Philosophy tends to be a specialised field of educational research, kept separate from empirical investigtions. Yet many educational issues are equally philosophical and empirical, involving philosophical problems with our conceptions as well as inadequate empirical data. To address these issues, I argue, we should combine philosophical and empirical approaches to research, and I explore how we might combine them in this paper. I start by outlining the differences between empirical research and philosophical research, and then describe some ways they might be combined. Finally, I discuss and illustrate one new hybrid method of philosophical-empirical research, the Community of Inquiry.

Empirical research, put simply, involves gathering and analysing data to draw conclusions. For example, comparing a control group with an experimental group to find out which method for teaching mathematics leads to higher exam scores. But in philosophical research we do not gather and analyse data to draw our conclusions. Instead we construct concepts, theories and arguments, employing logic and reasoning to resolve conceptual and normative problems. For example, we might analyse the arguments for and against so we can judge whether some ways of maximising exam scores should be shunned.

A more controversial distinction is that philosophical research is in the realm of what should be—conceptually and morally—while empirical research is in the realm of what is (see Golding 2008, 2011, 2013a). For example, a philosophical analysis of ‘education’ might involve arguing that we should understand education in a particular way because this most aligns with our intuitions, with other legitimate conceptions, and with what we know. Although some might take this to be defining what the concept is, I think it is more plausible to see it as arguing for how we should define the concept. Philosophical analyses of education might also involve arguing that we should employ some educational strategies and avoid others, based on normative considerations such as equity or autonomy. On the other hand, an empirical investigation might analyse how different documents or education systems use the term ‘education’, or study which methods are educationally effective.

Philosophy tends to be kept apart from empirical research—empirical researchers gather data while philosophers conceptualise, without fraternising (see Carr 2004). Philosophers might sometimes work on the same problem as empirical researchers, but this multidisciplinary approach still tends to involve philosophical separatism, because we choose either philosophical or empirical methods for approaching an educational problem. For example, philosophy might be seen as the “conceptual underlabourer”, clearing ambiguity and incoherence so the real empirical work can be done (see Blake et al. 2003, 14), or merely one research method alongside various quantitative and qualitative empirical options (see Burbules and Warnick 2006).

But philosophy need not be so sharply differentiated from empirical research. As Quine shows (1953, 20), there is a blurred boundary between the a posteriori and the a priori, the world and language, reality and conception, and hence between the empirical and the philosophical. All research involves both philosophical and empirical elements, but sometimes one is foregrounded while the other is backgrounded. Empirical research addresses questions best answered by gathering empirical data, but the research is always framed by philosophical theories and conceptions about reality, knowledge or methodology, even if they remain unexamined. For example, ‘Is X an effective method for learning algebra?’ is an empirical question, but it cannot be answered without an implicit philosophical theory of learning. Likewise, philosophical research addresses questions best answered by philosophical argumentation and conceptualisation, but empirical observations, experiences and knowledge are still relevant, even though they may not be given explicit attention. For example, ‘What is teaching/learning?’ is a conceptual question best tackled using philosophical methods, but it cannot be answered without drawing on empirical experience of both teaching and learning.

Rather than philosophy being sharply separated from the empirical, philosophy is continuous with empirical research, and there is a continuum from theoretical “armchair philosophy” through to atheoretical data collection (Prinz 2008, 206). Across this continuum there are various ways philosophical and empirical research might be combined, which I sketch in Fig. 1 and in the rest of this section.

Fig. 1
figure 1

Continuum of research methods from theoretical philosophy to atheoretical empirical research

Full size image

One option is that philosophy is interdependent, collaborative and co-operative with the empirical. Philosophers and empirical researchers can team up, combining their expertise to understand and resolve complex interdisciplinary problems—for example how to design an educational system for an impoverished country with complex cultural, religious, economic and technological needs. Examples of this interdisciplinary approach include Dohn’s (2011) view of philosophy as a dialogue partner with the empirical disciplines, and Hansson’s (2008) ‘philosophy with’.

An alternative is ‘applied philosophy’ (Blake et al. 2003, 15) where the theories, concepts, principles and methods of philosophy are applied to empirical problems in a one-way or ‘top-down’ direction (Bufacchi 2004)—for example, a philosophical analysis of student rights or equity applied to the physical design of schools and classrooms. Similarly, philosophical theories can be applied as a frame or lens for analysing data in qualitative empirical research, such as a Foucauldian analysis of educational findings. Going in the other direction, philosophers can draw from the empirical findings to inform, influence, enrich or test their philosophical conclusions (Mejia 2008)—for example, using the results of empirical studies to refine and elaborate a philosophical theory about indoctrination. Prinz (2008) calls this empirical philosophy.

A third option in the middle of the continuum from Fig. 1 is philosophical–empirical research that is at once philosophical and empirical. For example, we can go back and forth between philosophical arguments and empirical observations to form and test conclusions about the “meaning and worth of possible pedagogical strategies, outcomes and contexts” (Mejia 2008, 162). In other words, we find a reflective equilibrium between the philosophical and the empirical (Bufacchi 2004). We form philosophical conclusions about issues like classroom management, and then through empirical research we find a mismatch between our meanings and values and what we observe in the classroom, so we adjust our philosophical conclusions, and so on.

Another example of philosophical-empirical research might be empirically situated, practical reflection, similar to the Aristotelian tradition of practical reasoning (Carr 2004). For example, a teacher’s practical reflection about how to facilitate classroom discussion might involve mixing empirical considerations about seating arrangements and student behaviour patterns, with philosophical considerations about the ideals of dialogue.

A third example of philosophical–empirical research might be experimental philosophy. Philosophers could conduct empirical psychological experiments to illuminate philosophical concepts such as learning, typically by asking participants to identify their intuitions in response to a philosophical thought-experiment. This research method democratises philosophy (Prinz 2008, 199). We can analyse the intuitions about philosophical thought-experiments from many research participants, just as an individual philosopher might analyse their own intuitions.

The Community of Inquiry as a Research Method

In this section I describe one further hybrid method of philosophical–empirical research—the Community of Inquiry (see Fig. 1). A Community of Inquiry involves selecting a group of participants who are expert practitioners, experienced in the issue being researched, and then engaging them in collaborative, dialogical philosophical inquiry about this issue, guided by a philosophical moderator or facilitator. This method of collaborative inquiry is derived from Peirce’s theories about how scientific knowledge is produced (see Pardales and Girod 2006) and from Dewey’s (1938) pragmatist philosophy, and was subsequently developed as an educational tool by the Philosophy for Children movement (Lipman 2003).Footnote 1 Although the Community of Inquiry is most commonly used as an educational tool, it is easily adapted for research because the inquiry process needed for learning is similar to the inquiry process for research. The Community of Inquiry as a research method has the following characteristics:

  • It is used to address philosophical problems and questions;

  • The method blends collaborative philosophical inquiry and empirical data collection and analysis;

  • The research is thus done with the participants, and;

  • Produces philosophical and empirical results.

I discuss each characteristic in the next four sub-sections. I use broad brush strokes for this general account of the Community of Inquiry, and then provide a specific example of one study employing this method.

Address Philosophical Problems and Questions

The Community of Inquiry is useful for investigating philosophical problems and questions about educational conceptions, especially the practical conceptions educators use to guide their practice. A conception includes beliefs, assumptions and propositional knowledge as well as attitudes, norms and values. Conceptions are philosophically problematic when they are inadequate or incongruous. An inadequate conception would not enable us to understand or deal with the situations we face. For example our conception of assessment might be limited to tests and examinations, and so it is inadequate to deal with peer assessment or assessment as learning. An incongruous conception is incompatible or inconsistent with other important conceptions or our experience. For example, our conception of teaching might be incompatible with our conception of student autonomy, or with our experience of independent student learning. We can investigate problematic conceptions in a Community of Inquiry by addressing questions such as: ‘What conception of assessment or teaching would be better for guiding practice in this particular context?’

Such philosophical questions cannot be resolved by empirical methods because the problem is a conception that does not make sense, not a ‘gap’ in our empirical knowledge. We solve the problem by enriching our conceptions, or refining them to make them congruous with other conceptions and our experience. This is philosophical work, not empirical. For example ‘To what extent should teachers tell students what to do?’ can only be answered by giving a philosophical analysis of ‘teaching’ and of the normative implications of authoritative teaching. It cannot be answered by gathering more empirical data.

Yet, our answers to philosophical questions must also be informed by empirical knowledge. The philosophical analysis needs to be situated in actual practice, dealing with the concrete situations teachers face, and building on their current conceptions. It would be pointless to give a philosophical analysis of some ideal, ‘straw-man’ realm which is out of touch with practice. We cannot decide the better conceptions for particular educators in particular circumstances at a particular time, unless we grapple with their current problems and conceptions. And we cannot find out about their current problems and conceptions without some empirical research or observation.

Collaborative Philosophical Inquiry and a Method of Collecting Empirical Data

The Community of Inquiry uses collaborative philosophical inquiry as a new method for collecting data, and equally, uses a focus group, moderated by a philosopher, as a new way to do philosophical research.

The Community of Inquiry follows a typical process of philosophical inquiry, though using distributed thinking where different participants make the different philosophical moves (Lipman 2003, 95). Broadly speaking, participants start with a philosophical question, develop positions as answers to this question, and then consider the arguments for and against the positions, in order to judge which are philosophically better—for example, more congruent, adequate, reasonable or fruitful (see Table 1).

Table 1 A simple framework for a process of philosophical inquiry (based on Golding 2009)
Full size table

The researcher guides the inquiry in partnership with the participants, ensuring it is philosophically rigorous but without imposing the researcher’s conclusions. As such, they might be called a ‘moderator’, the term used in a focus group. The moderator has to be skilled in philosophical inquiry, and in facilitating group inquiry. They must keep the participants on track so they make progress, but without asking leading questions or directing participants to the moderator’s conclusions. Most importantly, the path of the inquiry is not predetermined, so the moderator cannot rely on set questions but must judge how to proceed moment by moment, by asking him or herself: What stage of the inquiry have we reached? What would be a useful next step? What question can I ask to prompt us to take the next step? (See Table 1 for prompt questions). For example, imagine a Community of Inquiry about the question, “To what extent should teachers tell students what to do?” If one participant claimed, “The teacher should never tell a student what to do”, then the moderator might judge that the next stage in the philosophical inquiry would be to explore the reasons for and against. They would neither agree nor disagree with the claim, but would instead prompt participants to give reasons for and against, or examples and counter-examples. After receiving several responses, the moderator might then judge that it was time to prompt the participants to weigh up the various considerations, pro and con. If the participants then decided there were stronger reasons for rejecting the original claim than for accepting it, the moderator would then prompt participants to suggest a better claim, which would subsequently be elaborated and evaluated. And so on, as the philosophical inquiry continues.Footnote 2

This collaborative method is both philosophical and empirical. The ‘community’ engages in the same sort of philosophical inquiry, analysis and argumentation as an individual philosopher, so it is a philosophical method. But the method also involves gathering data about the participant’s conceptions, so it is also similar to an empirical focus group. The results would thus have some features of a philosophical argument and some features of a qualitative focus group, hence philosophical-empirical research.

Participatory Research

The partnership and genuine dialogue involved in a Community of Inquiry means it is a participatory research method. Research is done with participants rather than to them.

The research partnership in a Community of Inquiry, between moderator and participants, also democratises philosophical work (but in a different way than experimental philosophy discussed earlier). It makes philosophical inquiry available to participants who would otherwise be excluded. The moderator brings their philosophical inquiry skills, and the participants bring their expert empirical knowledge, so together they can engage in collaborative philosophical inquiry that would otherwise be out of their reach.

The participants also benefit from a Community of Inquiry as well as the researcher. Participants:

  1. 1.

    Clarify their personal conceptions and make them more explicit.

  2. 2.

    Better understand the inquiry topic and the conceptions of other participants.

  3. 3.

    Arrive at better conceptions than they could have articulated before participating.

  4. 4.

    Become better at philosophical and collaborative inquiry.

  5. 5.

    Develop a stronger community.

Philosophical and Empirical Results

A Community of Inquiry is likely to follow a messy process, following several paths with many leaps and backtracks. This is the unanalysed data, so to speak. To analyse this data we reorganise it into clear lines of inquiry rather than identifing only the “bare results” (Dewey 1938, 16). In particular, we identify the milestones produced at each stage of the inquiry—the positions identified, the reasons for and against, the evaluation of the strength of these reasons, as well as the conclusions drawn (See Table 1).

From an analysis of a Community of Inquiry we can identify results broken into empirical and philosophical facets. The empirical facet of the results includes a description of the participants’ conceptions, both their initial, unreflective formulation of their conceptions and their refined formulations at the end of the inquiry. The philosophical facet of the results includes a philosophical analysis and evaluation of these conceptions. This philosophical facet also includes the milestones produced during the collaborative philosophical inquiry (See Table 1). These are the same sorts of results a philosopher would report in a philosophical research article.

The refined conceptions that result from a Community of Inquiry are different from those reached in an interview or focus group, because they are the outcomes of different processes. In a Community of Inquiry and in a focus group the refined conceptions result from intersubjective testing, but this does not happen in an interview. In a Community of Inquiry the intersubjective testing and refining occurs through group inquiry, while in a focus group the testing and refining is through group interview. In a focus group we interview participants to find out what they actually think. In a Community of Inquiry we inquire together about what we should think. The aim in a focus group is to uncover the participant’s conceptions, not to push them to develop better conceptions as it is in a Community of Inquiry, so the process of questioning and probing involved is different (see Table 1).

I do not suggest a focus group could never involve philosophical inquiry. Rather I suggest that the Community of Inquiry could be seen as a philosophical, inquiry-based version of a focus group, especially suited for addressing philosophically problematic questions and issues.

As an aside, I also argue that the Community of Inquiry can reveal participant conceptions better than other research methods, because the Community of Inquiry results in reflective, intersubjectively tested conceptions. I identify ‘my conception’ (or at least my conception about a philosophically problematic issue) with the reflective, refined formulation I arrive at after my ideas are tested in collaborative philosophical inquiry. This formulation is what I really mean. Whereas I distance myself from the unreflective, often inadequate and incongruous, ‘off-the-top-of-the-head’ formulation I might give when asked what I think in an interview or focus group. Even if this formulation of my conceptions is revealing, if pressed, I would say “This is not what I mean”. I only ‘own’ a conception once I have reflected on it and explicitly accepted it. My unreflective statements, which are often based on unquestioned assumptions picked up who knows where, are not mine in an important sense.

We evaluate the empirical facet of the results in a Community of Inquiry differently to the philosophical facet. The empirical results require empirical warrant, and the philosophical results require philosophical warrant. The empirical results are meant to be an accurate formulation of the participant’s conceptions, and we evaluate these results like any qualitative empirical research: Are the results likely to be accurate because they are derived from a trustworthy and reliable method? We might even ask ‘Can the results be generalised to the broader population?’ But a Community of Inquiry is primarily concerned with the philosophical facet, and we want to evaluate whether we have generated better conceptions, conceptions we should employ, so I will focus on how we evaluate these philosophical results.

The philosophical results of a Community of Inquiry are primarily evaluated according to the strength of the arguments, not the trustworthiness of the method (Golding 2013a, 154). Rigorous philosophical results are those with philosophical warrant: Well-reasoned, justified or supported, using the standards of philosophical inquiry, logic and argumentation. To evaluate the philosophical results we ask questions such as: ‘Have we identified the most important alternative resolutions? Have we identified plausible reasons for and against? Are there strong arguments to support our conclusions? Do the arguments provide sufficient support to justify the conceptions we reach?’

We also have a secondary warrant for adopting the philosophical results because they are the outcome of a Community of Inquiry; the outcome of rigorous philosophical analysis and inquiry, tested using intersubjective checks and balances, by expert practitioners with practical wisdom in the area. During their communal philosophical inquiry, the participants in a Community of Inquiry criticise, test, evaluate and refine suggestions from other community members. This is similar to peer-review, which lends extra philosophical warrant to the results. Furthermore, the results have the warrant of expertise, which should be considered alongside the philosophical arguments and the intersubjective justification. Because the participants have practical expertise in the inquiry topic, this gives extra warrant to their evaluations and revised conceptions, and makes it more likely these are the ones we should adopt.

The philosophical warrant for the results, however, depends on the moderator’s skill. They must operate as philosophical guide, enabling the participants to make progress in their joint inquiry while ensuring the inquiry is philosophically rigorous, but without leading participants to the moderator’s judgements and conclusions (For example, see the illustration of a Community of Inquiry about whether teachers should tell students what to do, described earlier in this article. Also see Golding 2013b for more on guiding an inquiry with participants rather than for them).

Fix, Build, Diagnose or Guide: Strong and Weak Metaphors for Learning Advisors

In this section I illustrate how I used a simple version of the Community of Inquiry to evaluate metaphors for being a University learning advisor.Footnote 3 A learning advisor or learning developer provides general support for student learning rather than teaching a specific discipline or profession. This frequently involves one-to-one consultations or workshops with students in such areas as study skills, writing, time management, or other academic advice (see Hartley, Hilsdon, Keenan, Sinfield, and Verity 2011).

Metaphors such as ‘coach’ or ‘counsellor’ are commonly used by learning advisors to guide their practice (Carter & Bartlett-Trafford, 2008). Yet these are often philosophically problematic, being both inadequate to deal with the learning situations an advisor faces, and incongruous with an advisor’s values. In my study I aimed to identify the strengths and weaknesses of different metaphors, what each hides and highlights, and so to evaluate which metaphors learning advisors should employ.

A philosophical–empirical Community of Inquiry was a useful method for addressing the philosophical problems and for evaluating the metaphors for being a learning advisor. If learning advisors participate in a Community of Inquiry, they can identify the metaphors they currently employ to guide their practice, and also evaluate which would be the better metaphors to adopt.

All the learning advisors at the University of Otago in May 2011 and 2013 were included as participants, because they had practical expertise in the topic. I was the moderator. For the method to be successful the participants had to be committed to collaborative inquiry about the topic, and willing to operate as a Community of Inquiry. They indicated they were keen to inquire together about this topic. Yet it takes time for any group to overcome power differences or potential conflicts, and to develop the required safety, mutual respect and skills needed to inquire together. The learning advisors already operated somewhat as a community because they had worked together, but I still had to spend some time managing the group dynamics, encouraging participation, and giving them time to learn to inquire together. I also had to give them time to learn to inquire with me as a guide and moderator rather than as the leader with the ‘right’ answers.

Although I was also a learning advisor, many of the participants had more experience, so my primary expertise was as the philosophical moderator, working in partnership with the expert participants. I had the necessary expertise to be the moderator. For over 15 years I had used the philosophical Community of Inquiry as a method for teaching and learning.

I moderated an extended Community of Inquiry, including two group discussions with participants, separated by 2 years to allow time for reflective practice as we worked together. Most participants were involved in both discussions, but some left the University after the first discussion, and some new participants joined. The process was a truncated collaborative philosophical inquiry framed by Lakoff and Johnson’s (1980) work on conceptual metaphors. We sought two main inquiry milestones, our philosophical results: the possible metaphors for being a learning advisor (the positions), and the strengths and weaknesses of these metaphors (the arguments for and against). We did not seek one best metaphor for learning advisors (a final position), because no single metaphor would encompass everything a learning advisor does. Instead we aimed to sort the better metaphors from the worse.

The research process followed these stages:

  1. 1.

    I became familiar with the literature on metaphors for learning advisors so, as the moderator, I could guide the participants through the scholarly terrain (though without trying to lead them anywhere in particular).

  2. 2.

    We then had the first Community of Inquiry discussion where we:

    1. a.

      Identified a learning advisor’s aims and values.

    2. b.

      Identified and elaborated the most important metaphors we could use to guide our practice.

    3. c.

      Evaluated the possible strengths and weaknesses of each metaphor, using the aims and values as criteria.

  3. 3.

    Next, to analyse the data, I categorised what was said into different families of metaphors, including their strengths and weaknesses, then wrote this up in a draft article.

  4. 4.

    I consulted the literature again to find any metaphors or strengths and weaknesses we had missed, and then added these to the draft article so we could build on and contribute to the literature.

  5. 5.

    The draft article was circulated to all participants, so we could have a second discussion to add to, confirm, clarify and evaluate the metaphors, strengths and weaknesses.

  6. 6.

    We continued our second discussion until we had reached saturation for the milestones; until we had exhausted suggestions for new metaphors and exhausted suggestions of new strengths and weaknesses. We did not seek agreement about the best metaphor because this was not the aim of the study. We did, however, find consensus about the strengths and weaknesses of each metaphor, and consensus about the better metaphors we should employ, and the worse metaphors we should avoid.

  7. 7.

    Based on the second discussion, the article was revised and circulated for final confirmation, before it was submitted for publication. In the spirit of participatory research, all participants were included as joint authors, rather than I, as researcher, ‘stealing’ their words and presenting them as my ‘own’ findings.

A Summary of Some Results

The first metaphor we considered was ‘the fixer’, which entails that learning advisors serve a remedial function, fixing ‘broken’ students. This is a common metaphor for learning advisors, but we were all uncomfortable with it, though initially we were unsure why. Through inquiry we recognised that this metaphor accurately entails that learning advisors help students to achieve academic success, and we judged this to be a strength. But we agreed that ‘the fixer’ is not an appropriate metaphor to guide our practice because it is incompatible with how we help students. There were three main weaknesses:

  • This metaphor entails a deficit model of learning, where learning involves fixing something broken. But we argued that a developmental model of learning is more appropriate for learning advisors, where learning is a process of improvement from wherever you happen to start.

  • The fixer metaphor entails that students are passive—they have a problem and we can fix them. This is not empowering for students or for learning advisors.

  • The fixer metaphor entails that learning support is only for ‘dummies’. This stigma restricts equitable access to learning advisors.

After clarifying the weaknesses of ‘fixing’ metaphors, we recognised that it would be better for learning advisors to be guided by ‘improving and developing’ metaphors such as builder, architect, coach or guide.

Next, we considered the ‘builder’ metaphor, commonly held by learning advisors, and evaluated to what extent we ‘build learning’. We judged that one strength of this metaphor was that we do help students to build their learning, whether they have a problem, or just want to improve further. But this metaphor is still weak: A builder does the work for their client, and this is inconsistent with one of a learning advisor’s main values, fostering independent learning. So, we judged that the ‘builder’ metaphor is better than the ‘fixer’ but it still has significant weaknesses.

The inquiry continued when we considered whether we diagnose and prescribe treatments for learning ailments, or whether we act as a learning guide. In brief, we thought the diagnosis metaphor was strong as it clarified how we judge what we can do to assist student learning, but it also entailed that students were ‘ill’ which was a similar weakness to the ‘fixer’ metaphor. We judged the guide metaphor to be one of the strongest metaphors for learning advisors because it was both informative about what we do, and consistent with our values: a learning advisor provides developmental guidance about central academic concerns, and partners and empowers (guides) students so they become independent learners.

The philosophical results from this study—the plausibility of the metaphors and their strengths and weaknesses—were tested intersubjectively in successive stages. They were initially tested during the first Community of Inquiry discussion (stage 2), then through a second discussion about the results so far (stage 6), and finally, through final confirmation by all participants (stage 7). The results could be further tested by conducting a new Community of Inquiry with different participants and seeing if they reach the same results. Alternatively, the new Community of Inquiry could inquire about, and test, the results of this Community of Inquiry.

This Community of Inquiry took a practical, pragmatist approach to philosophical inquiry, and should be judged by these standards. The aim was to evaluate which metaphors would be better to adopt in practice; which metaphors enable learning advisors to find their way around their practice. As such the inquiry was closer to Carr’s (2004) idea that philosophy is a practical reflection on practice rather than detached theorization. The metaphors were intended to have heuristic (and perhaps propaedeutic) value in practice; this is the same value Newtonian physics has for engineers constructing bridges, or the value geocentric astronomy has for navigation, even though they cannot be defended as the most comprehensive, precise, or accurate conceptions. Given our practical philosophical aim, we are warranted to adopt the metaphors we found to be the strongest, not because we have proved them philosophically true, but because they have intersubjective warrant, the strongest arguments supporting them, and thus because they are better than the alternatives.

Conclusion

My aim in this article was to open up new possibilities for research in education, in particular, new possibilities for blending philosophical and empirical research. I argued that as well as philosophy instead of empirical research, philosophy with empirical research, or philosophy applied to the empirical, there could also be philosophical-empirical research which is at once philosophical and empirical. I then showed that the Community of Inquiry provided a good example of such a hybrid research method. If participants engage in collaborative philosophical inquiry about educational conceptions, using distributed thinking, and guided by a skilled philosopher, then the result will be both an empirical account of their conceptions, and a philosophical analysis and argument for the better conceptions. In this way the Community of Inquiry provides a new method for educational research, and a useful means for combining philosophical and empirical research.