Ameliorating educational concepts and the value of analytic philosophy of education

Abstract R. S. Peters and a small group of contemporaries set the foundations for analytic philosophy of education in the 1960s, a field which continues to this day. This article asks about the value of analytic philosophy of education today, and proposes alterations to its initial aims and methods to make its value clearer. I outline some critiques of analytic philosophy of education, and respond by clarifying its aims. The key insight is that if analytic philosophy of education is explicitly aligned with recent trends in social philosophy to focus on ameliorative analysis, then its value and relevance as a method in educational studies becomes clearer. Ameliorative analysis starts with the social role that a concept plays, and analyses it with this purpose in mind. I illustrate how R. S. Peters analysis of education would be approached differently with ameliorative analysis in mind, and conclude by pointing to some interesting points of discussion that arise from this approach.


Introduction
When R. S. Peters took over the chair in philosophy of education at the University of London from his predecessor L. A. Reid, he set out a new direction for philosophy of education: There was a time when it was taken for granted that the philosophy of education consisted in the formulation of high-level directives which would guide educational practice and shape the organisation of schools and universities. These expectations of philosophy still persist in ordinary language. They are as implicit in the question 'What is the philosophy of education?' as they are in the question 'What is the philosophy of life?' Professional philosophers, however, are embarrassed by such expectations. For during the twentieth century philosophy has been undergoing a revolution, which has consisted largely in an increasing awareness of what philosophy is and is not. Few professional philosophers would now think that it is their function to provide such high level directives for education or for life; indeed one of their main preoccupations has been to lay bare such aristocratic pronouncements under the analytic guillotine. They cast themselves in the more mundane Lockian role of underlabourers in the garden of knowledge. The disciplined demarcation of concepts, the patient explication of the grounds of knowledge and of the presuppositions of different forms of discourse, has now become their stock in trade. (Peters, 1966, p. 15) & Peters, 1970, p. 4); and Scheffler in the United States whose analysis of the concept of knowledge and its place in education sought 'a general description or definition, a statement of criteria of knowledge which may serve to clarify its logical status ' (1965, p. 5).
Peters, Hirst and Scheffler pioneered analytic philosophy of education to the extent that Peters describes himself as 'charting a comparatively unexplored region where there are almost no signposts ' (1966, p. 7). A lively tradition exists in their wake, with analytic philosophers of education working in education, politics and philosophy departments across the world. However, the value of conceptual analysis has been called into question by philosophers of education who would consider their role as more synthetic and less analytic. The move from formulating high level directives in education to clarifying individual concepts has led to analytic philosophy facing charges of irrelevance, insensitivity and triviality.
Responding to Peters' work as the new chair of philosophy of education, his predecessor L. A. Reid claimed that a philosophy of education which does not seek to 'guide humanity' is missing the point of the enterprise. He says that 'the proper function of analysis is the better understanding of whole which are analysed; it is servant and not master… if synthesis without analysis is empty, or at least muddled, analysis without synthesis is blind, or at least pointless or feckless ' (1965, p. 24). conceptual analysis on its own is not worthwhile.
More recently, Paul Standish sets out this tension as it continues today: Some have thought that conceptual analysis can actually show you the heart of what education is aboutthat it has its own logic, which a close examination of its central concepts (such as teaching, learning, and knowledge) will reveal. (see for example John Wilson, as well as the work of Israel Scheffler, RS Peters, Paul Hirst, Harvey Siegel, and christopher Winch in certain respects)… Philosophers of education can challenge this in favour of the view that there is no avoiding the direct consideration of questions of value. (Standish, 2010, pp. 7-8) Standish points out that on its own, conceptual analysis is too abstract. charting the logical relations of concepts, trying to define them in a coherent way, and smoothing out technical flaws in our ordinary understanding of the world is not sufficient to support progress in educational studies. Standish argues that there is always going to be a role for normative considerations, for questions of value. Pure conceptual analysis either fails to be explicit about normative issues, or leads to trivial conclusions through failing to take them into account. Wilfred carr goes further, and claims that the focus of philosophers of education on conceptual analysis has led to a disconnect between philosophy of education and educational practice. The scientific, objective and value free focus of conceptual analysis has turned philosophy of education into an abstract pursuit with little value to those seeking guidance about educational practice. carr says that 'the problem of relating philosophy to education is deeply rooted in the conceptual foundations on which the contemporary philosophy of education has been erected and hence be nothing other than the entirely predictable outcome of a fundamental intellectual disorder internal to the contemporary condition of the discipline itself' (2004, p. 57). carr advocates postmodernism as the solution to this wrong turn in philosophy of education; essentially rejecting the value of Peters' analytic tradition completely.
In response to these critiques, I explain how conceptual analysis, conceived as ameliorative analysis (Haslanger, 2000), is socially and educationally valuable. Ameliorative analysis, as a methodology, explicitly considers the purposes of a concept while maintaining traditional standards of conceptual analysis. The slightly altered focus of ameliorative analysis or conceptual engineering leads to a methodology which is rigorous, practical and socially relevant at the same time. This is something that is needed in education, where policy and practice shapes individual lives and society.
conceiving of analytic philosophy as an ameliorative task counters some of the reservations held by Reid, Standish and carr about analytic philosophy of education. Re-aligning traditional analytic philosophy of education with ameliorative analysis also has implications for analytic philosophers of education, who should explicitly prioritise the role that concepts play in society and to individuals when defining and assessing concepts. Ameliorative conceptual analysis is by its nature rooted in normative concerns, the social world, the experiences of individuals, and practical questions; it can save analytic philosophy from accusations of being technical and academic.

Why educational concepts matter
There are many important educational concepts, and educational concepts are important because they influence educational practices, debates and attitudes. These, in turn, influence individual lives, and society. When analytic philosophers attempt to define educational concepts they are trying to clarify what we mean by ideas that are central to education. Educational concepts include 'education ' itself, 'intelligence' , 'knowledge' , 'inquiry' , 'justice' , 'equality' , 'teaching' , 'learning' , 'indoctrination' , 'belief' , 'morality' , 'schooling' , 'virtue' , 'flourishing' , 'well-being' , 'vocation' , 'skills' , 'rationality' , 'children' , 'state' , 'private' , 'interests' , and so on. All of these are terms which sit at the heart of debates in education and whose meaning has the potential to shape policy and practice.
Take the use of the term 'knowledge' in education. How this is conceptualised makes a difference to what happens in school, and what happens in schools makes a difference to children's lives, and to society as a whole. For example, in the UK, there is a 'Knowledge-Richcurriculum' movement. This promotes a curriculum that focuses on imparting knowledge to children. The trend is echoed in the United States by charter School groups such as the Knowledge is Power Programme, with its 255 schools, and Hirsch's Core Knowledge programme. It is also reflected in policy, with English curriculum reforms aiming 'to ensure that the acquisition of knowledge within rigorous subject disciplines is properly valued and cherished' (Gove, 2009, p. 17).
In this context, how 'knowledge' is conceptualised leads to different social outcomes. If knowledge is associated with scientific knowledge, then a resulting curriculum prioritises science subjects over arts and humanities subjects, and downplays educational goods such as character development, civic mindedness, or vocational skills. If a body of facts is seen as more important than inquiry, then the curriculum might turn to scripted lessons and memorisation, detracting from 'progressive' teaching styles. If knowledge is associated with traditional disciplinary canons, then the resulting curriculum might conflict with movements to decolonise the curriculum, potentially inhibiting social progress by excluding non-canonical ideas. All of these outcomes are in fact at play in the United Kingdom at the moment.
Similarly, how 'intelligence' is conceptualised affects which children are selected to enter selective schools such as private and grammar schools in the United Kingdom. These children are likely to take leading roles in society, to be more economically successful, and to be more able to shape the world according to their standards. How 'equality' is conceptualised shapes how resources are distributed in education; for example, which schools receive funding, whether funding is allocated outside of schools, whether parents are able to choose the sorts of schools that their children attend, and whether selective and private schools are permissible. Educational concepts matter because they shape the social world, and they affect the lives of individuals. They do this because they form the terms of the debate within which educational policy and practice are determined.
An analysis of an educational concept which does not explicitly focus on the role that concept plays in society, or the purposes that it serves, is missing an important part of the point of conceptual analysis. How a concept is understood can lead to social progress, but also social stagnation or damage. new analyses of concepts can lead to positive outcomes for individuals, but also harm. Educational concepts can be particularly wide-ranging in their effects on the social world. Schooling, for example, is possibly the largest (often state-run) intervention in many people's lives. Sally Haslanger's 'Gender and Race: (What) Are They?(What) Do We Want Them To Be?' (2000) is a seminal piece of work on conceptual engineering for the social world. Like questions about education, questions about gender and race are social questions. Like educational concepts, how we understand the concepts of 'gender' and 'race' has important implications for social institutions, practices, individuals and for society as a whole. Haslanger starts by outlining three approaches for answering the questions 'what is gender?' and 'what is race?' She calls these three different approaches: 'conceptual' , 'descriptive' and 'analytic' (she later calls 'analytic' approaches 'ameliorative').

Ameliorative analysis
A conceptual inquiry is the sort of traditional conceptual analysis pursued by Peters and his contemporaries. Here, 'one way to proceed would be to use the method of reflective equilibrium' (Haslanger, 2000, p. 33). This entails proposing definitions and testing them against counterexamples until a suitable definition of the word in question is found. This sort of inquiry takes our ordinary language and attempts to clarify it. Analytic philosophy of the 1960s took this approach, defining ordinary concepts with the aim of reaching clarity about their meaning. This approach is still valuable, as rigour and clarity are central to the ameliorative potential of any definition.
The second form of inquiry is descriptive; 'a descriptive project is not concerned with exploring the nuances of our concepts… it focuses instead on their extension. Here, the task is to develop potentially more accurate concepts through careful consideration of the phenomena, usually relying on empirical or quasi-empirical methods' (p. 33). This form of inquiry tracks the naturalistic turn in analytic philosophy, where real world phenomena are explored to narrow down the meanings of our concepts. Haslanger explains that a descriptive approach to 'a human right, a citizen, a democracy, might begin by considering the full range of what has counted as such to determine whether there is an underlying (possibly social) kind that explains the temptation to group the cases together' (p. 33).
Both conceptual and descriptive forms of inquiry in analytic philosophy are important, but the third is Haslanger's own contribution to what ought to also take place when considering socially important concepts. This third sort of inquiry is what Haslanger calls 'analytical' or 'ameliorative': on this approach the task is not to explicate our ordinary concepts; nor is it to investigate the kind that we may or may not be tracking with our everyday conceptual apparatus; instead we begin by considering more fully the pragmatics of our talk employing the terms in question. What is the point of having these concepts? What cognitive or practical task do they (or should they) enable us to accomplish? Are they effective tools to accomplish our (legitimate) purposes; if not, what concepts would serve these purposes better? (Haslanger, 2000, p. 33) With regard to gender and race, a traditional conceptual inquiry would reflect current ordinary usage of the terms 'gender' and 'race' . This analysis might be harmful, for example defining gender in terms of sex might harm transgender people and communities, or fail to call out harms experienced by people who do not clearly fall into one category or another.
Similarly, a descriptive approach to race or gender which tracked different intuitions and attitudes to race or gender across the world might conclude with concepts that aggravate stereotypes and injustice.
An ameliorative approach to analysis takes social issues seriously and prioritises them within conceptual analysis. concepts are not defined solely by the boundaries of ordinary language use, nor are they defined by the results of empirical investigations: 'the world by itself can't tell us what gender is, or what race is; it is up to us to decide what in the world, if anything, they are… My priority in this inquiry is not to capture what we do mean, but how we might usefully revise what we mean for certain theoretical and political purposes' (p. 34). Prior to any investigation, the ameliorative analytic philosopher begins 'by considering what we want the concept in question for' (Haslanger, 2000, p. 35).
Haslanger provides a solution to some of the criticisms faced by traditional analytic philosophers of education. Rather than accept that analytic philosophy of education is a technical pursuit with few real-world ramifications, or reject analytic philosophy of education in favour of a post-analytic or postmodern methodology, analytic philosophy can and should play an important role in shaping what we mean by key educational concepts with the aim of promoting social change. This can be achieved by considering what we want from a concept before beginning to map out a definition for it. Haslanger thinks that our conceptual analyses ought to serve our existing aims. She sees conceptual analysis as a form of activism, where philosophers can define concepts in such a way that they improve society. Traditional standards of conceptual analysis are still important because concepts need to be rigorously and clearly defined, however, normative concerns predetermine the field of conceptual investigation. This is helpful because it allows analytic philosophers to separate conceptual analysis from normative work; the normative work should guide conceptual analysis, but conceptual analysis remains a distinct methodology.

What are educational concepts for?
In this section, I will propose some very basic tests to check whether a concept has been analysed with its social purposes in mind. As a disclaimer, each educational concept can play many different roles, and which role in particular we want it to play, and so which resulting analysis will be best, differs depending on our ethical, political and contextual commitments. As such, I am not proposing a set of overarching educational aims that I think everyone should commit to. Instead, I am proposing three tests that are relatively uncontroversial, and can serve as a test for whether concepts have been considered in an ameliorative way.
I propose that at a minimum, there are three things that we should demand from any educational concept. At the very least, an educational concept (eg 'knowledge' , 'intelligence' , 'equality') should try to serve three purposes: (i) to be good for the individual being educated; (ii) to be good for society as a whole; and (iii) to be capable of yielding practical guidance for educators. If a defined concept meets these conditions, then the concept has been analysed with its social context and purposes in mind. It is worth noting that the additional step of determining what we mean by (i) and (ii) and how they interact is always going to be an important part of analytic philosophy of education, but will form part of the prior normative work. This normative work is a hugely significant philosophical task in its own right.
The first two of these tests have been chosen because without them, it seems hard to ascribe educational value to something. An action, institution, idea, practice, policy, school subject, etc, can have educational value in a variety of ways, but all of these ways are linked to some value for an individual or some value for society. This is reflected in Brighouse et al. (2016) language of educational goods. Here, educational goods span 'the knowledge, skills, dispositions, and attitudes that inhere in people and have the potential to contribute to their own flourishing and the flourishing of others' (2016, p. 5). Educational goods contribute to the capacity for economic productivity, personal autonomy, democratic competence, healthy personal relationships, treating others as equals, and personal fulfilment (p. 8). All of these goods either serve the individual being educated (contribute to their own flourishing), or serve broader society (contribute to the flourishing of others). If these educational goods did not serve either the individual being educated, or society, it seems that they would not be educationally valuable. At the same time, being good for the individual being educated is not sufficient for something to have educational value on its own, and neither is being good for society on its own. The goods on offer need to be weighed up against each other. nonetheless, (i) and (ii) can serve as a basic test for amelioration because without at least one or the other, it is hard to see how educational value has been considered at all.
Another way of thinking about these criteria is in terms of avoiding promoting harmful concepts. The first test (i) checks that our proposed educational concepts do not harm individuals being educated. For example, if the concept of 'intelligence' includes the ability to read and write fluently and thus excludes dyslexic students from being considered intelligent, then it is clearly harmful to dyslexic people being educated. If an educational concept harms the individual being educated, then it is unlikely to be a good educational concept. At a minimum it seems reasonable to hold that one of the purposes of education is to provide something good to the individuals being educated. If a concept did no good for most individuals, or actively harmed some individuals, then that concept may be in need of revision. Perhaps the social role of the concept and its potential real world implications were not prioritised.
The second test (ii) ensures that educational concepts do not harm society as a whole. For example, one conceptualisation of 'equality' which rests on equality of opportunity might allow for the existence selective schools where any student can apply, but only some will be accepted. This could (and does) harm society by excluding students from certain social groups from valuable or powerful roles in society. And this in turn can entrench social injustice. If an educational concept is harmful to society, then it does not serve a central purpose of education to improve society as a whole. Unless some normative reason for taking this approach has been explicitly argued for (maybe there is a clash with value to the individual), the proposed concept is likely to be in need of a revised ameliorative analysis.
Finally, (iii) if an educational concept is too vague or abstract to provide practical guidance, it is likely to need revision. For example, if 'knowledge' is conceptualised as 'justified, true belief' , this says virtually nothing about what ought to be included on the curriculum. There are too many justified, true beliefs to be included, and it is unclear whether we as a society, or educators as individuals have good access to which beliefs are justified and true, and which are not. other than narrowing the curriculum to exclude anything that is obviously false, justified, true belief is an impractical educational conception of knowledge. It is hard to understand how it might serve the purpose of influencing policy and practice in education. This criterion is shared with traditional methods of conceptual analysis and points to the importance of clarity. However, it also ensures that proposed concepts are directly linked to the social world; that they make a difference. Since rigour and clarity is still important, much existing work in analytic philosophy of education is still valuable when pursuing ameliorative analysis in education. Furthermore, there is room for traditional methods of analytic philosophy when it comes to evaluating the clarity of ameliorated concepts. Ameliorative conceptual analysis is does not break from the past, instead it links conceptual analysis to the social world.
If these tests are applied to analyses of educational concepts, then the resulting revisited concepts will aim to minimise harm to individuals and society, and at best promote positive social change with clear practical guidance, which is basically what we want educational concepts for. These tests work without the need to set out broader political or ethical aims for education beyond the basics of not causing harm to individuals or society and being clear enough to have practical implications. Any revised concept would, however, need to be guided by explicit normative theorising to account for the tensions between these three tests. It is worth bearing in mind that revisions to concepts might be ameliorative in the sense that they fix obvious problems, rather than revisionary in the sense that a whole new conceptual scheme is built from scratch. Since existing standards of conceptual analysis are still important (concepts should be logically coherent, clear and precise), ditching existing analyses is unlikely to be worthwhile. I am not proposing a new discipline, I am proposing an adjustment to conceptual analysis in education with an explicit focus on social aims.
Ameliorative analysis can link philosophy of education to practical issues and questions of value; something that Reid, carr and Standish say analytic philosophers do not prioritise enough.

The concept of 'education'
It would be unfair to accuse Peters of descriptively analysing the concept of 'education' without any concern for the social world. Peters and Hirst claim that 'we will not measure the success of conceptual analysis by the extent to which we can produce definitions. Rather we will measure it by the extent to which our understanding is thereby increased about how things are in the world and of the possible stances that we can adopt towards our predicament in it' (1970, p. 12). Furthermore, they recognise that there are 'practical purposes embodied in the concept of 'education'' (p. 15). despite Peters' awareness that the concept of 'education' is socially important and has practical relevance, he primarily engages in what Haslanger would call a conceptual inquiry, searching for a definition of education despite his reservations about whether a definitive account is possible. He does not explicitly prioritise the purpose that concepts can play in society, or start with a developed normative compass.
Peters' 'normative aspect of education ' (1966, p. 25) is present, but minimal: he says that like the concept of 'reform' , 'education' has 'the criterion built into [it] that something worth while should be achieved' (p. 25). He then launches into a traditional conceptual analysis. He says that education involves some form of 'task achievement' (p. 26) so that someone who has been educated has achieved something worth while. Peters continues that whatever it is that an educated person achieves, it is intrinsically worthwhile. He makes this move by suggesting that if the aim is extrinsically or instrumentally worth while, the concept of 'education' merges with the concept of 'training' , and someone who has been trained to do something worth while would not be considered educated (presumably this rests on the ordinary language distinction between education and training). He uses the example of training someone to act responsibly, as opposed to genuinely reforming them to illustrate this (p. 27). Finally, 'education' has a 'cognitive aspect'; an educated person should have (a) knowledge and understanding, and (b) cognitive perspective, where they know and understand a broad range of activities. In summary, Peters' analysis of education is that ''education' … involves the intentional transmission of what is worth while' in terms of knowledge and understanding of a broad range of activities' (p. 35).
Peters proposes a definition of education and considers cases that do not seem to fit with the ordinary usage of the word, thus engaging in reflective equilibrium as a form of inquiry. Peters searches for necessary and sufficient conditions for 'education' through considering potential definitions of 'education' and asking whether they can accommodate all examples of how we ordinarily think of education, and exclude all examples of activities we would not ordinarily think of as education, such as training.
Peters' concept of education as the intentional transmission of what is worth while in terms of knowledge and understanding of a broad range of activities can be assessed against the tests suggested for ameliorative analysis of educational concepts. These ask whether Peters' concept of education is (i) good for the individual being educated (ii) good for society as a whole, and (iii) capable of yielding practical guidance for educators.
i. Is Peters' concept of education good for the individual being educated? The educated person, according to Peters has a broad knowledge and understanding of intrinsically worth while activities. Since there are a vast number of intrinsically worth while activities, whether or not this sort of education is good for the individual depends on what these are. Peters' view was that these intrinsically valuable activities cover more or less traditional school subjects such as 'science, history, mathematics, religious and aesthetic awareness, together with moral, prudential and technical forms of thought and action ' (1966, p. 50). Peters holds that these particular intrinsically valuable activities broaden the mind and are endlessly engaging. While it seems fair to conclude that for many individuals, an education into traditional school subjects is not harmful and has certain benefits, it is unclear whether these are the best activities for an individual to engage in at school. This argument is made by White in his analysis of Peters' transcendental argument (1973). depending on the needs and interests of the individual, perhaps they would do better to spend their day perfecting their sporting abilities, or spending time with their family at home. depending on the society they live in and their social circumstances, perhaps it would be better for the individual to become a trained accountant, or to learn how to farm their family's land. Furthermore, some students are going to struggle to engage with an academic curriculum which might seem distant from their own interests and concerns; these individuals might leave school as 'uneducated' , a label which is likely to restrict the possibilities open to them in their lives and so harm them. Peters presents a narrow account of what is best for an individual being educated.
Arguably, it will not be the best for all individuals; it might harm some. ii. Is Peters' concept of education good for society as a whole? overall, it seems that a society where everyone has been educated in a range of academic activities is not particularly bad, but neither is it ideal. There is an elitism in Peters' concept of education which reflects the idea that someone with a broad academic education fits the mould of the 'educated man' , or gentleman. They are able to converse with other educated men and progress in society through accessing further education and highly paid jobs. While this is good for the individual, it is not good for society as a whole. It presents a vision of a society where only certain types of people, and certain sorts of achievements are valued. There are always going to be individuals who will not meet the standards of Peters' 'educated man' , and these people will be left with very little to show for their education. Furthermore, the social aims of education are often seen as involving purposes like preparation for participating in democratic processes, or economic participation. Peters' concept of education is not focused on these purposes and so only contributes to them tangentially. Just as importantly, a society which values the 'educated man' becomes blind to insights from other groups such as women, or ideas which are not equated with the white Western canon (see code 2014). Insights from these groups might help to shape a more diverse, knowledgeable and rich society; they would lead to social progress. Peters' concept potentially hinders this sort of progress. iii. Is Peters' concept of education capable of yielding practical guidance for educators?
Peters provides some practical guidance; the curriculum ought to be broad and academic. More detailed guidance comes from Hirst who developed an account of the liberal curriculum based on six or seven forms of knowledge. This is very clear in its curriculum guidance and aligns directly with Peters' analysis of the concept of education (Hirst, 1974). Hirst's curriculum guidance points to a set of subjects which ought to be taught in schools in order to constitute an education. In this sense, Peters' concept is a good one, it yields practical guidance. However, this potentially compounds the problems Peters faces with the first two criteria because the practical guidance provided means that Peters' concept has shaped education, influenced individuals and changed society; and not necessarily for the better. For example, the UK conservative government's interest in liberal education cites Hirst's work (conway, 2010).
The importance of educational concepts is clear: they can influence policy, they can underly social practices and they can shape the way the world around us looks. Haslanger points out that 'practices shape us as we shape them. This provides resources for understanding why social practices tend to be stable, but also reveals sites and opportunities for change. (challenge social meanings! Intervene in the material conditions!)' (Haslanger, 2018, p. 231). This urges analytic philosophers of education to set out to change society by ameliorating damaging concepts, or proposing normatively guided analyses which promote social progress. Before setting out to analyse the concept of 'education' , philosophers of education should consider what they want from a concept of 'education' . The same can be said for any other educational concept, such as 'knowledge' , 'intelligence' , 'equality' , or 'rationality' . These concepts, if adopted, have the potential to alter the social world. This places a responsibility on analytic philosophers to work for the good of individuals and society, and to employ clear normative compasses when undertaking their analytic work. It also highlights a reason why conceptual analysis is valuable: it can improve the social world.

Three interesting points
Three interesting points derive from taking ameliorative analysis in education seriously. These are: (1) The canon of traditional conceptual analysis in education is ready for renewed examination and revision; (2) It is likely that there is a distinct set of educational concepts which may or may not overlap with concepts from other areas of philosophy; and (3) We can identify criteria to guide ameliorative analysis for educational purposes. I am going to start with (2) because it has not been discussed yet, and because it draws the other two into sharper focus. These points are proposals for areas of discussion, they are not secure conclusions.
An ameliorative analysis of a concept is driven by the purposes that we want that concept to play in society. This means that ordinary concepts can be analysed differently according to the different purposes they are designed to serve, and these purposes will differ in different fields. An educational concept of knowledge needs to be able to guide the curriculum, to provide students with something that is good for them, and to improve society as a whole. The concept of knowledge, as it is used in the biological sciences, needs to pick out theories that lead to non-accidental biological progress in terms such as theoretical fruitfulness or technological advances. The epistemological concept of knowledge needs to distinguish knowledge from belief, truth from falsehood, and justification from unfoundedness. There is no reason to think that biological knowledge is the same concept as epistemic knowledge, or educational knowledge, although it's reasonable to expect them to overlap.
Eder makes this point about the concept of rationality. She argues that 'the normativity of rationality and the intended purpose, or goal, of the theory of rationality be taken into account when explicating the concept of rationality… Given the differences between such purposes, the method of explication that I put forward allows for different characterisations of rationality and, thus, for epistemic pluralism concerning rationality ' (2021, pp. 4976-4977). concepts need to be engineered to suit the purposes they are needed for. Eder identifies two potential purposes for the concept of rationality within the field of epistemology: a theory of rationality is needed to guide the formation of doxastic states; and a theory of rationality is needed to assess doxastic states. Someone might be rational because they form beliefs well, or their beliefs might be rational because they conform to standards of rationality. These are two different purposes for the concept of rationality which are best solved by two distinct concepts of rationality.
When it comes to education, there is no reason why existing concepts from other fields of philosophy ought to be the best at achieving educational aims. The epistemological concepts of rationality discussed by Eder are not necessarily good educational concepts. They might not yield clear practical guidance for educators, they might lead to attitudes and practices that are bad for the individuals being education, or they might be bad for society as a whole. If analysis is guided by the purpose of the concept, then educational concepts are likely to be distinct from non-educational concepts. We should expect that rationality means something different in education than in epistemology, and we should expect to need to engineer concepts that serve the purposes we want them for.
This leads to (1), that conceptual analyses in education should be revisited and concepts may need to be revised or rejected if they fall short of the tests I have proposed, or if new purposes are needed from concepts. First, our understanding of the purposes of education changes as society changes and it is very likely that what Peters considered the primary aims of education in 1966 are different from our current priorities. Ameliorative analyses of educational concepts are never final, they demand attention to ensure that they suit a changing social world. Second, any concept which has not been analysed in a way that prioritises social factors is going to need to be revised because concepts have real-world effects and a harmful concept is a bad concept. Finally, while concepts from other areas of analytic philosophy such as epistemology, political philosophy, or ethics are fruitful places to turn to for guidance about educational concepts, they are unlikely to offer up ready-made educational concepts. This is because the purposes of the analyses offered are not necessarily educational purposes.
The remaining point is (3), that we can identify tests to guide the need for ameliorative analysis in education. Three tests were outlined earlier, and provide guidance about the minimum that is needed for an educational concept. An educational concept should lead to something that is (i) good for the individual being education, (ii) good for society as a whole, and should be (iii) capable of yielding practical guidance. These tests are proposed to check whether concepts are potentially positive educational forces, or open to amelioration.
In this respect, analytic philosophy of education necessarily requires a prior normative task to determine the aims of educational concepts. Perhaps we want a concept of knowledge so that students are protected from lies or indoctrination, perhaps the purpose is to provide students with information that will help them to live goods lives, or perhaps the concept of knowledge is important because we want students to be able to inquire into the nature of world for themselves. These questions about what we want a concept of knowledge for in education are legitimate grounds for independent prior debate, and constitute a normative area of analytic philosophy aside from conceptual analysis itself. depending on the answer, the purposes that should be taken into account in an ameliorative analysis of the educational concept of knowledge will differ. So long as these purposes are made explicit when analysing educational concepts, the methodology of ameliorative analysis remains clear, transparent and rigorous while fully engaging with the complexities of the social world and normative aims.
Ameliorative analysis of educational concepts thus comes with a methodology. The methods of traditional conceptual analysis are still important: the analysed concept needs to be logically coherent, it needs to be able to account for examples and counter-examples and to adequately delineate what we mean when we use the concept. Additionally, prior normative work on the purposes of the concept in the context of education is needed, and then the concept arrived at should serve those purposes. The aims of education, lived experiences of the social world, and political and ethical commitments are central to ameliorative conceptual analysis. Ameliorative conceptual analysis is by no means a rejection of traditional conceptual analysis, turning to it provides a re-statement of the aims of social analytic philosophy and makes its value as a methodology in education studies clearer.

Conclusion
I have argued that analytic philosophy of education is due an ameliorative turn. Traditional analytic philosophy of education has not always been explicitly guided by the impact that educational concepts have on the social world. Educational concepts are social concepts in that they can alter how individuals are treated, and can change society as a whole. This means that when philosophers of education analyse educational concepts they ought to consider the purpose of those concepts, and how concepts might be analysed so that they can lead to positive changes for individuals and to society, guided by prior normative theorising. This brings clarity to the methodology and value of analytic philosophy of education: It can change the social world for the better through careful, socially responsible, conceptual analysis. An ameliorative focus in analytic philosophy of education would counter concerns that it sits outside of real-world concerns, experiences and social progress. concepts can shape society, and ameliorative conceptual analysis which serves clear social aims can lead to positive social change. This serves as a clarification for the value of analytic philosophy as a method of inquiry in education studies.

Disclosure statement
no potential conflict of interest was reported by the author.

Funding
This study was supported by the Economic and Social Research council (ES/W006146/1) and Society for Applied Philosophy.

Notes on contributor
Jane Gatley is a lecturer in education at Swansea University. Prior to this, she held an ESRc postdoctoral fellowship at Birmingham University, where she also completed her Phd. In a previous life, she worked as a Religious Education teacher in secondary schools in England. Her work is in philosophy of education, and the role of philosophy in education.