Sharing talk, sharing cognition: philosophy with children as the basis for productive classroom interaction

Taking the linguistic turn requires an examination of the relationship between thought and language: starting with Wittgenstein, we draw out the implications of expressing thought in a jointlyconstructed system of meaning. We then examine the relationship between thought and language in a pedagogical context by drawing on the practice of philosophy with children in the classroom, identifying key skills which are important in the development of meaningful classroom interaction, connecting this to philosophy with children practice. We go on to explore the consequences of educational attainment for those children who enter school with impoverished language, referring to a number of key studies, including our own, which highlight the detrimental effects not only on children’s outcomes at school, but also their ability to become equal inhabitants of a linguistic space. Therefore we argue that a focus on oracy skills crucially underpins wider outcomes. We conclude by looking at ways in which oracy skills can be developed in the classroom, arguing that although there are other techniques for developing oracy skills in the classroom, philosophy with children provides the most comprehensive way of doing this.


Introduction
Language in the context of pedagogical practice is a key preoccupation for Robin Alexander who has carried out extensive, multinational work into pedagogical practice, with a particular focus on classroom dialogue (Alexander 2000(Alexander , 2004(Alexander , 2010. For Alexander, it is the quality of classroom dialogue that constitutes effective pedagogy, as "talk truly empowers children as learners" (Alexander 2010: 105). This dialogic pedagogy is characterised by teaching allowing children to participate in extended discussion, and valuing talk as an outcome as much as written work.
Good talking skills often come under the name of oracy, which was coined as a counterpart to literacy and numeracy and which Mercer defines as "teaching people to use language effectively in whatever situation they're in" . The Oracy@Cambridge group in the UK provide a framework for teaching oracy skills which sets out a number of points on which children should be competent if they are to be successful orators. These comprise the physical, linguistic, cognitive and social-emotional components of using language. Although oracy is not as widely used as literacy and numeracy, we use it here as a pedagogical term for children's communicative competence in the classroom.
Given that there have been many questions raised about the relationship between thought and language (Hacker 2007;Harré & Gillett 1994), in considering the development of children's oracy skills, consideration must necessarily also be given to children's thinking skills. Lipman's (1980) claim for this relationship is that it is often misinterpreted: "the common assumption is that reflection generates dialogue, when, in fact, it is dialogue that generated reflection" (Lipman 1980: 22). Indeed, he suggests that conceiving of thinking as something "private and internal" (Lipman 1980: 22) is detrimental to pedagogy because the perception is that it is difficult for teachers to make children's thinking apparent, to differentiate good from poor thinking, and to improve thinking.
A key aspect of dialogic pedagogy, is the recognition that the learner "is not a selfsufficient entity; it needs the other, his recognition and his formative activity" (Todorov 1984: 96). The dialogic encounter between one learner and another is therefore one in which meaning is made between the two: as Holquist (1990: 38) writes, although multifaceted, dialogue "can be reduced to a minimum of three elements… an utterance, a reply, and a relation between the two". The third element here mirrors sociocultural, constructionist theories of children's development (Vygotsky 1978) and it is this which allows for the sharing of language and thought. This relation between the utterance of one speaker and the reply of the other is the focus of dialogic pedagogy as teachers employ strategies to facilitate understanding between learners.
If a dialogic pedagogy both allows children to learn effectively and equips them with the skills to become active and engaged citizens (Michaels, O'Connor and Resnick 2008), it follows that it is important for schools to provide opportunities for all children to develop good oracy competence. One way of implementing dialogic pedagogy and focussing on the interplay of thinking and oracy skills is through philosophy with children. Philosophy with children programmes originated with Matthew Lipman in the United States in the 1970s, when Lipman saw a need for educational reform which would give children a meaningful education and equip them with the skills to critically engage with the world (Lipman 1980(Lipman , 2003. This could be achieved through questions generated from philosophical stimuli.
The term Philosophy for Children refers to Lipman's specific programme, and is widely understood in that context. However there are a number of other permutations of philosophy with children, and so that is the term used throughout is philosophy with children, which refers to a range of philosophy for children practices that focus on a community of enquiry approach to dialogic practice.

The Philosophy of Language
The theoretical background to the case for pursuing philosophy for children lies in the so-called 'linguistic turn', namely the increasing preoccupation of philosophy in the early 20 th century with language itself as the underlying structure for, and articulation of, what Hacker (2007) refers to as our conceptual scheme. It is Wittgenstein who is most often identified as the progenitor of the linguistic turn; his work was suffused with the relationships between language, logic and philosophy. Starting from a conviction in his youth that logical analysis of language would disclose the substance of the world, his arrival at the position that philosophical problems stem largely from linguistic confusions and are to be resolved largely by clarification of the uses of words led him to conclude that philosophical questions were indeed essentially ones of language. "All philosophy is a "critique" of language," he says (Wittgenstein 1961, 4.0031). The goal of philosophy became the understanding of the structure and articulations of our conceptual scheme, and the resolution of the problems of philosophy, which he identifies with problems of language such as lack of clarity about the uses of words and covert misuses (Hacker 2007).
So a central method of philosophy after the linguistic turn is therefore to examine meticulously the uses of words in order to disentangle conceptual confusions. One of the points of this, according to Wittgenstein, is to arrive at an overview of a concept and to produce a representation of the field of concepts with which we are dealing. He compared this to drawing a map that will help us find our way around in the field of our concepts and conceptual structures (Wittgenstein 1961, 4.01). And there is also a clear link to the nature of thought itself. An extreme form of the linguistic turn (Dummett 1994) holds that there is no thought which is inexpressible in language. Wittgenstein says, 'We may say that thinking is essentially the activity of operating with signs' (Wittgenstein's Blue Book, cited in Harré & Gillett 1994: 50). The linguistic turn therefore also implies that the discipline of philosophy does not constitute a contribution to knowledge about reality (as with scientific knowledge) but is rather a process, contributing to a particular form of understanding.

Philosophy with children: linking theory to practice
As has been often noted (Daniel and Auriac 2013;Jorgensen 2009;Vansieleghem 2012), when children philosophize they are not learning the canon of philosophical thought, but are engaged in an active process. The nature of that process centres on the relationship between thinking and speaking, and therefore forms a central part of this paper. For Wittgenstein, we operate in social spheres in which "the workings of each other's minds are available to us in what we jointly create conversationally" (Harre and Gillett 1994: 27). Language functions as a "shared symbolic system" (p. 44), and individuals having a share in that system are able to communicate meaningfully with others.
This shared system means that the very act of discursive philosophical activity in the classroom automatically engages learners as 'concept users', namely as competent manag-ers of systems and signs (Harré and Gillett 1994). And we do not engage them as isolated individuals. Harré and Gillett point out that "the grasp of (the use of a word) [a concept] is an active discursive skill… It is built on participation in discourse, and it is governed by rules or norms that tell the thinker what counts as an item of this or that type" (p. 48). Philosophy in the classroom therefore has the potential for learners to work together to build and discover meaning for themselves and to hone and develop their systems of understanding.
Questions regarding the inclusion of philosophy in a curriculum often lead to questions of the purpose of education (Martens 2009). If a Wittgensteinian perspective tells us that private experiences can be shared in common language, then children's ability to effectively express their cognitive experiences in language and to use this ability to further their educative and life experiences is, for Lipman (1980) and for Dewey (1933), the hallmark of education. Therefore it is important to examine what it is about philosophy with children that results in improved oracy skills which allows children not only to attain more highly at school but also to participate more effectively in civic life beyond school.
As previously mentioned, doing philosophy with children is conceived of as a process -of philosophizing. This has led to some criticism from academic philosophers who do not see their subject in this way, and question whether or not what children are really doing can be called philosophy (Vansieleghem 2012). It is an important point that there can be no doubt that there are forms and functions of dialogue that are not philosophical (referred to by Hacker (2007) as natural language). That is one of the reasons why some philosophy for children programmes (Cassidy and Christie 2013) insist on the facilitators of the programmes having a philosophical background in order to recognise and develop philosophical themes. Moreover, although children have been described as 'natural philosophers', what people tend to mean when they say this is that children ask lots of questions. This is of course true, but that does not make those questions philosophical. Martens writes that "spontaneous flashes of insight or iterations of wisdom picked up" (Martens, 2009: 101) are not enough to fulfil the requirement of being a philosopher, and neither is a "general child's curiosity".
This indicates that for dialogue to be philosophical requires some kind of facilitation, moving towards an educational aim. This concept of philosophy for children aiming at a pedagogical goal, is not without issue, which will be discussed subsequently. However, from Lipman onwards, philosophy for children programmes have come to be associated with critical thinking skills. For Daniel and Auriac, in fact, this association is "the main common trait… between P4C and philosophy" (2011: 416).
This commonality between philosophy with children and critical thinking refers to philosophy as a process, different from the acquisition of knowledge of the thoughts of previous philosophers. The very process of engaging in philosophical discussion is one which involves the process of learning to think critically. As Martens (2009) writes, "[p] hilosophy with children should be characterised by dialogue, concept analysis, argumentation and action" (p. 18). Daniel and Auriac also go on to point out that the parallels between philosophy and critical thinking in terms of "questioning, conceptualizing, evalu-ating" (Daniel and Auriac 2011: 421). Indeed there is a substantial body of research which shows that philosophy for children has a positive effect on children's higher-order thinking skills (Garcia-Moriyon, Rebollo and Collom 2008), and it seems clear that there is a strong case for identifying the development of critical thinking skills with philosophy -skills which, of course are translatable across the curriculum.
However, although critical thinking is, for many, inextricably linked with philosophy with children programmes, it is only one strand of the conception of such programmes. Englhart (1997) outlines three aspects -of which Lipman's critical thinking strand is just one. A second strand developed with Matthews, who does not see the 'goal' of philosophy with children as one which promotes developmental stages, but rather opens up spaces in which children can explore their own thinking. A final strand was a critical emancipatory one, in which philosophy with children can disrupt the power relations between the powerful and the oppressed. These aspects will be explored further in the context of the aims and outcomes of philosophy with children programmes.

Children's experience of language
In the UK, there has been concern over the educational attainment of disadvantaged children (Department for Education 2015; National Literacy Trust 2014, Educational Endowment Fund 2015), namely children who receive the school pupil premium, an extra sum of money paid to the school for children who fit certain socio-economic criteria. These include primarily children who receive free school meals because of a low household income level. (Education Funding Agency 2016).
One way in which disadvantaged children are different from their non-disadvantaged peers is that they are likely to have had less experience of a range of talking situations in the home. Hart and Risley (1995) are well-known for their study in the United States: a large-scale, longitudinal study which examined talk in the homes of American preschool children and found that by school entrance age, children from welfare homes had had exposure to 30 million fewer words than children from professional homes. In addition, the talk in the welfare homes was command or instruction driven and with a high level of negative interaction. Conversely, in professional homes children were more likely to participate in a range of talk situations, including extended discussion.
In the UK, children are assessed at the age of four when they leave pre-school on a range of competencies based on the Early Years Framework. The National Literacy Trust released a 2014 report in the UK which indicated that when children leave preschool at the age of four, disadvantaged children are already 13 percentage points behind non-disadvantaged children when assessed for the Communication and Language module. This indicates that the findings of the Hart and Risley study are still pertinent in the current UK context.
We recently conducted a study in a Primary school in the South of England during which we made observations of children during whole class discussions. We gave teachers of Key Stage One children (age 5-7) a number of philosophical questions from which to choose and asked them to have a whole class discussion for around fifteen minutes. We then analysed the discussions based on pre-defined categories based in part on Cazden's (2001) work on classroom discourse (see Figure 1). We referred to each specific utterance by the child as a 'talk moment'. Not only did the disadvantaged children in the class tend to make fewer contributions to the discussions, but the way they spoke was different, characterised by contributions of three words or fewer which indicates a call-and-response answer to a teacher's prompting rather than a sustained contribution to the dialogue. Given this, it is clear that although the Habermasian discursive goal may be "a free space in which all persons involved in the inquiry have an equal chance to bring their arguments forward" (Vansieleghem: 6-7), children's experiences of talk in the home provide very different experiences when they come to school. It is therefore very difficult to entirely agree with Matthew's (1984) argument that children can enter into successful philosophical inquiry precisely because they lack the socialisation of adults. He intends this to mean that there is therefore a pre-socialised 'space' in which children are free to voice thoughts and frame questions in a way which would be more constrained for adults. However, as is shown by our findings in relation to disadvantaged children, that 'space' has already been defined at least to some extent by children's early experiences, particularly with language.
Therefore, consideration needs to be given to how to decrease the skills gap between children while advancing the skills of all children. On the one hand, it could be that the discursive practices that constitute philosophy with children allow children to experience the sort of talk that may not have been present in their home lives, and that actually it is merely the practice of the skill of talking and listening that leads to improvement in the same way as the practice of any other skill. If this is the case, it is not necessarily philosophy with children itself that leads to these deficiencies being remedied. On the other hand, there are specific practices associated with philosophy with children which studies have shown to lead to an improvement in oracy skills, which link linguistic ability to critical thinking skills which have been previously discussed as a mainstay of philosophy with children practice. Nippold, Frantz-Kaspar, Cramond, Kirk, Hayward-Mayhew and MacKinnona (2015) conducted a study on adolescents' speech in which the syntactic complexity resulting from a critical thinking task and a conversational task was compared. Not only did the study find that the "critical-thinking task prompted substantially greater use of complex syntax than the conversational task" (Nippold et al. 2015: 329), but they claim that "conversational tasks are unlikely to reveal a speaker's best use of complex syntax" (Nippold et al. 2015: 325).
Therefore, referring back to the Oracy@Cambridge's framework of oracy skills, it seems that a focus on the cognitive aspects of oracy is also a focus on the linguistic aspects in the context of critical thinking. The philosophy of language as conceived by Wittgenstein also leads us to this conclusion. For Wittgenstein, "words are deeds" (Wittgenstein, 1980: 46). What renders a sign meaningful is how it is used by linguistic agents, not how it relates to the metaphorical idea of the agents' inner minds, and therefore cognition is "not-just-in-the-head" (Susswein and Racine 2009: 185). It has a strongly sociocultural element, as more recent theories of distributed and situated cognition attest.

Collaborative learning and attainment as improved social interaction
Theories of distributed and situated cognition imply a model of classroom activity into which philosophy for children fits well. As a discursive and inherently collaborative activity, as well as giving children the opportunity for improving their linguistic sophistication, it teaches children social skills and expects them to use these to solve problems. It result in the most educationally productive form of classroom discussion, known as exploratory talk (Mercer 2012). It therefore straddles the Social/Emotional strand of the Oracy@Cambridge framework of skills as well as the Physical, Linguistic and Cognitive strands.
Educators have been aware for some considerable time that learning is a fundamentally social activity: studies in the US in 1980 and 1994 identified the key benefits of collaborative learning as increased effort, positive relationships, improved psychological health and improved attainment (Johnson and Johnson 1989;Johnson, Johnson and Holubec 1994). Schools are today increasingly aware of the need to become "learning communities" because classrooms that operate as communities get better results (Education Scotland 2016).
As a potential tool for helping to develop learning communities, the practice of philosophy for children often relates to theories of education as a whole. Indeed, Oral (2013) claims that in today's policy-driven educational climate in the UK, philosophy programmes are one of the only ways to provide a focus on the primacy of collaborative social transaction. He attributes this to an exclusive focus on knowledge, and as practitioners in the UK education system we would extend that to include a narrow focus on assessment as the perceived aim of education. Alexander would certainly decry this notion of educative purpose as un-dialogic, resulting in the sort of classroom dialogue which results in "guess what I'm thinking" test questions. (Nystrand 1997in Alexander 2010 Even subjects that have traditionally been considered individualistic in the UK and US, such as maths, have been identified as a community enterprise. Jo Boaler's 1999 comparison of two radically different approaches to teaching maths showed that a more project-based, collaborative approach resulted in better attainment at GCSE as well as in tests devised for the purposes of the study. Boaler identifies the capacity of the 'collaborative-taught' students to think for themselves, select relevant information from irrelevant, and work together proactively to find and define problems as well as solve them as key factors in the students' improved performance. These findings related to 9-11 year olds but their relevance for primary learners is reinforced by practitioner Mike Askew, who is clear about the importance of talk for learning in primary mathematics: "Talk is central to mathematics lessons… It's not a list of words that you select from to put your ideas 'into'; it is the words through which ideas are formed" (Askew 2012: 136).
Taking us back to Alexander, Askew and Boaler both emphasise the centrality of dialogue to maths learning. Though for Askew any social skills that might be developed as part of the process are secondary to the main objective of progress in maths, the link between collaborative learning, social skills, language and improved attainment is clear. There is also an increased sense of interdisciplinarity: maths does not exist in a vacuum, or behind walls, but is relevant to, and has exchanges with, other subjects and the environment in which learners are situated -whether that be in the classroom environment or the outside world.
If these skills are important to attainment in maths, it must be postulated that they will boost attainment in other areas as well. The key question is whether improved social/ linguistic/collaborative learning skills can be developed or in a cross-disciplinary manner such that attainment is improved across a wider range of subjects. Fortunately there is evidence here too -and this evidence returns us to Philosophy for Children. The UK's Educational Endowment Foundation (EEF) undertook a study in 2013 into the dialogic practice of the Philosophy for Children (P4C) programme designed by the Society for the Advancement of Philosophical Enquiry and Reflection in Education (SAPERE). In a total of 48 schools across a wide range of geographies, with an above-average number of children from disadvantaged backgrounds, pupils in years 4 and 5 received one session of P4C per week between January and December 2013. These sessions were based on a Community of Inquiry approach, where teachers used a stimulus such as a story, image or artefact to encourage children to develop their own philosophical questions and then discuss the questions generated by the group.
The evaluation of the project, undertaken by Durham University, found that P4C had a positive impact on pupils' Key Stage 2 average progress in reading and maths (with effect sizes of +0.10 for maths, and +0.12 for reading) (Education Endowment Foundation 2015: 3). This equates to two months' additional progress in relation to both activities. Significantly, the gains for children from disadvantaged backgrounds were even more significant: three months' additional progress in maths (effect size +0.20) and four months additional progress in reading (effect size +0.29). Feedback from teachers and pupils also suggested a belief that P4C had a beneficial impact on wider outcomes such as pupils' confidence to speak, patience when listening to others, and self-esteem. Some teachers also perceived that P4C had a positive impact on general classroom engagement and may have resulted in some pupils asking more questions across all lessons.

Conclusion
As Alexander emphasises, and this paper reinforces, the case for dialogic approaches to teaching is clear. So is the capacity of philosophy for children to improve core skills that will underpin attainment in other disciplines. But are philosophy with children programmes the only (or the best) way to achieve this? The International Baccalaureate Organisation clearly values philosophical discussion and incorporates elements of learner-led epistemological enquiry as part of all its programmes, including the Primary Years Programme.
However, it is clear from the scope of the Oracy@Cambridge skills framework that these skills have the capacity to be developed in other ways, using different pedagogical strategies such as storytelling or formal debate. This clearly merits further study. Nonetheless, it appears that both these alternatives have certain limitations compared with philosophy. Formal debate focuses on discussion rather than dialogue, tending to reject plurality of opinion and co-construction in favour of refutation of others' points of view; storytelling, whilst promoting imaginative co-construction, does not engage as extensively with critical thinking and, as Nippold et al (2015) had found, is less likely to prompt the same degree of syntactic complexity as critical thinking activities.
It is our view that a philosophy with children approach engages with the widest range of criteria from the Oracy@Cambridge skills framework and therefore has the greatest potential to offer a technique for improved academic attainment -and also (in the light of the EEF study) for narrowing the gap for children from disadvantaged backgrounds. The delivery of philosophy for children needs careful consideration in order to ensure it benefits all children. At a practical level this must entail balancing teacher workload with the need for autonomy and efficacy, certainly in the UK, where initiative overload is a common complaint across the profession. Further work is currently underway to develop models which will enable and promote such balance, and we would be happy to have further discussions on this point. One of the aims of this work is to introduce a greater degree of coherence in philosophy teaching, in particular as regards its development of critical thinking skills.
As has been noted throughout this paper, social interaction through language requires users to be aware that it is governed by rules, and to engage in language use through those rules in order to make coherent meanings with others. Lipman argues that the same is true of philosophy with children programmes, in which children have their "social impulses" (Lipman 1980: 24) developed through the rules of engagement of philosophical discussion. Therefore, we also argue that the explicit teaching of oracy skills (rather than allowing them to develop through everyday classroom interaction) is vital, as the teacher plays a crucial role in fostering a community of enquiry which is governed by a shared set of rules, and which allows philosophy to promote a "fundamental respect for the other and his often confusing and peculiar opinions" (Martens 2009: 105) It is our conclusion that philosophy with children is a "cultural technique" (Martens 2009: 16), one which has benefits for children's learning which go beyond critical thinking skills, or linguistic skills but one in which those skills entwine. The Bavarian pre-school curriculum document includes a statement on philosophical education for children: "It challenges the entire person of the child to understand and position itself in relation to the immeasurable, uncountable, imponderable things that form the foundations of our reality" (in Martens, 2009: 101). Perhaps, then, the value of philosophy as a means to improving oracy skills is that it also allows the child to understand and reposition itself in relation to others in a classroom context and so stand in a different relation both to others and its former self in terms of language and, through it, cognition and collaboration.