Harkness Learning: Principles of a Radical American Pedagogy

Guy J. Williams, Philosophy and Religion, Wellington College

Contact: gjw@wellingtoncollege.org.uk

Keywords: Harkness, pedagogy, dialogue, Socratic dialogue, learning, American, education, problem-based learning, pragmatism


This paper investigates and argues for Harkness learning: an approach to education that inculcates a culture of enquiry, driven by students in dialogue around a table. Tracing the history of the Harkness reforms in education at Phillips Exeter Academy, their pedagogical and philosophical roots are considered. Although partly inspired by the Oxford tutorial system and the Socratic concept of dialogue, Harkness departs from there to a radical classroom dynamic. The teacher is required to be more open-minded and less controlling over outcomes, to take the risk of listening more and saying less. This shift in emphasis fits with a distinctly American philosophical respect for the sanctity of the free thought and originality of the individual, which here is traced to Transcendentalist ideals that have permeated American culture. It can also be compared with Problem-Based Learning and there are certain significant overlaps between these pedagogies. However, Harkness is sometimes narrowly misunderstood as a technique for teaching, which takes the approach out of context. It needs to be understood in the broad frame of cultural reform of an institution – it is a useful symbol for a community committed to student discourse and problem solving. In this respect, Harkness owes a further debt to pragmatism – another great American philosophical-educational tradition.


I would make education a pleasant thing both to the teacher and the scholar. This discipline, which we allow to be the end of life, should not be one thing in the schoolroom, and another in the street. We should seek to be fellow students with the pupil, and should learn of, as well as with him, if we would be most helpful to him (Thoreau, 1906 [1837]).

In the United States, specifically in New England, there is a well-known and celebrated intellectual culture of individualism, freedom of conscience, and (theoretically, at least) egalitarianism. It goes right back to the Pilgrim Fathers, in their flight from what they saw as a corrupt monarchy, government, and Church of England. Later on, in the nineteenth century, Henry David Thoreau, famed for his self-sufficient and slightly reclusive flight to Walden Pond, also put education within a radical frame, seeing the position of the student and teacher in terms of a journey shared. Even in 1837, Thoreau was talking about learning of and with a student, to be of help in his studies. The kind of agency this assumes is striking in term of getting alongside the learner and being part of (not dictating) the process. Similar sentiments were raised by Thoreau's Transcendentalist colleague Ralph Waldo Emerson in his contention that 'the secret of education lies in respecting the pupil.' (Emerson, 1883 [1863], p. 143). In their writings on education, both Thoreau and Emerson opposed institutional control, formality, and the instrumental view of learning as a means of gaining credentials. Weary of the drudgery of conventional schooling, they wanted to provide for an autonomous learner. Such ideas were surely ahead of their time in America, as indeed they would have been in Britain. Today, however, they serve as helpful reminders that concepts like student-focused learning have long histories way beyond modern pedagogy.

One striking product of American educational ideals has been the Harkness table, as first adopted by Phillips Exeter Academy (PEA) – the apparatus symbolic of collaborative learning that is driven by students, not teachers. Classes are characterised by a small group (typically, twelve) gathered around a seminar table for discussion and problem solving. Although this approach has been implemented and promoted consistently by PEA for nearly one hundred years, it has quite recently become the subject of growing interest for schools all around the world, including the UK [1]. Its supporters would say that Harkness could become a fresh paradigm for learning in the secondary sector and beyond. I would like to contribute to this current and important conversation here by setting out some of the pedagogical and philosophical principles from which this approach to learning has developed.

Initially, I would like to give Harkness quite a broad definition. It certainly is not new, and I would contend that it is not really a method per se either. Harkness is found in forms of education that are committed to maximal student preparation, problem solving, discussion, enquiry, and a diminished role for the teacher's authority or ego. Harkness learning is well symbolised in the classroom layout of the Harkness table, but the physical artefact itself does not guarantee any particular learning style. So, as I will argue here, an understanding of Harkness requires more than a furniture supplier – it requires a consideration of the educational culture or ethos where it is employed. Because I think it has value as a pragmatic approach above all else, I would also advocate that discussions move beyond simply talking about 'Harkness lessons' towards inculcating a school-wide 'Harkness culture'. It could be a kind of 'central metaphor' (Heskel and Davis, 2008, vii) by which a school could orient its pedagogy.

History and development

Harkness learning takes its name from the American businessman, philanthropist and educational reformer of the early 20th century, Edward Harkness. As an alumnus of St Paul's School (a New Hampshire rival to PEA), he felt that his own educational experience at school had been restricting and not suited to his needs. Students sat individually in a front-facing desk arrangement, attended to instruction ('recitations') from a teacher on a raised platform at the front of the room. It was a competitive environment, which was affirming only for those with quick answers to questions. Whether by accident or design, Harkness felt that this arrangement created a hierarchy which did not favour those with less confidence or who wanted more support (and he bracketed himself in this group). This pedagogy could be compared with that used in the schools of Victorian and Edwardian England and, in fact, the same style of lesson was also used at PEA at that time (i.e. lectures, with spot questions for students). It was even customary for students to click their fingers loudly when another student faltered, to show that they had a ready answer (Towler, 2006).

Due to a friendship with its then Principal Lewis Perry, Harkness – an heir to a multi-million dollar fortune – set upon PEA as the possible recipient of a major donation. Tasked with making a proposal for how the funds could be used, a group of PEA faculty took a tour of 'great schools' in Britain (including my own current employer) to investigate best practice. The results of this fact finding mission were varied proposals, including an enhanced tutorial system, reorganisation of residential houses, visiting lecturers, ability sets, and new methods of science teaching. Harkness rejected the initial plan on the grounds that it suggested mere enhancements, insufficiently radical, and did not contain the kind of educational revolution that he had been hoping to fund. He was looking for a complete overhaul in teaching (Heskel and Davis, 2008, 3-4).

Harkness favoured what he called a 'conference' pedagogy, which bore some relation to the tutorial system of small group discussion teaching in Oxford. He hoped to see groups of about ten students, gathered around a table, and engaged much more in conversation than in recitations. The PEA faculty then worked on his counter-proposal and agreed to the wholesale change in teaching at the school. The funding for this new approach totalled in excess of five-million dollars and instigated the construction of new classroom blocks, facilities, and of course the famous tables. Harkness was thus advocating (and getting) what he described as changes 'of a fundamental nature' but was far less successful in pursuit of his wider hope that 'the whole educational system in our Secondary Schools would be changed' ([letter to Perry] Heskel and Davis, 2008, 3). Thereafter, all learning at PEA took place within the new Harkness framework, and many other American boarding schools in due course followed suit.

In essence, therefore, against its founder's original intentions, the Harkness conferences became a pedagogy for elite schools only, and that is largely where this approach has remained until recently. Clearly, Harkness hoped that children from all backgrounds would in time come to benefit from an approach of small-group collaborative enquiry. After all, what is more important, or a better investment than education? In this aspiration, it could be argued that Harkness was unrealistic – a criticism for a later discussion – because the most obvious objection to Harkness learning is its apparently prohibitive cost in class sizes of ten or twelve. Nevertheless, putting that problem aside for time being (and I would argue that there are practical workarounds), it seems undeniable that Harkness instigated a pedagogy that has endured at and enriched some of the world's most successful schools.

A radical pedagogy

Despite the radical intentions of its founder, the actual practice of learning and teaching around a Harkness table developed only gradually over a long period of time. The new approach was not invented in complete form, with a coherent theory of how precisely the student and teacher should interact. The Harkness philosophy originated very simply from observing the benefits of small group teaching, but ended up taking a departure from, for example, the tutorial system at Oxford. The implicit pedagogy of the ancient Oxbridge system (which has never been formally codified) lies in the foundational assumptions that (a) dialogue is good for learning, (b) personal provision is good for learning, (c) it's easier to ask questions in small groups, and (d) conversation with a subject authority allows for the acquisition of knowledge. In that sense, the Oxbridge system could be seen as a form of Socratic conversation (see Lane Fox, 2008, 55-60). Ultimately, however, despite substantial overlaps with Oxbridge tutoring, Harkness learning has moved in a different direction, in no small part because it has drawn deeply from radical American pedagogical thinking. Suspicious of authority and submission to expert opinion, Harkness learning now probably owes more to Emersonian concepts of self-realisation and self-reliance than it does to the British interpretations of Socratic pedagogy that may have first instigated it.

Going back to Emerson's ideal of respecting the pupil, this concept forms the basis and title of a classic pedagogy text produced by PEA, which attempts to explain the principles and practice of Harkness (Cole and Cornell, 1981). It explains that there is a balance to be struck between not stifling the authentic voice of the student and insisting upon rigorous content: as Emerson put it, respect for the pupil and respect for oneself as a teacher. So, in practice, Harkness learning is not madly liberal, allowing students to say whatever they want about whatever topic they want, whenever they want. It works within a disciplined framework with prescribed content, high expectations for reading and writing, and a demanding pace for course progression. So, the framework is tough and strictly delineated, but what happens within that framework is truly open-ended. Teacher planning is for possibilities, not outcomes, in that Harkness learning simply enables students to have an intellectual conversation. It is planned that all the students will have to read a certain text, or write a certain paper, but what then happens around the table really comes from them.

The freedom of the student is therefore a central concern, albeit within a well-defined structure. It is a very high degree of freedom in educational terms, in the sense that the students have genuine control and are not driven towards a pre-conceived idea of what 'good' understanding is. The underlying assumption for this seems to be that it is important to make room for the authentic voice of the student, to provide encouragement, build confidence, and to embed positive attitudes to learning. In this regard, it seems to mirror the Transcendental individualism of Emerson's essay 'Education':

The great object of Education should be commensurate with the object of life. It should be a moral one; to teach self-trust: to inspire the youthful man with an interest in himself; with a curiosity touching his own nature; to acquaint him with the resources of his mind, and to teach him that there is all his strength, and to inflame him with a piety towards the Grand Mind in which he lives … I believe that our own experience instructs us that the secret of Education lies in respecting the pupil. It is not for you to choose what he shall know, what he shall do. It is chosen and foreordained, and he only holds the key to his own secret. By your tampering and thwarting and too much governing he may be hindered from his end and kept out of his own. Respect the child. Wait and see the new product of Nature. Nature loves analogies, but not repetitions. Respect the child. Be not too much his parent. Trespass not on his solitude (Emerson, 1883 [1863], p. 135, 143).

However, a full Transcendental revolution of education would be radical indeed, and both Emerson and Thoreau had deep suspicions of systematising and formal education – it 'makes a straight-cut ditch out of a free, meandering brook' (Thoreau, (1906) [1850]). Judging them harshly, it could be said that they held romantic ideas that lacked practical application. In 'Education' Emerson goes on to associate much schooling with mindless machinery, criticises mass education, and advocates educating children 'one by one … the whole world is needed for the tuition of each pupil.' (Emerson, 1883 [1863], p. 153). Nevertheless, there is arguably something truly valuable and interesting in the insistence upon an ethic of (paradoxically) teaching self-trust. The established educational idea of measuring outcomes contains an accidental determinism, because teachers decide ahead of time what they are looking for: Thoreau's straight-cut ditch. Why should a teacher keep checking for a pre-conceived measure of knowledge and understanding, when a pupil is supposed to be on a voyage of self-discovery? If the voyage is in evidence, why should the destination matter?

And so back to Harkness. Like an Oxford tutorial it assumes that (a) dialogue is good for learning, (b) personal provision is good for learning, and (c) that it is easier to ask questions in small groups. However, as the approach has developed in secondary sector teaching, the role of the teacher as a subject authority is not the same. Teachers have the expertise to demand the right processes and frameworks, but the discussion does not belong to them. The pedagogical philosophy of Harkness lies in a shift in the balance of power, which even separates it from the Socratic concept of dialogue.

Harkness dialogue and Socratic dialogue

Superficially, Harkness seems to be another incarnation of Socratic dialogue and indeed is sometimes referred to as 'Socratic' by its practitioners. The teacher is present to coordinate truth seeking in the conversation, and to ensure that the knowledge is brought out from the students through intelligent questions. The teacher is a facilitator and mediator of learning. This would seem to correspond to a central metaphor of self-understanding, that Socrates is the 'midwife' of ideas who enables the students to give them birth (as in the Theaetetus and Symposium). On the one hand, there is a problem with this on the side of Plato scholarship – the actual 'Socratic' method seems to vary greatly in the dialogues from genuine conversation to blatant lecturing (see Lane-Fox, 2008, 56). What is the authentic Socratic approach? A much discussed problem. Yet, on the other hand, there is a further problem of identifying Harkness dialogue with a Socratic approach, because even in considering the more interactive mode of conversation in the early dialogues, the comparison is not apt. Quite simply, a Harkness teacher has to give up on the authority and control of a midwife.

Even in genuine discussion with friends, the Socrates of Platonic dialogues is the dominant interlocutor, this presumably being a joint product of Plato's reverence for his mentor and his desire to use the well-known teacher as a mouthpiece for his own ideas. Also, given the strong theme of Platonic epistemology running through many of the dialogues – that knowledge is intrinsic in the soul and must be recollected from a past life – the Socratic metaphor of the midwife connects with a kind of objectivism: the 'right' answer is in there somewhere. The centrality of the teacher in this model contrasts with Harkness learning, because such classes are not like Platonic dialogues, venerating the master. This distinction is neatly summarised by Peter Vorkink, of the Religion department at PEA:

As generally understood, Socratic teaching involves the teacher asking a series of questions and the student(s) refining his/her responses—with the teacher continually pointing out weaknesses in the student's (s') ideas until the student(s) 'realizes' the right answer … teaching via open-ended questions [i.e. Harkness] does look like Socratic teaching, but in the end the real questioning is going on not teacher-to-student but student-to-student (Vorkink, 2013, 40).

This can be neatly summarised in the following diagrams, in which the lines represent movement from one speaker around the table to another:

Socratic dialogue with primary interlocutor Harkness dialogue

The obvious objection to this, and a point in favour of a classically Socratic approach, is that student-to-student interaction lacks the same kind of quality control that is brought about by a true philosophical midwife – an expert authority. What is there to stop the students stumbling down a blind alley of educational relativism? Are no points or perspectives more valuable than others? Socrates, being the dominant interlocutor, would always correct the reasoning of his young associates, and this would fit with the established role of the teacher.

However, there are possible responses to this objection. It assumes a certain epistemological goal in education which could be challenged, centred on imparting truths, what Pritchard calls the 'historically popular' assumption 'that we merely want our children to acquire lots of true beliefs' (Pritchard, 2013, 236). The problem with this is that it mismatches with our most important goal for education, that we 'provide children with the cognitive skills to be able to determine truths for themselves' (Pritchard, 2013, 237). Is it really the case that the fear of epistemological error should trump the cognitive agency of students? Is it such a problem if they publicly air weak reasoning or false claims? Taking a virtue perspective on the matter, a good case could be made for seeing the 'truth' of education as the experience of discovery, shaping ideas with one's peers, rather than being directed – however gently – towards them. There are also practical ways out from the accusation of unchecked relativism: inculcating good fact-checking routines, setting reading that could address misconceptions, and teaching the subtle art of polite disagreement. The teacher may not be king in Harkness pedagogy, but that does not make him delinquent.

Comparison with problem-based learning

Harkness has a clear dialogical structure and orientation. More than that, though, it is way of prioritising educational goals towards tackling problems in free and imaginative ways, and this makes a connection with a wider array of approaches. As a student-centred pedagogy with North American origins, Harkness has to an extent a parallel history with problem-based learning (PBL), which has originated and spread from various medical schools and HE institutions. A brief comparison here can help to elucidate some of the practical features and advantages of Harkness as an adaptable pedagogy that inculcates skills.

PBL in fact encompasses a variety of practices, though its core is a simple idea: the learning process is enhanced by setting problems for students, so that the achievement of a practical goal necessitates independent research and discussion within a team (for a comprehensive review and clarification, see Savin-Baden and Major 2004). So, for example, medical students might be presented with a case of an individual displaying certain symptoms that could be indicative of various diagnoses. The case works as a stimulus, because it is sufficiently ill-defined that it requires deeper examination. In other words, medical students have to do what doctors do; they have to work out what they are dealing with and how to respond by asking the right questions and accessing the right resources. It is an active and constructive process, drawing on contextual and social considerations.

How does that compare with the Harkness, 'conference' dialogue? Firstly, both approaches are process-driven, in the sense that priority is given to the how of learning, rather than the what. In PBL, students have to learn how to learn, because for example the rapid advance of medicine makes it impossible for any one individual to hold all the necessary knowledge at a given moment. What matters, therefore, is the development of a practitioner who knows how to find things out efficiently and accurately, and then who is able to test ideas before applying them (e.g. in treatment). So too with Harkness, the most important hurdles for students are learning how to prepare for the discussions and then how to act in the discussions. This also links to the second connection, in that both approaches are social models of learning. Although instructor-to-student interaction retains value, in both PBL and Harkness peer-to-peer interaction is the most decisive form of engagement. Dialogue with peers is the required method for solving the problems and, looking long-term, the wider field of society and employment puts great weight upon intellectual and professional dialogue. Why not start this at university or at school? Finally, there is a third area of overlap in the cyclical structure of problem solving. Dialogue helps to construct problems; it phrases and clarifies them in a common language. Dialogue also proposes solutions and verbally tests them among peers. It is this element of discussion that re-constructs problems or replaces them with better problems. Both Harkness and PBL are concerned with the refinement of ideas in a social setting, and in that sense could be interchangeable in certain situations. PBL could be implemented through Harkness dialogue. Harkness could be problem-based in the way it is set up.

Harkness method Vs. Harkness culture

Having now set out the theory, philosophy and history, what should we make of it? All of the points above could be seen as expounding a classroom methodology – there is a 'way' to 'teach' Harkness lessons through subtle encouragement of the students to dialogue with one another. There may be some truth in that, in the sense that experience of such lessons makes the teacher a skilful listener, able to move things along without intruding on the conversation. Having said that, I would argue that labelling Harkness as a teaching method can be counter-productive and misleading on other levels. It implies that the teacher can create Harkness as a style of lesson – another tool in the box – that can be deployed as and when it seems suitable. Yet, this is certainly mistaken, because teachers don't have the power to impose student leadership on occasional lessons in an otherwise didactic pedagogical culture. It involves fundamental beliefs about learning, about the agency of learning. The idea that Harkness lessons can be conceptually borrowed as a form of novel academic extension therefore involves a certain absurdity.

Going back to the hopes and expectations of Harkness himself, he anticipated a 'conference' style of learning, specifically to build the confidence of learners, rather than catch them out on knowledge recall. There is something fundamentally humane about that, because his ideas started out with development of young people in mind, while so many educational reformers look to content first. Thus, there is a deeper question about employing Harkness to do with the cultural fit in an institution. If there is a culture in which students 'have to' know the 'right' things to 'succeed' then lessons will naturally be risk-averse and paternalistic – the teacher 'helps' students to 'get through' school. It would be possible to purchase oval shaped tables for such a school, but they would surely have no impact without a wider examination of the learning culture. Isolated from an appropriate context, a method of Harkness lessons would be ineffective.

This brings us back to the assertion of PEA teachers that Harkness is 'central metaphor' for the school's learning culture (Heskel and Davis, 2008, vii). It is a way of putting the intellectual development of young people at the heart of the school – nothing more, nothing less. It is not a technique, or even an ideal, such that would need to be observed in a pristine state. Institutional symbols are much more pragmatic than that, and they can be moulded to the varied changes in educational, social and technological circumstances that may arise.

The idea of reforming a learning culture, rather than technicalities, is a way of seeing education in a much more fluid state. Despite the challenges of implementing something as radical as Harkness, I would argue that it is worth the risk, to see how education could be of use to the student's own development, on the student's own terms. That is how the Transcendentalists saw education: a practical and collaborative enterprise, and one that draws out the inner voice. As a genuine programme for reforming education, however, we must look further along in pedagogical philosophy – finally, in this paper – and acknowledge the proposals of the pragmatist school, and of John Dewey particularly. This provides, I think, a meaningful way of defining the way Harkness (or other pedagogies like it) could be introduced into the wider discussion of our educational priorities.

Conclusion: Harkness as a social, pragmatic symbol

Another great intellectual tradition of North America, pragmatism, sees knowledge itself as a fluid reflection of what is expedient. Where epistemology leads, education follows, and so there is no perceived value in simply handing down the supposed timeless truths. If what we regard as knowledge is simply the best fit with our social systems, then education becomes an adaptive enterprise of allowing students to develop capacities that are suitable to their time and place.

Foremost among pragmatic commentators on education is undoubtedly John Dewey, who could reasonably be described as an educator as much as a philosopher (in fact, his gravestone describes him precisely thus). It is not possible to survey his vast contributions to pedagogy here (for a helpful overview: Garrison, 1999), but there is a particular conception of the school and the teacher in Dewey that ties the themes of this paper together, and makes sense of Harkness as a powerful symbol for educational institutions:

I believe that the school is primarily a social institution. Education being a social process, the school is simply that form of community life in which all those agencies are concentrated that will be most effective in bringing the child to share in the inherited resources of the race, and to use his own powers for social ends ...The teacher is not in the school to impose certain ideas or to form certain habits in the child, but is there as a member of the community to select the influences which shall affect the child and to assist him in properly responding to these. Thus the teacher becomes a partner in the learning process, guiding students to independently discover meaning within the subject area (Dewey, 1897, article II).

Dewey's own ideas for the application of such philosophy encompassed the foundation of a school with a strong focus on practical tasks and manual skills (in the University of Chicago Laboratory School), and they have been a forming influence over PBL. However, applying his words to Harkness, his conception of the school and the teacher fits with and indeed concludes the argument of this paper. Learning is a social process that occurs within the specific culture of an institution. The teacher is a representative of a community that has certain values or assumptions about what learning is. The goal of learning is the use of a student's powers for social ends. Therefore, a school needs to have a way of representing its understanding of learning to itself: a social and pragmatic symbol. That is what Harkness achieves so effectively.


  • Cole, D. and Cornell, R. (ed.) (1981) Respecting the Pupil: Essays on Teaching Able Students (Exeter: PEA)
  • Dewey, J. (1897) My Pedagogic Creed, School Journal 54, pp. 77-80
  • Emerson, R.W. (1883) Education [1863], The Writings of Ralph Waldo Emerson (Boston, Houghton Mifflin)
  • Garrison, D. (1999) John Dewey (Encyclopedia of Educational Philosophy and Theory) online at: www.ffst.hr/ENCYCLOPAEDIA/doku.php?id=dewey_john (accessed on 15 September 2013)
  • Heskel, J. and Dyer, D. (2008) After the Harkness Gift: a History of Phillips Exeter Academy since 1930 (Exeter: PEA)
  • Lane-Fox, R. (2008) Tutorials in Greats and History: the Socratic Method, in D. Palfreyman (ed.) The Oxford Tutorial: thanks, you taught me how to think (Oxford: OxCHEPS)
  • Paton, G. (2008) Oval Harkness Table Lesson Plan to Stop Pupils Hiding in Class, The Daily Telegraph (London). Online at www.telegraph.co.uk/education/secondaryeducation/3146468/Oval-Harkness-table-plan-to-stop-pupils-hiding-in-class.html (accessed 31 August 2013)
  • Pritchard, D. (2013) Epistemic Virtue and the Epistemology of Education, Journal of Philosophy of Education, 47.2 pp. 236-247
  • Savin-Baden, M. and Major, C.H. (2004) Foundations of Problem-Based Learning (Maidenhead: Open University Press / McGraw Hill)
  • Thoreau, H.D. (1906) Letter to Orestes Brownson [December 1837], The Writings of Henry David Thoreau (Boston, Houghton Mifflin)
  • - (1906) Journal [1850], The Writings of Henry David Thoreau (Boston, Houghton Mifflin)
  • Towler, K. (2006) The Men behind the Plan, Exeter Bulletin, pp. 24-33
  • Vorkink, P. (2013) How one Translates a Graduate Theological Education into a Secondary School Curriculum: this I Believe and how I Came to Believe it (Harvard: reader's version of doctoral thesis draft, as yet unpublished - kindly shared by Peter Vorkink)
  • [1] Including quite a large-scale roll-out at my own institution, Wellington College, as reported in the Daily Telegraph in October 2008: www.telegraph.co.uk/education/secondaryeducation/3146468/Oval-Harkness-table-plan-to-stop-pupils-hiding-in-class.html (23/05/2013).


Academy for Learning and Teaching Excellence
University of Bedfordshire
University Square
Luton, Bedfordshire