WHEN PRESIDENT Charles W. Eliot explored ways the University could improve secondary education in the 1890s, his goals were few and simple. To guarantee Harvard a large, well-qualified pool of applicants, he led a small recruiting campaign, helping high school instructors shepherd young men past the imposing gates of the Yard. But as Eliot's cause expanded, eventually inspiring the establishment of the Graduate School of Education in 1920, skeptics repeatedly challenged the movement, questioning the suitability of secondary education as a university concern.

Arthur G. Powell confronted the legacy of Eliot and his predecessors while serving as associate dean of the Ed School from 1968 to 1976. In his history of the institution, The Uncertain Profession: Harvard and the Search for Educational Authority, Powell chronicles the metamorphosis of the "Division of Education," which limped through the first 20 years of the century under the burden of unending criticism, into a respected graduate school. The saga, one of bitter disagreement and jarring philosophical gyrations, provides a sober backdrop for the current dissatisfaction with efforts to improve American education.

Although Powell's chronology is often tedious, his concern with detail emphasizes the book's most alarming point: the Ed School never established a clear path for itself after the initial transition from Eliot's movement for secondary school reform to the acceptance of responsibility for improving the entire public school system.

After tracing the growing tension between the research and professional training functions of the infant school, the author methodically presents a half century of programmatic jousting. Highlights include Henry W. Holmes's failure to nurture an elite corps of super-educators in the 1920s, followed by Frank E. Spaulding's crusade for a scholarly research revival. Francis Keppel replaced Spaulding and led a charge back in the other direction. Programs that trained teachers and administrators again tool their place next to faculty research projects, a compromise Powell says "caused substantial friction and unhappiness." Despite its inherent instability, the Keppel combination survived the 1950s and struggled into the 1960s, fueled by free-flowing federal dollars.

BY THIS POINT in his narrative, Powell has established the disturbing rhythm of change at the Ed School. His description of the clumsy attempt to bolster state and federal agencies in the campaign against urban decay does not surprise us. Powell remembers that "an emphasis on educational jobs distant from children reemerged. Unlike earlier times, many of these jobs were located outside school systems or schools of education."

Finally at the end of this sluggish carousel ride, we appreciate the author's dogged pursuit of a common thread with which to bind the many tasks Harvard educators have undertaken. His disgruntled conclusion:

These various notions of education's tasks overlapped and frequently worked at cross-purposes. All were typically pursued without any clear sense of whether they were being met. Decisions to intensify or retreat from a task were not made on the basis of whether the objective seemed nearer or farther away, but on the basis of suddenly conflicting new tasks. Problems were rarely solved; they were survived.

Although deep down Powell may have wanted to end his work with this blunt indictment, he allows himself to revert to "Edthink" in two final paragraphs of qualification and apparent apology to his former employers. He praises "Harvard's investment in education" for providing self-consciousness, but not self-confidence, to the profession, implying that this contribution makes up for the 60 years of indecision.

Yet, leaders of the Ed School have always claimed that their institution is an innovator, not merely an expensive chalkboard on which theorists and practitioners compete for space.

Nevertheless, the facts Powell relates far out-weigh his gratuitous sympathy, and the timeliness of his observations is affirmed by the school's latest readjustment, a move away from the support of agency-designed education, back toward more direct contact with individual institutions and "school leaders." Like so many of the realignments Powell describes, the current shift seems to correct an imbalance which was partially caused by uncontrollable social and economic conditions. But when viewed with the persepective furnished by this book, such circumstances cannot hide a reality that professors in Longfellow Hall have not fully addressed: in Powell's words, "The School never developed a clear sense of how educational knowledge was produced."