomes possible to hold teachers responsible for their failure to educate the young. Indeed, it seems that the causal thesis of teaching directly implies that teachers should be held strictly accountable for their students’

Correspondence: Both authors, Graduate School of Education, University of California, Los Angeles, Los Angeles, CA 90024.

1. Robert H. Ennis, “On Causality,” Educational Researcher 2, no. 6 (1973): 4-11, and “Is Answering Questions Teaching?” Educational Theory 35, no. 4 (1986): 343-48; C. J. B. Macmillan and James W. Garrison, “An Erotetic Concept of Teaching,” Educational Tbeory 33, nos. 3-4 (1 983): 157-67, “Erotetics Revisited,” Educational Tbeory 35, no. 4 (1986): 355-62, and “An Erotetic Notion of Causality” (Paper delivered at the Forty-second Annual Meeting of the Philosophy of Education Society, Montreal, Canada, 13 April 1986. James W. Garrison delivered a version of this paper under the title “Intentional Causation in Educational Research” at the Annual Meeting of the American Educational Research Association, San Francisco, 20 April 1986).

277 VOLUME 37, NUMBER 3

Page 2: Teacher Accountability and the Causal Theory of Teaching

278 EDUCATIONAL THEORY

learning, or failure to learn.‘ Thus, philosophy, educational research, and commonsense ideas of teaching all help to undergird the teacher accountability movement.

As we shall attempt to show, this version of the causal thesis and its associated implications require serious modification. Though we, too, maintain a causal theory of teaching, we intend to establish that it does not follow from a more adequate causal understanding of teaching that teachers are solely or even mainly responsible for dismal educational results. The widespread misconception about teaching is due to (1) an inappropriate conception of the form of causal relations in understanding the teaching/ learning situation and (2) the omission of the student’s role in that situation.

In what follows we shall consider a number of philosophical views on the causal nature of teaching. In part 1 we review Robert H. Ennis’s theory of causality. In parts 2 and 3 we examine the “erotetic” theory of teaching proffered by C. J. B. Macmillan and James Garrison, as well as the views of Paul J. Diet1 and Thomas F. Green. For a variety of reasons, we shall show why each of these accounts, though bearing merit, is wanting in various ways. In part 3 we shall also put forward our own views of the teachingllearning situation that builds on the strengths of these other conceptions, but avoids, we hope, their weaknesses. Finally, in part 4 we shall briefly discuss some policy issues surrounding teacher accountability.

I.

Robert H. Ennis’s main article on causation appeared in the Educational Re~earcher.~ As we shall show, his analysis of causal statements implies a direct link between a teacher’s causing learning and a teacher’s responsibility. Macmillan and Garrison, whase works we consider in parts 2 and 3, have also been influenced by Ennis’s views of causation.

First, after giving a number of examples of implicitly causal linguistic expressions such as “bring about,” “results in,” “contributed to,” “affected,” etc., Ennis writes, “Two less obvious examples are taught and teach. . . .”4 This language yields a few lines later to this argument: “The education enterprise attempts to bring about changes in students. Bringing something about is unavoidably a causal notion. Therefore the central thrust of educational research taken as a whole must be toward the establishment of causal statement^."^ Since teaching is aimed at bringing about changes in students, we can be fairly confident that Ennis does maintain a causal theory of teaching -that is, that teaching causes learning - even though Ennis never explicitly says so.

Then, after distinguishing between general and specific causal statements,‘ Ennis offers the following analysis of specific causal statements:

A person making a specific causal statement of the form ‘X caused Y,’ (1)

2. The stakes in this issue are far from small. B. Rodman. in “Rating Teachers on Students’ Test Scores Sparks Furor, Legal Action in St. Louis,” Education Week 6, no. 2 (17 September 1986): 1 and 18, reports that the St. Louis Board of Education has begun evaluating teachers on the basis of student test performance (using a national standardized norm-referenced test and soon to be implemented criterion-referenced tests). If a given class of students falls below certain predetermined test levels, then their teachers are to be given “unsatisfactory” ratings. An “unsatisfactory” rating automatically earns a teacher “probationary” status (including a pay freeze) and can lead to eventual dismissal if student test scores fail to improve within one hundred days. The new School Board policy is currently being challenged in the courts by St. Louis Teacher Unions. It may be presumed that other school boards are following the case with interest.

3. Ennis, “On Causation.” 4. Ibid., 4. 5. Ibid., 4. 6. In ibid., 7, the distinction he makes between general and specific causal claims is not fully

clear since the two types of test situations for specific causal claims that Ennis cites involve generalizations. For example, one way of attacking specific causal claims “is by showing that under relevantly similar conditions the occurrence of something just like X is not followed or accompanied by something just like Y.” Controlled experimentation is the other test situation cited.

SUMMER 1987

Page 3: Teacher Accountability and the Causal Theory of Teaching

TEACHER ACCOUNTAEILITY

asserts that, given the other existing conditions, X, a particular thing, was sufficient to bring about the occurrence of Y, and (2) holds X responsible for that occurrence. I offer this as a complete analysis, though not a reduction.’

279

Now several things are interesting about Ennis’s account. First of all, he appears to rule out a priori the possibility of probabilistic causal relations in which a cause increases the probability of the occurrence of the effect, but is not sufficient for its occurrence. Recently, many philosophers of science, social scientists, and physical scientistis have pointed to the widespread importance of this conception of causation. In probability terms it can be roughly stated: 1 > P (Y/X) > O.* Although Ennis does reject this notion with little or no argument, this conception, or more importantly a related conception of causation, INUP causation, may prove vital in analyzing the teaching/learning situation. We shall reserve this discussion for part 3.

What we need to concentrate on here, however, is (2), Ennis’s “responsibility” criterion in analyzing causal claims, for it seems to be highly pertinent to teacher accountability, Indeed, though Ennis only states that “X caused Y” means in part that we hold X responsible for Y’s occurrence, in the causal theory of teaching it takes on a new meaning. For if teaching causes learning, and if the students fail to learn, then for Ennis, it would seem, the teacher failed to teach. But since it is the teacher’s job to teach, we may hold the teacher responsible for failing to bring about learning just as much as we would say the teacher is responsible if the students do learn. (Omissions and absences can be causes for Ennis; see footnote 7). Thus, Ennis appears to build teacher accountability into the causal theory of teaching as a result of his general analysis of causal statements.

Ennis, unfortunately, is very brief in his analysis of “responsibility.” At times, he seems to mean that we hold the cause responsible in its ordinary-language sense of being “morally, socially, legally, or institutionally answerable for its effect.” At other times, he seems to mean only that, of a number of actual causal conditions, we pick out one or two “as responsible.” Whatever the case, if he means the first, his general account is false; but if he means the second, then his general account is vacuously circular.

Concerning the first, it is clear that we do not hold natural processes morally, legally, socially, or institutionally answerable for their effects. Even in torts, the legal fiction “act of God” is used to dismiss responsibility. As Hart and Honor6 succinctly put it:

We still speak of inanimate or natural causes such as storms, floods, germs, or the failure of electricity supply as “responsible for” disasters; this mode of expression, now taken only to mean that they caused the-disasters, no doubt originated in the belief that all that happens is the work of spirits when it is not that of men. (Our empha~is)~

Now while this normative or legal reading of Ennis’s responsibility criterion may be used nonvacuously to elucidate the meaning of causal statements, it has the obvious defect of imputing liability to natural phenomena under legal or moral, etc., rules and so renders such phenomena blameworthy and liable to sanctions. If God or spirits

7. Ibid., our emphasis. Though Ennis calls the cause “a particular thing” here, elsewhere (p. 5) he is widely tolerant in his causal ontology; beyond Davidsonian events, he admits states of affairs, acts, processes, and lacks (nonobtaining states of affairs).

8. In Frederick S. Ellett, Jr., and David P. Ericson, “Causal Laws and Laws of Association,” No& 19, no. 4 (1985): 537-49, and “Correlation, Partial Correlation, and Causation;’ Synthese 67 (May 1986): 157-73, we provide a precise rendering and qualification of this notion. We also canvass the views of a number of philosophers and social scientists who regard it as important. See also our “An Analysis of Probabilistic Causation in Dichotomous Structures,” Synthese 67 (May 1986): 175-93.

9. H. L. A. Hart and A. M. Honore, Causation in the Law (Oxford: Oxford University Press, 1959), 61.

VOLUME 37. NUMBER 3

Page 4: Teacher Accountability and the Causal Theory of Teaching

280 EDUCATIONAL THEORY

were the causes of such fortune and misfortune, then this reading may have some merit. It is difficult to see how such an account is to be reconciled with the picture of a spirit-less nature as given by our contemporary life and natural sciences, however. And given Ennis’s own allegiance to the critical spirit of science, it is further difficult to believe that he would accept this rendering.

Unfortunately, Ennis suggests that he might accept this rendering. In discussing the role of value judgments in responsibility ascriptions, he uses the example of students’ being raised in the rural South as causing educational disadvantagement. In contrast to that conclusion, he argues that “living in the rural South was not really the cause, but rather a number of features that seemed to go along with living in the rural South. On this revised view it is these features that are being held responsible, and thus are candidates for praise and blame, and are recommended intervention points” (our emphasis).” The connective “and thus” in this passage suggests that Ennis’s concept of responsibility is logically linked to the ethical/legal concept of responsibility. Strictly interpreted, therefore, this is a version of the spirits-as-cause view. Elsewhere, however, he seems to imply that not all “responsibility” ascriptions need or require the assignment of praise or blame.”

Whatever the case, the principle of charity in interpretation can perhaps be invoked to conclude that Ennis is guilty of no more than imprecise language here. Therefore, let us interpret him as holding no version of the spirits-as-cause view. But if one takes this interpretation, a major difficulty arises. For if “X caused Y” means (along with the sufficiency criterion) that we “hold X responsible for that occurrence,” and if “holding X responsible for that occurrence” simply means that we pick out X “from a number of causal conditions and [call] it the cause,”12 then the responsibility criterion is patently and vacuously circular. The responsibility criterion was to be used to elucidate “X caused Y.” But if “holding X responsible” merely means “picking out and calling X the cause,” then, by substitution, “calling X the cause” does no work in analyzing “X caused Y.” For Ennis has not explicated any of the factors involved in picking out and naming one of the causal conditions “the ~ause.” ’~ (The Hart and Honor6 quote above makes that very clear as well.)

Given these interpretations, then, the “responsibility” criterion is either false or vacuously redundant. We shall suggest in part 3, nonetheless, that Ennis’s distinction between “the cause” and various kinds of “partial causes,” when suitably reinterpreted, does have a role to play in discussing the causal theory of teaching. It is sufficient now to note that, lacking any other interpretations of the “responsibility” criterion, ‘causation’ and the ordinary sense of ‘responsibility’ appear to be wholly independent concepts. This is important to keep in mind as we turn to Macmillan and Garrison’s erotetic theory of teaching, since they appear to rely in crucial places on Ennis’s account.

11.

In several published articles, C. J. 8. Macmillan and James W. Garrison have developed what they call an “erotetic” theory of tea~hing.’~ In this they are building upon Jaakko Hintikka’s attempt to develop an erotetic logic of questions and answers: the logical presuppositions of questions and what it is for something to be a determinate answer to a particular que~t i0n. l~ For Macmillan and Garrison understand the providing

10. Ennis, “On Causality,” 10. 11. Ibid., 11, point 14. 12. Ibid., 7. 13. In his “Causal Judgments and Causal Explanations,” Journal of Philosophy 62, no. 23

(1965), Samuel Gorovitz presents an analysis of those features (the standards of comparison) that are involved in singling out one causal factor from among the other causal factors and calling it “the cause.” Gorovitz’s analysis shows that it would be extremely misleading to talk about these singling-out features as “holding the factor responsible.”

14. Macmillan and Garrison, “An Erotetic Concept of Teaching,” and “Erotetics Revisted.” 15. Macmillan and Garrison, “An Erotetic Concept of Teaching.”

SUMMER 1987

Page 5: Teacher Accountability and the Causal Theory of Teaching

TEACHER ACCOUNTABILITY 281

of answers to the students‘ questions to be the logically central content of the concept of teaching, while all else a teacher might do is considered to be peripheral at best. Moreover, in an unpublished paper that was presented in several public symposia, Macmillan and Garrison connect their erotetic theory of teaching to an “erotetic conception of causation.”16

As might be surmised, their erotetic theory of teaching is preeminently intellectualist and intentionalist in orientation. As they put it:

It is the intention of teaching acts to answer the questions that the auditor (student) epistemologically ought to ask, given his intellectual predicaments with regard to the subject matter. Insofar as these questions can be put in a clear and unconfused way, the questions will have exact and determinate semantical (and possibly syntactical) content. The correct answer to a properly posed question must be couched in the terms within which it was asked. We may say. . . , then, that intentional intellectual acts of teaching hit their marks when they satisfy the semantical and syntactical demands of the questions the students epistemologically ought to have asked given their intellectual predicaments.”

Two aspects of their position are important to note here. First, erotetic logic is concerned foremostly with subject matter that yields rather precise semantical and syntactical conditions of satisfaction. As we shall see, it is not at all clear how it can be applied to the teaching of subject matter that lacks a determinate semantics (e.g., normative language, procedural and pragmatic principles, methodology, all of mathematics, to wit: all that is essentially nonempirical). The second, somewhat related, aspect involves their apparent dismissal of the pragmatic dimension involved in the relations between teacher and student. As they write, “It is not that the auditor does ask the question, or that the teacher believes that he would ask the question, but rather the teacher believes that in some sense the student ought to ask the question [relative to the student’s intellectual predicament]” (original emphasis).” Beyond being in an intellectual predicament, the learner plays a small role in Macmillan and Garrison’s theory. In their view, the central feature of the teaching/learning situation is the teacher’s answering the questions regardless of what the learner does. Pendlebury has questioned this omission of any mention of other features of the ~tudent . ’~ And in our symposia responses to Macmillan and Garrison, we have questioned whether it is not a logically central teaching act to put students into a position in which they are intellectual perplexed, not merely answer the questions (see also Scheffler).20 In responses to these questions, Macmillan and Garrison are rather sketchy. They simply assert that their theory of teaching is both subject and student centered, but they provide no details.2’ And a bit more mysteriously they claim, “There is a place for learning theory here, but it is a different place from one that puts it logically ahead of teaching.”22

It is, however, with Macmillan and Garrison’s attempt to develop an erotetic-cum- causal theory of teaching to serve “as the basis for the empirical study of teaching as an intentional activity”23 that a rather different pictures emerges. Perhaps anticipating

16. The draft, “An Erotetic Notion of Causality,” is intended to appear as a chapter in a book they are writing on their erotetic theory of teaching. We responded to this paper at both symposia. Since Macmillan and Garrison are reconsidering some of the ideas of this paper, the full views attributed to them here may not be current. Still, the direction of their thinking is rather clear.

17. Macmillan and Garrison, “An Erotetic Concept of Teaching,” 160. 18. Ibid., 159. 19. Shirley Pendlebury, “Teaching: Response and Responsibility,” Educational Theory 36, no.

20. Israel Scheffler, “Reflections on Educational Relevance,” in his Reason and Teaching

21. Macmillan and Garrison, “An Erotetic Notion of Causality,” 13. 22. Macmillan and Garrison, “Erotetics Revisted,” 356. 23. Macmillan and Garrison, “An Erotetic Notion of Causality,” 4.

4 (1 986): 349-54.

(Indianapolis, Ind.: Bobbs-Merill, 1973).

VOLUME 37, NUMBER 3

Page 6: Teacher Accountability and the Causal Theory of Teaching

282 EDUCATIONAL THEORY

Pendlebury’s and our complaint that an essential pragmatic dimension (one which would mention the students’ attitudes and beliefs) is missing in their account, they attempt to supplement Hintikka’s erotetic logic with a salient feature of Ennis’s work on causation?“ Beyond the syntax and semantics of erotetic logic, they introduce a third “epistemological” or pragmatic condition: “A satisfactory answer must satisfy the questioner and not just the question” (our emphasis).25 Thus, while Hintikka notes that a quesrioner’s question (1) “Who lives in that house?” becomes under erotetic trans- formation rules (2) “Bring it about that I know who lives in that house,” Hintikka is really unconcerned with the imperative component of “bring it about.” Macmillan and Garrison, on the other hand, see this component as the foundation for a causal (and erotetic) theory of teaching. For, following Ennis, (2) above may be rendered as (3) ”Cause it that I know who lives in the house.”

Dubbing this a notion of “erotetic causation,” they proceed to argue that erotetic logic and erotetic causality are, nonetheless, scarcely indistinguishable and to show how erotetic causation differs from the “traditional analysis” of causation.26 In doing so, they attempt to show not only that the relation between teaching and learning is an empirical, causal one (suitable for educational researchers), but also that it is a logical one. Rejecting Hume’s insistence that cause and effect must be logically independent, they argue that the relation between the answer (cause) and the satisfaction of the question (effect) is one of logic (semantical and epistemological, not merely ~yntactic).’~ As they put it, “the [correct] answer must bring about the satisfaction of the content and epistemological conditions of the question” (original emphasis).*’ Thus, on Macmillan and Garrison’s view, the learner is satisfied (a pragmatic issue) when the teacher satisfies the syntactic and semantic demands of the question. Or in their terminology, when the teacher correctly answers the question, “the student will experience the erotetic causal nexus.”2g

Now it can be made clear how Macmillan and Garrison’s erotetic/causal theory of teaching results in a determinate position on teacher accountability. They do not hesitate to state that “a complete answer brings about a change in the epistemological state of the student who epistemologically ought to ask it” (original emphasis).3° Since for Macmillan and Garrison the logically central task of teaching is to answer the questions, it is not difficult to see who will be held accountable should the student’s epistemological state not be changed in the requisite way. It will be the teacher‘s fault. In this, they approvingly cite Paul Dietl’s view that teaching without learning is of logical necessity a “mi~fire.”~’ Going apparently beyond the public’s view of teacher accountability, they avow, “The teacher’s responsibility may be broader than ordinarily thought” (Out‘ emphasis).32

Macmillan and Garrison clearly recognize that certain factors “external” to the teaching tasks may intervene to mitigate the teacher’s responsibility. It is interesting to note, however, how far the “internal” or teacher-related factors extend. In keeping with their erotetic conception, the primary “internal” factors for teacher failure include the

24. In their reply to Pendlebury, this supplement is missing. 25. Macmillan and Garrison, “An Erotetic Notion of Causality,” 9. 26. Ibid., 13. 27. Ibid., 27. 28. Ibid., 13. 29. Ibid., 15-16. 30. The discerning reader may see that already there are major difficulties in Macmillan and

Garrison’s formulation. First, since their account excludes the student actually having or raising a question, what sense can be given to the notion of satisfying the questioner when there is no questioner? Second, even if the student poses a question and the teacher answers it Completely (i.e., satisfies the semantical and syntactical demands), no epistemological state of the student need necessarily be changed. The student may simply not understand the answer or may not believe it. (We comment on this last point below.) Moreover, it should be noted that in the original Hintikka discussion, the question, with its imperatival aspect, is addressed to no one in particular.

31. Macmillan and Garrison, “Erotetics Revisted,” 358. 32. Ibid., 359.

SUMMER 1987

Page 7: Teacher Accountability and the Causal Theory of Teaching

TEACHER ACCOUNTABILITY 283

teacher’s answering the right questions incorrectly, the teacher’s answering the wrong or irrelevant questions, and the teacher’s failing to speak to students’ epistemological oughts.33 But they include more, as this passage illustrates:

Bertha, we find, has not been attending to the demonstration of different moves which constitute swimming and cannot execute them as a result. This is a failure in teaching, since Albert was not teaching her how to swim [sic]; the teacher’s responsibility extends to being aware of and taking care that the conditions of her getting the material are met. Clara, on the other hand, attends carefully, but gets a cramp and is paralyzed; she too cannot swim, but it is not a failure of teaching that makes it so. (Our emphasis except for “her”)34

Now it is not fully clear to us why Macmillan and Garrison extend the teacher’s responsibility to include making sure students attend to the lesson. And since making sure they attend is obviously not the same as correctly answering the questions, it scarcely appears to satisfy their criteria for a central (intellectual) teaching act. At best, in their scheme, it is a candidate for what they call peripheral or strategic teaching acts. It may be, however, that its inclusion results from a pressure upon their analysis of the relation between ‘teaching’ and ‘learning.’ In taking that relation to be both causal and logical, once the teacher satisfies the semantical and syntactical conditions of the question arising from the student’s “ought,” necessarily the questioner is satisfied too (an epistemological, pragmatic condition). (But remember, they do not require that the teacher answer the question a student may or may not have: only the questions the student ought to have.) To allow, then, a situation in which the syntactical and semantical demands of the question are satisfied, but in which the student fails to learn, is something necessarily forbidden by their theory. In this way, it becomes imperative that the teacher be held accountable not merely for answering the questions, but also for taking care that the epistemological conditions for the student‘s getting the material are met.

Unfortunately, this at once undermines their distinction between central (i.e., intellectual) and peripheral teaching acts - the very core of their erotetic conception of teaching. So, in order to preserve that distinction, they might be well advised to limit the teacher‘s responsibility to answering correctly the questions ( i a , satisfying the semantic and syntactic demands only). (Bertha’s problems of inattention might be hers alone. Indeed, in our own analysis of the teaching/learning situation, we shall suggest something of this sort.) Unfortunately for Macmillan and Garrison’s theory, this is, to repeat, an impossible avenue of escape. Once the syntactic and semantic demands of the questions are met, it is logically necessary that learning ensues (the student necessarily “experiences the erotetic causal nexus”). To allow otherwise is to acknowl- edge a logical gap between teacher and learner, between answering the question and the student’s learning. And it is precisely that gap that their erotetic logical and causal theory of teaching forbids.

Thus, in whichever direction Macmillan and Garrison turn, they must either yield the distinction between central and peripheral teaching acts, a direction that undercuts the usefulness of erotetic logic in analyzing the concept of teaching but which apparently preserves their account of teacher responsibility, or else they must discard their notion of erotetic causation, a direction that opens the gap between answering the question and learning and places the usefulness of their erotetic framework for educational researchers in question. Can they split the horns of this dilemma and locate a saving solution? We are doubtful. Even so, there are many useful ideas in their theory that we shall now attempt to sketch - and that is all we can do here - into our own.

33. Ibid. 34. Ibid.

VOLUME 37, NUMBER 3

Page 8: Teacher Accountability and the Causal Theory of Teaching

284 EDUCATIONAL THEORY

Ill.

As mentioned early in the previous section, we have grave doubts about the general success of erotetic logic in elucidating the commonsense concept of teaching. It may prove useful in giving an account of teaching a very small body of subject matter in which questions can be formulated with precise syntactical and semantical conditions and in which the answers to those questions are precisely determinable (e.g., certain empirical knowledge). Most subject matter of educational importance, however, is not of this sort. Normative subject matter, methodological canons, and mathematics have no determinate semantics. And, interestingly, insofar as causal explanations in the sciences are pragmatic in nature, erotetic logic cannot hope to give an adequate analysis of their conditions of satisfaction (on this, see van Fraassen and Achinstein).= There is, here, an even graver difficulty. Hintikka’s erotetic logic is designed to cover questions that admit only of one, uniquely determined, answer. It cannot deal with questions that admit of more than one “reasonable” answer, since no more than one reasonable answer can be true of the world, and all may be false.

Macmillan and Garrison implicitly recognize this constraint when they state that the teacher’s answer must be And though they mention that they can deal with matters of belief, they have yet to show how erotetic logic can handle them. In education, as in life and science, very little is of certainty, for which reason many philosophers of education have held that the aim of teaching is to develop reasonable (or evidential) belief in students - or to help them become evidential believer^.^' Consider, moreover, the difficulties involved in accommodating skill acquisition, moral, social, and emotional learning to their

Though we do not see much hope, beyond an extremely narrow range, for an erotetic analysis of teaching, several aspects of Macmillan and Garrison’s program seem to us crucial and true. The first is their emphasis, shared by a wide variety of philosophers of education, on the intentionality of teaching in all of its forms. Whatever else it is, teaching is an intentional activity in relation to learning. Second, we underline our support for the view that it is logically impossible to characterize teaching without reference to bringing about learning. The intention of teaching acts is the bringing about of learning. Moreover, like Macmillan and Garrison and, we think, Ennis, we take the relation between teaching and learning to be causal in nature.

Now it is this last claim that, for some, is controversial. Yet we think its truth is crucial not merely for the efficacy of educational research on teaching, but also for showing what the limits are for teacher accountability. For those who believe that teaching is logically related to learning, such as Macmillan and Garrison, are also likely to believe that teachers are far more accountable for the failure to bring about learning than we think. Indeed, it follows from this view that if the teacher is unsuccessful in bringing about learning, the teacher may have been doing anything but teaching - absent “external” defeating factors, we have here the mark of negligence or incom- petence.

The “logical relationists” fall into two camps: (1) those who believe teaching is both logically and causally related to learning (Paul J. Diet1 and Macrnillan and Garrison are examples here); (2) those who believe that the relation is logical in some sense, but who believe logical relations rule out causal relations (see Melden and Stoutland

35. Bas C. van Fraassen, The Scientific Image (Oxford: Oxford University Press, 1980); and

36. Macmillan and Garrison, ”Erotetics Revisted;’ 359 and 360. 37. See, e.g., Thomas F. Green, The Activities of Teaching (New York: McGraw-Hill, 1971). 38. In fairness to Macmillan and Garrison, they gamely try to assimilate teaching someone

how to swim to their framework. As they recognize, practice is necessary in this. Yet they regard practicing episodes (supervised or not) as question answering just as much as intellectual episodes of teaching that. Though they protest that this is not ”verbal magic,” it is difficult to see how it is anything else. In practicing, if anyone is answering questions, it is the would-be swimmer, not the coach. See “Erotetics Revisited,” 359-60.

Peter Achinstein, The Nature of Explanation (Oxford: Oxford University Press, 1983).

SUMMER 1987

Page 9: Teacher Accountability and the Causal Theory of Teaching

TEACHER ACCOUNTABILITY 285

on this general The latter accept some version of Hume’s analysis of causation that holds that cause and effect must be logically independent (under some description); the former reject the Humean analysis (see Macmillan and On either account, however, a “strong” thesis that teachers are strictly and solely ac- countable for lack of student learning necessarily follows.

Thus, it will be sufficient for our purposes here to show that teaching is not logically related to learning in any interesting sense that rules out causal relations or supports the strong thesis for teacher accountability. Teaching and learning in particular are logically independent in the Humean-Davidsonian sense, as we shall show. (This claim does not generalize, however. We do believe that in some cases there are genuine causal connections between phenomena that also involve logical connections that do violate Humean-Davidsonian desiderata - see, e.g., Taylor and Er i cs~n .~ ’ Thus, in those cases we may well side with Dietl and Macmillan and Garrison. We simply deny that this holds for teaching and learning.) In an opposite direction, we reject the position that there are neither logical nor causal relations between teaching and learning. Thomas F. Green, most notably, champions this p~s i t ion.~ ’ There is much to be learned from Green on teaching and on learning and teacher accountability, we think, but his view of causation is not sufficiently subtle.

Our point of departure is the well-known task-achievement analysis of teaching and learning, a point we share with Ennis. Suitably understood and modified, we believe it is the correct analysis. In commenting upon it, Macmillan and Garrison note that it has two principal goals: (1) it drives a wedge between teaching and learning, so that in the task sense of ‘teach’ one can say that “X is teaching“ without implying that the subject matter is being learnt; (2) it clarifies the relation between the activities of teaching and the success of them, so that, like seeking and finding, teaching and learning are related as task and a~h ievemen t .~~ Macmillan and Garrison offer several reasons for thinking the task-achievement analysis is inadequate: first, Dietl’s argument for showing that teaching without learning is “necessarily a second, their own argument based on their notion of erotetic third, their gloss on Sylvain Bromberger’s analysis of activity and accomplishment and fourth, their claim that task-achievement analyses leave the relation between teaching and learning a mysterious In rebuttal to these arguments, we shall provide evidence that leads one on rational grounds to prefer a task-achievement analysis to any rival candidate.

First, consider Dietl’s argument in his classic, but all too neglected, posthumous article “Teaching, Learning and Kn~wing.”~ ’ After reviewing the arguments of B. 0. Smith, I. Scheffler, P. Komisar, and T. F. Green, all distinguished adherents of the view that teaching and learning are not logically connected, Dietl offers his own positive argument to show that they are mistaken. Without offering any explanations for his choice, Dietl takes acts of explaining as the “most promising” teaching act in showing a logical connection between teaching and learning and asks “whether or not explaining

39. A. I. Melden, FreeAction (London: Routledge & Kegan Paul, 1961), and Frederick Stoutland, “The Logical Connection Argument,” in Studies in the Theory of Knowledge, American Philosophical Quarterly Monograph Series (Oxford: Basil Blackwell, 1971).

40. Macmillan and Garrison, “An Erotetic Notion of Causality.” 41. Charles Taylor, “Explaining Action,” lnquiry 13; and David P. Ericson, “Cognition, Emotion,

and Attribution Theory in the Psychology of Motivation,” in Philosophy of Education, 1982, ed. Donna H. Kerr (Normal, 111.: Philosophy of Education Society, 1983). and “Emotion and Action in Cognitive Psychology: Breaching a Fashionable Philosophical Fence,” in Philosophy of Education, 7984, ed. Emily Robertson (Normal, 111.: Philosophy of Education Society, 1985).

42. Green, The Activities of Teaching, 140-42. 43. Macmillan and Garrison, “Erotetics Revisited,” 357-58. 44. Ibid., 358. 45. Macmillan and Garrison, “An Erotetic Notion of Causality.” 46. Macmillan and Garrison, “Erotetics Revisited,” 358. 47. Ibid. 48. Paul J. Dietl, “Teaching, Learning and Knowing,” Educational Philosophy and Theory 5,

no. 1 (1973).

VOLUME 37, NUMBER 3

Page 10: Teacher Accountability and the Causal Theory of Teaching

286 EDUCATIONAL THEORY

can be explicated without reference to understanding (its analogue to learning).”4s Upon dispensing with the deductive-nomological model of explanation (that makes no ref- erence to understanding), Dietl concludes:

It appears, then, that ‘explain’ and ‘understand’ are conceptually connected. One could not perfectly well understand the former without understanding the latter. Explanations which do not provide understanding are necessarily ab- normal. . . . The point Of engaging in the activity would be lost. . . . And teaching is to learning here, as explaining is to Understanding.. . . [Without students present and my speaking at an audience-appropriate level of complexity] it could not be my serious intention to bring [learning] about by what I am doing. So teaching appears to be both causally and definitionally connected to learning. It is fogically necessary that normally the result of teaching be learning. Teaching without learning is necessarily a misfire. (Emphases ours; author’s original emphasis removed)50

Dietl is right. ‘Explaining’ cannot be definitionally characterized independently of ’understanding’; and ‘teaching’ cannot be definitionally characterized independently of ‘learning’. To explain is to intend to bring about understanding; and, as he further says, to teach is to intend to bring about learning. ‘Teaching’ and ‘learning’ are logically connected in these ways. But in both cases Dietl fallaciously argues from the fact that ‘explaining’ (and ‘teaching’) must be characterized by the intent to bring about under- standing (and learning) to the conclusion that one logically cannot be engaged in the activity of explaining (and teaching) without the occurrence of understanding (and learning). Quite obviously, however, the characterization of an activity is one thing, but the success (or nonsuccess) of the endeavor is quite another. It is only if engagement in teaching somehow guaranteed (its own rate of) success that we would have a case in which the phenomena of teaching (and not its characterization) are logically related to the phenomena of learning. Intentions to bring about learning, like other intentions, are often thwarted. Even Dietl admits ”the undeniable fact that teaching, even good teaching, can take place without the occurrence of learning.”” Thus, it is NOT the case that “teaching without learning is necessarily a misfire.” They are logically independent phenomena, regardless of their logically connected characterization (and here, of course, ‘learning’ makes no essential reference to ‘teaching’).

Concerning Macmillan and Garrison’s second argument, it follows from the failure of Dietl’s argument that their own argument concerning erotetic causation must fail. For if good teaching, to be construed in terms of satisfying the syntactical and semantical desiderata of the question (and with the intention of bringing about learning), need not result in learning (necessarily experiencing the erotetic causal nexus), then exposed is the logical gap between teaching and learning. Now they do say that the teaching must satisfy the questioner, not merely the question. But here, as is typical, they slide from their own account of teaching in which teaching is aimed only at answering the questions students ought to ask to a view that they deny: viz, that the teacher must address questions that the student actually asks. They cannot have it both ways.5* Moreover, though there may be a logical connection of some sort between syntactically and

49. Ibid., 7. 50. Ibid., 9-10, 51. Ibid., 4. 52. And, at any rate, it would seem strange to tie teaching - and especially good teaching -

to satisfying the questioner. The question actually raised by a student may be irrelevant, out of bounds, etc. Decent teaching may call for rejecting the question in part or total, or may even require refusing to answer it because the teacher wants the students to solve it on their own. One noted educator, John Passmore, in “On Teaching to be Critical,” in Education and Reason, ed. R. F. Dearden, P. H. Hirst, and R. S. Peters (London: Routledge & Kegan Paul, 1975), 39, writes that “one should add that the very existence of the problem is not, normally, known to the student. One of the educator’s tasks is to make his students puzzled. . . . The fact remains that unless his pupils leave school puzzled his teachers have failed as educators, however successful they may have been as instructors.”

SUMMER 1987

Page 11: Teacher Accountability and the Causal Theory of Teaching

TEACHER ACCOUNTABILITY 287

semantically specifiable questions and their answers, the relation between teacher and learner here IS, as they admit, a pragmatic one. And it is most difficult to see how one could transform pragmatic relations into logical relations, except by a most stipulative definition.

The third argument that Macmillan and Garrison provide for thinking that teaching and learning are so tightly logically related depends upon Sylvain Bromberger’s distinction between “activity” and “accomplishment” For example, the “taught” in “X taught Y to Z“ can be taken to mean that “X was engaged in teaching Y to Z“ (activity term) where there is no imputation of success. Or else it can mean “X was successful in teaching Y to 2” (accomplishment term). Now, of course, it would be an utter contradiction to say that “X was successful in teaching Y to 2, but Z did not learn Y.” No doubt there. Surely, bringing about learning is the intended aim of teaching. And when the aim has been reached, the teaching has come to a successful cessation. If the aim has not been reached, the intention embedded in teaching activities has been frustrated. But none of this shows that teaching logically requires learning to occur. For that to be true, it is necessary to show that “X was engaged in teaching Y to Z (activity term), but Z did not learn Y” is a logical contradiction. In such a case, failure to learn would a priori rule out the fact that X was engaged in teaching at all. Macmillan and Garrison have said nothing in support of the claim that the activity term results in a contradiction. Indeed, it is difficult to see what could be said in support, since failure to learn implies only that the teaching was unsuccessful, not that teaching did not occur at all.

The fourth and last argument concerns the so-called mysterious nature of task- achievement analysis. This tag, borrowed from Dietl’s analysis, was first used against Paul K o m i s a r ’ ~ ~ ~ and Tom task-achievement analyses. Komisar, in our view correctly, argued against a logical connection between teaching and learning but admitted that there is something of a “conceptual communion” between the ideas. Since there is an implication of divine mysteries in the notion of theological communion, Diet1 can readily be forgiven. To demythologize the situation, however, we suspect that Komisar was taken with the characterization of teaching as logically involving the intent to bring about learning. But that is a communion of a different sort from one that transubstantiates (necessarily) teaching resolve into success. (That would be the real miracle to explain if teaching actually did imply learning.)

The alleged mystery of Green’s analysis is far more profane, however. As we mentioned above, Green denies that teaching and learning are either logically or causally related. And though Green calls learning the “upshot” of teaching, he provides no further analysis of “upshottedness,” as Diet1 correctly notes. Green, we think, is wrong in denying the causal relation, but his analysis of the teaching/learning situation is by far the most perceptive in the literature. For he introduces, though shades of Socrates, the idea that the learner has more than a supporting role to play in the teaching/ learning interaction. It is an idea that philosophers of education and educational researchers have yet to appreciate.

Despite his discussion of teaching and learning within the task-achievement framework, Green ultimately avoids committing himself to that analysis. In contrast he states, “Furthermore, it may be a mistake to think of teaching and learning as task and achievement simpliciter, because teaching and learning are terms each of which can be used both in a task and achievement sense.”56 Several things are bothering Green here. First, he correctly understands that teaching and learning are not logically related. But he is also concerned about viewing learning as the achievement, causal effect, or product of teaching. He is troubled by this, because “if teaching and learning

53. Macmillan and Garrison, “Erotetics Revisited,” 358-59. 54. 6. Paul Komisar, “Teaching: Act and Enterprise,” in Concepts of Teaching, ed. C. J. B.

55. Green, The Activities of Teaching. 56. Ibid., 142.

Macmillan and Thomas W. Nelson (Chicago: Rand-McNally, 1968), 63-88.

VOLUME 37, NUMBER 3

Page 12: Teacher Accountability and the Causal Theory of Teaching

288 EDUCATIONAL THEORY

are causally related or productively related, then the failure to produce learning is the teacher’s re~ponsibility.”~~ Green recoils from the strong thesis of holding the teacher fully responsible for the student’s learning, since he clearly realizes that the student is implicated in the teaching/learning situation as well. In an important passage Green continues, “If teaching and learning . . . are to be understood as task and achievement, then they are the task and achievement of different persons. Perhaps we would be on better ground if we viewed teaching as the task of which getting someone to learn is the corresponding achievement, or learning is the achievement of which studying, investigating, or practicing is the corresponding task.”58

Green just about has it right here. What is missing from Macmillan and Garrison’s analysis, Dietl’s analysis, and even other task-achievement analyses, is the crucial activity of the learner in learning. On the Macmillan and Garrison erotetic analysis, for example, the learner is so far removed from the teacher/learner interaction that the role played by the learner is scarcely more than that of a passive recipient of the answer provided by the teacher. For once we take the teacher’s answering the questions to be the heart of the teaching/learning situation, it is difficult to see what else the learner might be. It is this aspect of their erotetic theory that bothers Pendlebury, and it is an aspect that bothers us as well. But once, as with Green, the learner’s actual situation (not merely some epistemological ought) is brought into the center of the picture, then the issue of teacher accountability is transformed. For if learning is in part the achievement of the learner, then not all hangs on the activities of the teacher.

But if Macmillan and Garrison’s analysis implies that learning is too much of a passive thing where one simply receives the teacher’s answers, one can go to the other extreme in thinking that learning itself is an action or activity.59 Now, obviously, one can learn without being taught. And even in teaching/learning situations, good students can generally overcome inept teaching performance by means of their own interaction with texts and subject matter materials. (This is why researchers on teaching must be as much concerned with what students are doing as they are with what teachers are doing.) But learning is not an action as such; rather, it is the achievement of the student’s acts and activities of attending, perceiving, studying, practicing, and investigating.60

Now it looks, however, that we have followed Green too far in decoupling teaching from learning. By placing much greater responsibility on the learner for learning, we may have succeeded in exonerating teachers from responsibility for the failure of students to learn; but to say that learning is merely an upshot of teaching, as Green does, we have also apparently reinforced Dietl’s and Macmillan and Garrison’s charge that this relationship is mysterious at best.

The answer, as we have suggested, is to decline to follow Green’s lead into those murky waters. To use Ennis’s terminology, teaching is a cause of learning, but only a partial cause. It is important to recognize that Green rules out a causal connection between teaching and learning on the grounds “that teaching can occur when learning does not.”6’ In other words, Green’s resf for causality is one of suficient-causation: if C (the purported cause) occurs and E (the purported effect) does not, then no causal relation exists between C and E. We agree, of course, that teaching and learning do not pass this sufficiency test for causal relations. Though Green is correct in thinking

57. Ibid., 143. 58. Ibid., 142. 59. Paul Hirst, in “What Is Teaching?” The Philosophy of Education, ed. R. S. Peters (Oxford:

Oxford University Press, 1973), 163-77, for example, calls learning an activity. 60. It has just been brought to our attention, though too late to be considered here, that

Gary D Fenstermacher, “Philosophy of Research on Teaching: Three Aspects,” in Handbook of Research on Teaching, 3rd ed., ed. Merlin C. Wittrock (New York: Macmillan, 1986), 37-49, has developed more thoroughly some of Green’s insights on the role of the student. Like Green, however, he believes that teaching and learning are causally unrelated. We hope to discuss Fensterrnacher’s views in a later paper.

61. Green, The Activities of Teaching, 140.

SUMMER 1987

Page 13: Teacher Accountability and the Causal Theory of Teaching

TEACHER ACCOUNTABILITY 289

that this means that there can be no genera/ recipe that empirical researchers can discover for producing learning,52 it doesn’t entail that teaching and learning are causally unrelated. The sufficiency test is too stringent for dealing with nearly all causal phenomena (and especially those in social and educational matters). It is Green’s use of the sufficiency notion of causation that creates the problem and leads him to conclude wrongly that there are no causal relations between teaching and learning.

The kind of causal relations that are most frequently to be found in the teaching/ learning situation are “INUS” and/or “INUP” causal relations. An INUS cause is (roughly) defined as an lnsuffificient but Nonredundant condition that is part of a set of conditions that is Unnecessary but Sufficient for its effect. An INUP cause, in contrast, is similar in every way with the exception that when conjoined with the other INUP conditions, it forms a set of conditions that merely confers a certain probability on the occurrence of the effectB3 The basic idea behind INUS (and INUP) causation when applied to teaching and learning is as follows. As everyone should know, teaching by itself is rarely, if ever, sufficient for learning. If the student fails to attend to the teaching, fails to practice, fails to study or do homework, etc., obviously the student has little chance of learning the subject matter. This gives insight into Dietl’s observation that even excellent teaching can fail to bring about learning. It also shows why the teacher’s answering the questions, even when appropriate and done correctly, hardly carries its own guarantee of success. Teaching is an INUS (or INUP) cause of learning. In the same way, the student’s attending (to the teaching), studying, practicing, etc., is seldom sufficient for learning. Though bright students, given appropriate background knowledge or native abilities, may master a subject matter or set of skills by themselves, without (or in spite of) the teaching, in most academic contexts positive student efforts are not enough. But student efforts in interaction with the teacher’s guidance, clarifications, and explanations, etc., may be sufficient to produce learning (may because in some contexts or in dealing with some subjects, even the greatest efforts and the best teaching may fail short - this is indicative of lNUP causation, not INUS). Typically, then, student efforts and teaching activities are INUS (or INUP) causes of student learning. Thus, Green is less than correct in suggesting that learning is the achievement of the learner, where learning is merely the (mysterious) upshot of teaching. Student learning, most often, is the joint achievement of the learner and teacher.

Given this way of clarifying the task-achievement analysis of the teaching/learning situation, we have at once a nonrnysterious causal theory of teaching that does justice to both teacher and learner; to both the activities of studying and activities of teaching in the production of learning. At the same time, it is a theory that allows us to explain why researchers seeking invariant recipelike generalizations relating teaching behaviors to student achievement have been singularly unsuccessful.@ In these respects, the

62. Ibid., 141. 63. These definitions and distinctions between sufficiency causation, INUS causation, and

pure probabilistic causation are treated in depth in Ellett and Ericson, “Causal Laws and Laws of Association,” “Correlation, Partial Correlation, and Causation,” and “An Analysis of Probabilistic Causation in Dichotomous Structures.” INUS causation was first formally introduced by J. L. Mackie in “Causes and Conditions,” American Philosophical Quarterly 2 (1 965): 245-64, though he was anticipated by the sociologist/philosopher of science Stephan Nowak in “Some Problems of Causal Interpretation of Statistical Relationships,” Philosophy of Science, 27 (1 960): 23-38. We are responsible for the rather obvious extension of INUP causation; we think that it is most likely this notion that captures the flavor of causal relations in teaching.

64. INUS and INUP conceptions of causality are interactive in nature. As a result they are ideal for capturing the interactions between teacher, learner, and subject matter inherent in the teachingpearning situation. They contrast clearly, then, with sufficiency and probabilistic notions of causation. Because educational researchers typically seem to assume either sufficiency or probabilistic causal relations in their investigations, they have an exceedingly difficult time in capturing the nonlinear interactions in teaching and learning. Lee J. Cronbach’s essay, “Beyond the Two Disciplines of Scientific Psychology,” The American Psychologist 12, no. 11 (1975): 116- 27, on Aptitude X Treatment Interaction research is indicative of the difficulties of forcing interactive phenomena into sets of linear relations. Now it is true that the interactions between teachers and

VOLUME 37, NUMBER 3

Page 14: Teacher Accountability and the Causal Theory of Teaching

290 EDUCATIONAL THEORY

INUS (or INUP) causal theory of teaching seems to provide a more adequate conceptual framework for the empirical investigation of the teaching/learning situation than alter- native formulations. And lastly, such a theory permits the foundation for the assignments of credit and blame in just those cases where credit and blame are due. To that final task we now turn.

I v.

Earlier we examined Macmillan and Garrison's example of the swimming instructor who was blamed for failing to reach the student just because the student did not attend the in~t ruct ions.~~ It was our intuitive response to mention that things had gone awry with such a theory that holds the teacher responsible. Any theory that encourages that counterintuitive inference, as all theories claiming a logical relation or sufficient causal relation between the phenomena of teaching and learning do, seemed on the face of it false to educational phenomena. But the kind of causal theory that we propose here now explicitly enables us to explain the fallacious nature of that inference.

Let us say that one's teaching has been successful when the student has learned the intended material; we shall say that one's teaching has been unsuccessful when the student has not learned the material. To judge the success of a teaching activity is to focus primarily on whether the student learns the material.

But there is another way in which a teacher's activities can be judged which is independent of the student-success criteria. Let us say that one's teaching has been good in so far as the teaching reflects an adequate understanding of the material and provides a reasonable instructional sequence that is appropriate to the students' learning level. We shall say that one's teaching is bad (or a failure) in so far as it reflects an inadequate understanding of the material and fails to provide a reasonable instructional sequence.

The important point is that lack of success in teaching does not entail that the teaching is bad (or a failure), for it is possible that the teaching be good, but for all that the student does not learn. Also, it can be seen that failure in teaching does not entail lack of success in teaching, for it may be that the teacher is failing to perform the right activities in the right ways, but the student learns the intended material anyhow.

We have a conception of teaching, therefore, which differs in important ways from the Macmillan-Garrison conception. In our view, teaching is that activity that aims to help the student learn the intended material by means of providing a reasonable and appropriate instructional sequence. Not only must a teacher regard the instructional sequence as being conducive to the student's learning; more importantly, in our causal theory of teaching, the primary focus of the evaluation is whether the means selected by the teacher are conducive to student learning. Given this conception of teaching, good teaching does not entail that every single student learns the intended material. We believe it can even be shown that good teaching does not even entail that a specific proportion of the students in the class will learn the material. As Max Black and Michael Scriven have pointed out,66 the determination of whether a set of teaching activities is good (or adequate) essentially involves a comparison with the alternative approaches and methods for teaching the material. The essential feature of good teaching is not some specific success rate, but the fact that it is comparatively better at bringing about learning than its alternatives. And this comparative analysis will use the concept of the proper reference class to carry out the inquiry. The specification of the reference class

learners become exceedingly complex and difficult to research. But unless researchers adopt the INUS and INUP notions, we believe that there is not much likelihood of progress here.

65. Macmillan and Garrison, "Erotetics Revisited," 359. 66. Max Black, "Reasonableness," in Education and the Development of Reason, ed. R. F.

Dearden, P. H. Hirst, and R. S. Peters (London: Routledge & Kegan Paul, 1972) 194-207; and Michael Scriven, "The Methodology of Evaluation," Principles of Curriculum Evaluation, ed. R. Tyler and M. Scriven (Chicago: Rand-McNally, 1967); 39-83.

SUMMER 1987

Page 15: Teacher Accountability and the Causal Theory of Teaching

TEACHER ACCOUNTABILITY 291

will involve the characterization of the level and range of the class members’ information and cognitive abilities as well as the range of motivation.

Thus, our theory, unlike the others, permits us to make a strong distinction between “failure in teaching” and a “lack of success in teaching” (i.e., no learning) and to draw this distinction on grounds infernal to the teaching/learning situation. While “failure in teaching” does imply that the teacher can be held accountable for his or her performance, “lack of success in teaching” does not. Rather, it merely means that the aim of teaching has been thwarted, perhaps, by circumstances outside of the teacher’s control.

Now of circumstances beyond the teacher’s control, there are two kinds: (1) those external to the teaching-learning situation and (2) those internal to it. The former range from the mundane, such as student leg cramps in swimming, to the sublime, such as acts of God. These external interruptions, of course, can limit teacher responsibility even on the most ardent logical or causal sufficiency thesis. However, we maintain that internal circumstances can also absolve the teacher of responsibility. Of particular importance are those circumstances related to the aspects and features of the learner. For example, if teachers were to make a reasonable effort to ensure that students are attending to the teaching, but the students were to refuse to concentrate, then the subsequent student failure to learn would probably not be the teacher’s fault. There are situations in which it may well be the student’s fault. Similar remarks can be made about the student’s failure to study, to investigate, to practice, or to do assigned homework. Notice that, according to our view, the teaching/learning situation itself extends well beyond the activities and processes of the classroom.

In some ways our view is similar to a view advanced by Israel Scheffler. Scheffler holds that

to teach, in the standard sense, is at some points at least to submit oneself to the understanding and independent judgment of the pupil, to his demand for reasons, to his sense of what constitutes an adequate explana- tion. . . . Teaching, in this way, requires us to reveal our reasons to the students and, by so doing, to submit them to his evaluation and criticism. . . . To teach is thus, in the standard use of the term, to acknowledge the “reason” of the pupil, i.e., his demand for and judgment of reasons, even though such demands are not uniformly appropriate at every phase of the teaching interval.67

It can be seen that Scheffler’s view also maintains the distinction between “successful” teaching and “good” teaching.68 Notice, however, that although Scheffler puts forward a quite specific view of the teacher’s responsibilities, he fails to discuss what the student’s responsibilities are. Though we can do hardly better here, we can in a few final words mark out this area as an important one for further inquiry.

That there has been a strong tendency by philosophers, educational researchers, and the public to focus on only what goes on in the classroom and especially on what the teacher does, shows how narrow is their understanding of the educational context. The limited nature of this focus may well be responsible for ever-increasing “methods” of classroom-management and student-motivational techniques in which one makes the teachers responsible for mastering. But such a focus directs our attention away from the important fact that teaching, even the best teaching, cannot by itself bring about learning. Student efforts (at various activities) are almost always required too. In many contexts, students should be seen as agents of their own learning; they should take on their proper share of the responsibility. It does a gross injustice to point a blaming finger only at teachers, especially for that which is beyond their control.

67. Israel Scheffler, “Educational Metaphors,” in his The Language of Teaching, (Springfield, 111.: Charles C. Thomas, 1960). 57.

68. Though Scheffler in the above quote seems to be offering an analysis of ‘teaching’, it has become clear over the years that he is advocating an ideal, a view of what good teaching amounts to.

VOLUME 37, NUMBER 3

Page 16: Teacher Accountability and the Causal Theory of Teaching

292 EDUCATIONAL THEORY

Many teachers, we think, have known these things for years. They can recite anecdote after anecdote of students who could not care less, of parents who are unbothered or not even there (and surely it is parents who must be held responsible for the actions of the younger children),69 and of administrators who are unsupportive. It sometimes seems that only philosophers of education, educational researchers, and educational “statesmen” with their plethora of reforms to transform the teaching profession are unaware. What is required, then, beyond our conceptual analysis of the teaching/learning situation? Obviously, we require improved empirical research designs that take into account the cocentral position of the activities of the learner - in and out of the classroom. But an ethical theory is also required to determine, given a suitable factual basis, what the responsibilities of the teacher and learners are. Even if the teacher has the responsibility to help bring about certain kinds of learning, in certain contexts there may be good reasons for the student to bear a parallel responsibility to effect his (or her) own learning.’O Although we cannot go into the details here, we believe it can be shown that in almost all educational situations, the students, as well as the teacher, have certain responsibilities for learning.

A final note is in order on those now trumpeting educational reform by pointing to the “sorry” state of the teaching profession. No doubt a number of excellent measures can be taken to strengthen the art and practice of teaching in our classrooms. Though we have major reservations about identifying and removing the ablest teachers from the classroom in order to make quasi-administrative instructional leaders out of them (nearly tantamount to removing the best surgeons from their practices), freeing up teacher time and some form of differential career teaching ladder could be very beneficial. It is difficult to see, however, how continued efforts to denigrate unjustly the performance of the present cadre of teachers can aid in strengthening the profession or enhancing its public image.

There are some, nevertheless, who (cynically?) suspect that the teacher account- ability movement is something of a smoke screen, merely a means to advance other interests that lie beyond teacher reform itself. In this case, teachers are simply playing their assigned role of convenient scapegoat. If so, this gives an indication of the depths to which our current modes and terms of public discourse have plunged. And if so, it is certainly something far more troubling and well beyond the power of philosophy, or any other human inquiry, to cure.

But the fact remains that it is clearly time to quell the clamor for holding teachers mainly responsible for all of our educational ills. Teachers should be held properly accountable. But we maintain that they should be held accountable only for that which is within their power to control. Thus, one might more profitably review the responsibilities of boards of education and state legislatures who have the power to increase educational funding, lower class sizes, and otherwise create external conditions conducive to success. And, perhaps most importantly, we might more profitably urge that students (and parents) recognize and accept the responsibilities that are uniquely their own in ensuring that students become active agents of their own learning. Should we approach education in these ways, we may well attain a better understanding of the various

69. Such a view should be tempered, however, by the realization that some parents may be so crushed by the nature of their own circumstances that they are in no position to urge their children on. A similar point may be made about some children, too.

70. It is important, too, to note how class size plays a role here. The empirical literature on the relation of class size to educational performance has yielded equivocal results. To our knowledge, however. student attention and effort have not been adequately controlled for in these studies. Common sense suggests, however, that the smaller the class, the easier it is for teachers to determine and monitor student attentiveness and effort. Though, obviously, the teacher cannot compel such behavior, it seems reasonable on the whole to hold teachers less accountable for student achievement in larger classes than in smaller ones. For this, consider large college lecture classes in which students are mainly held responsible for their own learning.

SUMMER 1987

Page 17: Teacher Accountability and the Causal Theory of Teaching

TEACHER ACCOUNTABILITY 293

responsibilities in education and of the roles of those who contribute to carrying them out.

The authors contributed equally to this manuscript and, therefore, are equally accountable for its fortunes. We would like to thank the members of the California Association for Philosophy of Education and George F. Kneller for their helpful comments.

VOLUME 37, NUMBER 3