Expanding Davies’s Pragmatic Constraint: A Pragmatist Principle for Philosophizing about Art

I. Introduction

In Art as Performance, and in his subsequent papers ‘The Primacy of Practice in the Ontology of Art’ and ‘Descriptivism and Its Discontents’, David Davies argues for a methodological principle that would govern the ontology of art, which he calls the ‘pragmatic constraint’.¹ This principle aims to formalize the commonly accepted view that the objects posited by an ontological theory of art should be the kinds of thing that feature in the practices of artists, critics, audiences, and so on; in other words, that the things philosophers of art theorize about are the things that artists make, critics evaluate, and audiences encounter and appreciate. Although this basic idea is generally endorsed, the principle as Davies formulates it has not been universally accepted, in part because it seems bound to lead to disagreements that come down to differing intuitions.²

In order to avoid this type of impasse while preserving the insight that philosophers’ uses of the term ‘art’ should not fail to map onto any of the things that feature in artistic practices,³ I argue for an alternate methodological principle that is expanded to cover not only ontology but all philosophizing about art while remaining compatible with Davies’s constraint, insofar as views in the ontology of art that would meet Davies’s constraint will also meet the proposed principle. This principle is also ‘pragmatic’ – arguably, more philosophically so – insofar as it draws on two foundational thinkers in the American Pragmatist tradition, William James and John Dewey.

I begin by summarizing Davies’s pragmatic constraint and consider the worries that have been raised for it, including the controversy, emerging from a debate between Davies and Julian Dodd, over how to determine whether a view or argument in fact adheres to this constraint. I then look to ideas from classical pragmatism as the basis for a new methodological principle that, for the sake of clarity, I refer to as the ‘pragmatist principle’ to distinguish it from Davies’s constraint. Finally, I consider some advantages of and possible worries about this pragmatist principle and its adoption as a methodological standard for the philosophy of art.

II. Davies’s Pragmatic Constraint

Davies introduces his constraint by contending that ‘“artistic practice”, properly construed, must serve as the touchstone for our philosophical theorizing about art’, and that, when we are trying to answer ontological questions about the kind of entity artworks are, ‘the default assumption must be that those things treated as artworks in our artistic practices are indeed artworks’.⁴ He specifies that by ‘artistic practices’ he means ‘the ways in which we treat and characterize the things we term “artworks” in our critical and appreciative engagements with the arts’ and insists that ‘claims about the ontological status of artworks are importantly constrained by those features of our creative, critical, appreciative, and individuative practice in the arts that would withstand rational scrutiny’ or would be ‘upheld on rational reflection’.⁵ This leads to his pragmatic constraint, which holds that

Another way Davies puts this point is by saying that positions in the ontology of art must conform or be ‘answerable’ to our best considered answers to questions in the epistemology of art; that is, questions about how artworks are known or apprehended and how they are properly understood and appreciated.⁷ For example, if apprehending and evaluating artworks require one to have first-hand acquaintance with those works, and if aspects of an artwork’s history are relevant for properly understanding and appreciating it qua art, then the kind of entity an artwork is must be one that can have historical properties and that people can be directly acquainted with.

One worry is that it is not clear which critical and appreciative practices should be appealed to or reflected on. Just whose ‘rational reflection’ is to determine which properties are rightly ascribed to the things that are treated as art in these practices, or how they are properly understood and appreciated, is similarly left open. Robert Stecker raises a version of this worry, noting that the interpretive and appreciative activities of recognized artworld ‘experts’ – professional art critics, exhibitors, art historians, and so on – are ‘informed by widely divergent theoretical perspectives’, so there is no stable or uniform convergence among practitioners that could, after suitable reflection, be taken to establish a single standard to which ontological theories of art are answerable.⁸ Moreover, given the plausibility of the view, argued for by Morris Weitz, that art is an inherently ‘open’ concept and so the set of things that count as artworks is always open to expansion,⁹ we might also worry that there will be widely divergent creative practices to contend with in addition to critical or appreciative ones, with different creative practices potentially characterizing and treating ontologically different kinds of things as ‘art’. This makes the problem greater than Stecker acknowledges, since a wide diversity of creative practices – and so, of things characterized and treated as ‘art’ therein – will only increase the divergence of theoretical perspectives held by critics.

This diversity of creative, critical, and appreciative practices, along with the things characterized and treated as ‘art’ in these practices, raises another worry. Davies’s call for ontology of art to be answerable to our best epistemological views about art might be thought to risk circularity if ontology of art is understood to encompass such questions as ‘by virtue of what does something count as an artwork?’ or ‘what distinguishes art from non-art?’ This is because the selection of a set of things or practices to be the object(s) of the epistemological reflections that are taken to ‘constrain’ ontology already presupposes some view of what things count as art, and which practices count as artistic ones. In other words, it is not obvious that what Davies calls epistemology of art is wholly separable from ontology of art such that epistemological views can be logically prior to views about the existence and nature of artworks.

Davies addresses these worries in his later papers but not, I think, in a way that entirely removes ambiguity from his pragmatic constraint regarding what exactly the ontology of art is answerable to, and thus, how his principle is to be applied. In response to Stecker, Davies clarifies that it is not practitioners’ own reflections or beliefs about art that he takes ontology to be answerable to.¹⁰ Were that the case, it would amount to a ‘strong descriptivism’ according to which philosophers would be limited to taking artists and critics at their word with respect to issues and questions in the epistemology of art, leaving philosophers only to identify which ontological kind maps onto or best explains what practitioners already say about the objects of their practice, with no room for philosophers to critically analyse, disagree with, and correct or improve how practitioners understand their practice and its objects at a given time.¹¹ However, critically analysing and possibly improving how people understand the world, including how they understand what they do, surely falls within the legitimate scope of philosophy. Against this worry, Davies clarifies that what he is advocating is for philosophers’ own analyses of, and reflections on, the activities of practitioners in the arts (including critical reflection on, rather than mere acceptance of, practitioners’ beliefs) to precede and inform ontological theorizing about art.

It is relevant to the worry about possible circularity that Davies restricts the scope of the pragmatic constraint to ontology in a narrow sense, being exclusively concerned with what ontological kind(s) artworks are – for instance, whether artworks are physical objects, eternal forms, temporally extended gestalts, performances, and so on – and not to questions such as which things count as artworks, and by virtue of what they so count.¹² It is not clear that views in the epistemology of art will necessarily presuppose answers to ontological questions in this narrower sense, or at least not in the way that they presuppose answers to definitional questions, so this worry also seems to be diffused.

Even if Davies’s response disarms Stecker’s version of the worry by showing that he does not take ontology of art to be answerable to practitioners’ beliefs about what they make and do, it still does not tell us which philosophical view(s) of the individuative, appreciative, and creative practices of artists, critics, and so on we should hold ontology of art answerable to. Moreover, we can hardly expect to find a more stable and uniform convergence of views among philosophers than we can in the views of artists or critics. Davies’s principle does not tell us how, given the divergence of philosophical views, we are to determine which epistemological ones are the ‘best’ upon rational reflection: for instance, which ascriptions of properties to artworks are the ‘right’ and ‘reasonable’ ones, or which accounts of what it is to understand or appreciate artworks are ‘proper’.

In short, because the pragmatic constraint, on Davies’s formulation, does not give us a way to determine precisely what it is that ontology of art must be ‘answerable’ to in order to follow the pragmatic constraint, and because disputes about epistemological matters concerning art are not obviously any easier to resolve than ontological disputes, Davies’s formulation leaves it ambiguous as to exactly when this constraint is or is not being followed, which weakens its methodological usefulness. We might say that the pragmatic constraint, on this formulation, is either incomplete as an applicable methodological principle or that it merely ‘passes the buck’ from ontology to epistemology – where ‘buck-passing’ accounts are only helpful and informative when the area of inquiry to which the buck is passed is less controversial than the inquiry from which it is passed. If this is right, Davies’s version of the pragmatic constraint would seem not to be adequately informative or applicable as a methodological guideline without some further principle for distinguishing between better and worse views in the epistemology of art.

III. The Davies–Dodd Dispute

An example of this ambiguity concerning when the pragmatic constraint is being followed is found in the disagreement between Davies and Julian Dodd over whether Dodd’s argument for what he calls the ‘simple view’ of musical works adheres to this constraint. Dodd takes himself to be starting from a reflective consideration of some central aspects of critical, appreciative, and individuative practice involving music: namely, that musical works are commonly treated as audible, as repeatable, and as created by their composers. He argues that their audibility and repeatability are best explained by a type-token theory according to which musical works are types of sound sequences of which performances are their tokens. However, he takes this to be in tension with the idea of musical works being created since, he argues, types are abstract entities and metaphysicians generally agree that abstract entities are eternal and thus not creatable. This leads Dodd to conclude in favour of the simple view, which can be described as a kind of musical platonism since it holds that musical works, as types, are eternal and so are discovered by their composers rather than created, where these sound-sequence types set the norms by which their tokens (for example, performances of a given work) can be judged to be properly or improperly formed, and by which two performances can be judged to be of the same work.¹³

Because Dodd starts by considering how musical works are treated in practice and then brings metaphysical considerations to bear on his analysis of this treatment, one might think – as Dodd himself does¹⁴ – that his argument follows the pragmatic constraint. Rather than taking Dodd to begin with an ontological view that he then imposes onto elements of musical practice, we might characterize his approach as him saying ‘let’s look at some of the prominent ways that musical works are treated in practice and the properties thereby ascribed to them, and see which ontological category best accounts for the greatest number of these’ – where, presumably, if there were a single ontological category that could coherently accommodate all these practices and properties, Dodd would prefer it.

Davies, however, gives two reasons for thinking Dodd’s methodological approach is not in accord with the pragmatic constraint. One is that Dodd does not first ‘appeal to broadly epistemological considerations’ prior to arguing for his ontological view,¹⁵ where Davies notes that the order of the discussion in work that he takes to follow the pragmatic constraint – here he names Gregory Currie, Jerrold Levinson, Guy Rohrbaugh, and himself – is epistemology first, ontology second.¹⁶ But, Davies writes, ‘Dodd […] only turns seriously to such epistemological matters in the final chapters of his book’, where ‘[a]n evaluation of the specific judgments that we make about particular works is postponed until the basic ontology is already in place’.¹⁷ However, it is not obvious why the order in which points are discussed should be relevant to whether an argument follows the pragmatic constraint. The constraint, as presented in both Art as Performance and ‘The Primacy of Practice in the Ontology of Art’, says nothing about the order in which points must be presented but only that the conclusions of arguments in the ontology of art must ‘conform to’ the features of artistic practice that would be upheld on rational reflection. It is unclear why it would be incompatible with this constraint for one first to outline an argument for one’s ontological view and then discuss that view’s epistemological entailments, so long as these entailments were in line with whatever features of artistic practice would be upheld on rational reflection. It would, of course, go against the pragmatic constraint if what a certain ontological view entailed about how works ought to be individuated, evaluated, and so on did not hold up in light of rational reflection on the features of artistic practice, but one might think the opposite to be a virtue for an ontological view; that is, that it would count in favour of a view if what it entailed about these matters did happen to align with what an independent rational reflection on how these matters are handled in artistic practice would uphold.

One might think Davies’s concern is not merely with the order in which the points in Dodd’s argument are presented but with Dodd’s use of ontological premises to argue for epistemological conclusions. For instance, consider where Davies writes that, for the authors he names as adhering to the pragmatic constraint, the conclusions to their epistemological arguments concerning ‘what bears on the proper appreciation of artworks […] are essential premises in [their] arguments for ontological conclusions’.¹⁸ The opposite – ontological conclusions featuring as premises in arguments on epistemological matters – occurs in Dodd’s appeal to the type-token theory to discount certain properties from being properly ascribed to works of music: Dodd writes that ‘[o]nce one appreciates that works of music are abstract objects, ascribing such properties to them becomes a strange business indeed’.¹⁹ Although this worry is less trivial than taking issue with the order in which points are discussed, it also does not obviously go against the pragmatic constraint as Davies formulates it since that constraint does not say anything about which sorts of premise are legitimate to use in which kinds of argument, where it seems possible for an ontological argument not to contain epistemological premises but nevertheless to conform to whatever features of artistic practice rational reflection would uphold, and where an epistemological argument that contained specific ontological views in its premises might also conform to these features.

The other reason Davies gives for claiming that Dodd is not following the pragmatic constraint is that, while Dodd does consider some features of how musical works are treated in practice, he does not take enough features of the practice into account. For Davies, Dodd’s consideration of how musical works are treated as audible, repeatable, and created does not sufficiently ground his argument in practice; not sufficiently, that is, for him to discount creatability just because it is incompatible with the ontological theory he thinks best accounts for the other two features. As Davies insists, the pragmatic constraint ‘requires that we take account, in doing ontology, of our practice as a whole, rather than focusing on certain [features of practice] to the exclusion of others’, where this includes ‘tak[ing] account of the goals it is reasonable to ascribe to that practice, the values reasonably posited as sought within it, and the values reasonably ascribable to the works that enter into it’,²⁰ and he faults Dodd for not considering these things when developing his ontological theory. Artistic practices must be considered as a whole, Davies argues, because he takes philosophical interest in art to arise from our encounters with practice and to aim primarily at explaining and understanding practice, including the objects that feature within it. Thus, he writes, ‘[o]ntology must confront our practice taken holistically; we cannot discount prominent aspects of that practice on the basis of an ontology that is established merely by reference to a selective sampling of some of its features’.²¹

If what is meant by the need for ontologists of art to account for practice ‘as a whole’ or ‘holistically’ is that an incompatibility between a particular ontological theory and any feature that would be included, after rational and critical reflection, in the best unified conception of this practice will count against that theory, then it seems right as a regulative ideal (assuming that the unity and coherence of theories is desirable). However, this does not seem to be Davies’s criticism of Dodd since Davies does not specify which features of musical practice other than creatability – of which Dodd does take account – the ‘simple view’ is incompatible with; rather, the criticism seems to be that the number of features of musical practice that Dodd discusses is not enough. It is unclear, though, how many features of practice someone would need to discuss in order for their theory to be sufficiently grounded in practice so as to adhere to the pragmatic constraint. If the talk of needing to account for ‘practice as a whole’ is meant to imply that one would need to discuss every feature of an artistic practice in order to meet the pragmatic constraint, this would be implausibly demanding. There is not obviously an exhaustive list of all the features of an artistic practice that someone could be expected to account for and discuss when making an ontological argument. Even if there were, it is likely that at least some features of the practice would be trivial or otherwise irrelevant for considerations of ontology, where the question of which features were necessary to account for in order to meet the pragmatic constraint would itself be open to debate.

Moreover, it is not clear how this talk of ‘practice as a whole’ fits with Davies’s desire to avoid a strong form of descriptivism; that is, how it fits with his allowance that some features of an artistic practice may be revisable or may come not to be endorsed by philosophers after critical scrutiny and reflection. As noted, Davies clarifies that his constraint ‘permits us to reflect on practice when elements in it seem to conflict with one another or to require more perspicuous description’ and that it ‘allows for revisions not only in our beliefs about our practices and the things that enter into them, but also in that practice itself’.²² As such, the fact that Dodd ends up discounting some common aspects of how musical works are treated by critics and appreciators – namely, creatability and the relevance of the historical context of a work’s composition for its proper appreciation – in favour of others – namely, audibility and repeatability – will not itself discount him from having followed the pragmatic constraint. As Dodd asks, if the pragmatic constraint ‘has the normative, rather than purely descriptive, character Davies presents it as having, then what is to stop [a theorist from] claiming that [certain] properties’ that get attributed to musical works in practice, such as creatability, ‘are not “rightly” ascribed to works?’²³

In the absence of a more precise specification of which, or how many, elements of artistic practice must be accounted for in order to meet the pragmatic constraint, it is not clear what basis there is for declaring Dodd’s approach to be methodologically suspect with respect to the pragmatic constraint (although there may well be other reasons to object to his argument). Or, rather, it is unclear unless it is because Dodd’s understanding of the elements of musical practice he considers and how they relate to one another – that is, the conclusions he draws from his ‘broadly epistemological’ reflections – is not the understanding of musical practice that ultimately would be upheld upon rational reflection.

This leads us back to the worry about tying the pragmatic constraint to the right or best views in the epistemology of art, since it is not clear how these are to be determined, especially in the face of disagreement over how various elements of artistic practice are to be weighed against one another in reflective equilibrium. A consequence of this being built into the constraint is that disputes over whether the constraint is being followed will ultimately revert to disputes over which views in the epistemology of art are the correct ones. Not only is this unhelpful in resolving the dispute over methodology but it departs from the goal of the pragmatic constraint, which is to keep ontology grounded in practice by ensuring that ontologists’ explananda are the same things that artists typically make, audiences typically appreciate, and so on, since it would seem ultimately to make ontology answerable not to practice but to philosophers’ epistemological reflections and theories, with nothing to guarantee that philosophers’ epistemological explananda are the things that feature in artists’ and critics’ practices.

What is needed to avoid these worries and to be able to determine whether a purported appeal to practice in support of an ontological view does in fact accord with the pragmatic constraint is a further methodological principle that would apply to philosophers’ appeals to practice more generally. Because such a principle would apply to the epistemology of art as well as to ontology, it would better serve the goal of ensuring that the objects referred to in philosophical theories of art are the same things that are made by artists, apprehended by spectators, and so on, whereas the current formulation of the pragmatic constraint only ensures that the objects referred to in ontological and epistemological theorizing about art will be the same. This further principle would need to strike a balance between being more directly connected to what it is that practitioners do, make, appreciate, and so on, while avoiding a strong descriptivism that would make philosophy answerable to what practitioners believe or say about what they do. In what follows, I argue that such a principle can be formed by drawing on ideas from two paradigmatically pragmatic thinkers in philosophy: William James and John Dewey.

IV. Expanding the Pragmatic Constraint via Classical Pragmatism

In Some Problems of Philosophy, published posthumously in 1911, James discusses the relation between what he calls ‘percepts’ and ‘concepts’, where much of his discussion focuses on a problematic tendency of thought he calls the ‘abuse of concepts’. As an empiricist, for James all concepts arise from and trace back to percepts, which are immediate affective and sensory experiences of the phenomenal world. Intellectual activity, or thinking, involves substituting concepts for percepts and relating them by noting similarities and differences, drawing inferences, and so on.²⁴ The abuse of concepts results from merely assuming, without checking, that the relations we draw between concepts in our thinking about them will also hold of the objects and experiences from which these concepts arose. As James puts it, this amounts to ‘assimilating [the] concrete terms [of] a mass of perceptual fact […] one by one, to so many terms of the conceptual series, and then […] assuming that the relations intuitively found among the latter are what connect the former too’.²⁵

James argues that the way to avoid abusing concepts is to hold theoretical conclusions answerable to future experience by checking whether they do, in fact, map onto and cohere with the experiences we have when acting under the guidance of these conclusions, instead of taking ourselves to be justified in assuming our conclusions to be true of the things we are thinking about simply because they work ‘in theory’. In other words, the content and results of our thinking must both arise from past percepts and tie back into future percepts in order to count as a legitimate use of concepts rather than their ‘abuse’. This idea of thinking needing to start from and terminate in experience is also found in chapter one of Dewey’s Experience and Nature, where he argues for an empirically minded methodology for philosophy in which philosophy begins and ends with directly experienced subject matter.²⁶ Dewey proposes a ‘test for the value of any philosophy’, which is to ask: ‘Does it end in conclusions which, when they are referred back to ordinary life-experiences and their predicaments, render them more significant, more luminous to us, and make our dealings with them more fruitful?’ – as opposed to philosophical theories and conclusions being ‘left to dwell in separation in some technical realm of their own’.²⁷ He goes on to argue for the value of this principle, writing that

Following Dewey and James, I propose a methodological principle for the philosophy of art, one that might also be called a ‘pragmatic constraint’ owing to its basis in the pragmatist tradition and not merely because it appeals to the idea of practice. This principle would hold that philosophizing about art should both start from and tie back into practices of creating or engaging with art as these are concretely experienced. A more formal statement of the principle would hold that, in order to ensure that philosophers of art are talking about the same ‘things’ that practitioners create, appreciate, and so on, any view or conclusion argued for in the philosophy of art should (1) work as a possible answer to a question that could arise for a philosophically reflective practitioner from their engagement in an art-related practice, where this includes both the creation and the reception of art, and (2) potentially inform and make some fruitful difference for how that practice is pursued, for example, for how art is made or engaged with.

By a ‘philosophically reflective practitioner’ I mean one who is intellectually curious about the nature of their practice and its ‘objects’ and how these relate to other practices, areas of life, dimensions of experience, and so on, but not necessarily one who is formally trained in philosophy or who applies preformulated ideas from other areas of philosophy or other kinds of theory (for example, psychology, critical theory, and so on) to their practice.²⁹ Specifying this is important to ensure that questions arise ‘organically’ from the practice itself and to avoid philosophers importing established ideas from some other area of philosophy and seeing artistic practice through the lens of these ideas, since questions that arise for one area of concern (for example, philosophy of language) may not apply to or be relevant for another area of concern (for example, philosophy of art).

It might still be thought that this principle will not guarantee that the questions philosophers discuss and attempt to answer will be ones that genuinely or ‘organically’ arise from practice, since pseudo-questions can arise even for philosophically reflective practitioners. The principle’s second condition takes care of this worry, since a conclusion that is meant to answer a pseudo-question will not ‘tie back into’ the practice in the sense of usefully informing or making a helpful difference for how this practice can be conducted; if it did, that would suggest that the question it was answering was not a pseudo-question after all, since it seems right to think that, if an answer is useful in some practical endeavour, the question to which it is an answer is one that could arise in the pursuit of that endeavour.³⁰

How, it might be wondered, would the pragmatist principle be applied in (philosophical) practice? If the conclusion to a philosophical argument about art does not meet both conditions, we could say that it does not apply to the actual objects and activities that feature in artistic practice and so is misguided if it is meant to be a conclusion about art, or that the argument to which it is a conclusion – and perhaps the larger discussion in which the argument features – is misdirected and so not a fruitful area of inquiry for philosophers of art to be thinking about and working on – at least not qua philosophers of art. However, this does not mean that the argument and the broader discussion are not ‘properly philosophical’, since they may be answering questions that legitimately arise for other areas of philosophy; it only means that such arguments and discussions do not centrally belong to the philosophy of art but to these other areas. For example, arguments concerning the semantics of the language used to talk about art may be deemed to properly belong to the philosophy of language rather than to the philosophy of art, just as arguments concerning what it is for a mathematical formula to be beautiful properly belong to aesthetics rather than to the philosophy of mathematics.

The pragmatist principle calls for philosophers of art to engage more with artists and art practices, including art criticism, art history, and committed ‘connoisseurship’, so as to be receptive to the kinds of philosophical question that can arise from these practices and be in a position to judge whether the questions that other philosophers are asking could plausibly arise from practice. This engagement might be first-hand, with philosophers of art themselves engaging in these practices with some degree of competence so as to understand them from the inside, or it could be second-hand, with philosophers engaging with the first-hand experiences of practitioners through, for example, sustained conversation, reading artists’ statements and interviews, or collaborative inquiries. Whether first- or second-hand, this engagement must be done in a robust way that goes beyond the casual consumption of artworks, the casual reading of artists’ statements, and so on. It should also go beyond simply looking to artistic practice for answers to questions that arise ‘externally’ to the philosophy of art – for instance, questions from other areas of philosophy – or for examples to use in arguments in these other areas of philosophy.

What I am proposing does not strictly require every philosopher of art to engage with artistic practice in one of these ways, since it only stipulates that the conclusion of a philosophical argument about art should work as an answer to a question that could arise from some experience had in the course of artistic practice, but not that this question must arise from the philosopher’s own practical experience with art. This helps philosophers avoid inappropriately importing questions and problems from other areas of philosophy and imposing them onto art, which can distort our views of art-related phenomena by insisting they conform to a framework or set of concepts originally meant to apply to some other type of phenomenon. Significantly, while the principle does not require all philosophers of art to have first- or even second-hand knowledge of the practice, this kind of knowledge of artistic practices will be required to apply the pragmatist principle or to determine whether it is being applied properly by another, since this will be needed to judge what questions could plausibly arise for a reflective practitioner, or whether an answer to such a question could inform and usefully direct future practice – so the principle does require a robust engagement with artistic practice from at least some philosophers of art.

V. Advantages of the Pragmatist Principle

The principle proposed above avoids the problems faced by Davies’s formulation of his pragmatic constraint. While, like Davies’s version, it does not specify what the substance of philosophical views about art will be – which is desirable for a methodological principle – it gives a criterion for determining whether or not a particular philosophical discussion or argument adheres to the principle, whereas Davies’s version is less practically useful since it offers no way of determining which views in the epistemology of art are ‘best’ or would be upheld after rational reflection. It also helps to avoid disputes of the kind between Davies and Dodd, since it not only governs the ontology of art but also applies to epistemology, including views on which properties are rightly ascribed to artworks or how artworks are properly appreciated.³¹ Although it does not directly tell us which views on epistemological matters are correct, it does give a principled way of judging some views as less legitimate: namely, those that could not serve as answers to a question, or solutions to a problem, that could plausibly arise for a reflective practitioner in the course of making or engaging with art. Moreover, the pragmatist principle avoids these problems while achieving the aim of Davies’s pragmatic constraint insofar as it ensures that ontological theories will align with epistemological ones in the way Davies wants, since theories of both types will be compatible if both are connected to artistic practice in the way the pragmatist principle requires.

There are additional reasons for adopting the pragmatist principle as a general methodological principle in the philosophy of art. First, it achieves another aim of Davies’s constraint on which there is widespread agreement – namely, to make artistic practice the ‘touchstone’ for philosophizing about art – where Davies’s constraint effectively makes epistemology of art the touchstone for ontology with no guarantee that the explananda of epistemological discussions are in fact the same things that feature in artistic practice. Second, by encouraging philosophers of art to take artistic practice as the starting point for their inquiries, rather than starting from the existing philosophical literature, it promotes the discovery of new philosophical questions while avoiding recycling familiar or fashionable problems from other areas of philosophy and applying them derivatively to art. This could help to counter prejudices against aesthetics held by some specialists in other areas of philosophy who see it as a less-serious subfield in the discipline, since this would show that philosophy of art, having its own distinct philosophical problems, is not merely derivative of other ‘core’ areas of philosophy. In my own observation, attempts by aestheticians to get philosophers working in other areas to take aesthetics seriously tend to involve showing how familiar problems from those areas can be applied to art, or how examples from art can provide interesting or puzzling cases for these other areas. However, this strategy is rarely effective in changing assumptions of aesthetics’ ‘second-class’ philosophical status but is more likely to enforce the idea that it is derivative of, and so secondary to, what are seen to be the ‘core’ areas of philosophy.

Third, philosophizing about art in accordance with this principle could make philosophy of art more relevant to artists and other practitioners, including ‘committed engagers’ with (rather than ‘casual consumers’ of) art. This could open the door for a greater range and frequency of interdisciplinary collaboration between philosophers and practitioners. For philosophers who are not themselves artists, such collaborations could be a further way to discover new philosophical questions and problems that are distinctive of philosophy of art. This could also benefit work in other areas of philosophy, with philosophers of art finding in artistic practice counterexamples to theories in, say, metaphysics that could inform these theories ‘from the bottom up’ rather than the theories’ validity being assumed and imposed on art ‘from above’. For example, a plausible case for artworks being both created and best understood on a type-token model could challenge the assumption of metaphysicians that abstract objects are eternal and so cannot be created. And, when it comes to artists, they stand to gain a clearer understanding of what it is they do, with what is implicit in their practices being made explicit by philosophers acting as co-investigators rather than ‘top-down’ imposers of theory.³²

VI. Conclusion

I have discussed some problems for the applicability of the methodological principle that Davies’s calls the ‘pragmatic constraint’ and proposed an expanded version of this constraint that, to differentiate it, I call the ‘pragmatist principle’, which solves these problems while retaining and better fulfilling the motivation for Davies’s constraint, which is to keep philosophical theorizing about art grounded by making artistic practice its ‘touchstone’ without collapsing into strong descriptivism. I have also noted some benefits of adopting the pragmatist principle that go beyond solving the aforementioned problems with Davies’s constraint. Nevertheless, I suspect that some philosophers of art will be reluctant to adopt the pragmatist principle, at least in a widespread manner, and so in conclusion I should like to address some potential worries.

Why, it might be asked, should we seek to limit rather than expand the scope of philosophical thinking about art? Why should we not ‘let a thousand flowers bloom’? This worry might also be expressed by philosophers protesting the idea of anyone telling them how they ‘should’ be doing philosophy. However, this mistakes the purpose of the principle and the spirit in which it is proposed, which is not primarily to ‘tell philosophers what to do’ but to ensure that philosophical claims purporting to be about art are in fact about art as it occurs outside philosophical theorizing. Presumably every philosopher of art would agree that they want to be thinking and drawing conclusions about something that exists in the world prior to their theorizing, and not merely to be performing logical operations on a variable called ‘art’ that could just as well be called ‘X’, where the pragmatist principle gives a way of determining when this is (and is not) being done.

As for the worry that the principle would unhelpfully limit philosophical thinking by stipulating where philosophers should look for ideas and start their theorizing from, this also mistakes what the principle is calling for, which is not for philosophers of art to go about doing their thinking in any particular way but for the conclusions they draw and the views they develop to meet the conditions laid out in the principle. In other words, the principle gives a way of telling when a train of philosophical thought is on the wrong track, but does not specify a particular track that philosophers ‘should’ be thinking along. Thus, the principle does not prevent philosophers of art from drawing on other areas of philosophy and applying them to artistic phenomena to see what results, so long as the results work to advance our understanding of art – where, if they do, it will entail that the resulting view works as a solution to a problem that could arise from practice, even if the problem happened to arise in that case from a philosopher’s thinking about ideas in, say, the philosophy of language.

Another worry is that the principle assumes that philosophical theorizing should be practically applicable; however, some might say that philosophy is a theoretical discipline and does not need to – and perhaps should not – concern itself with practical ‘impact’ for its legitimacy. Such a view might see philosophy as tasked with clarifying concepts without making a difference for how those concepts are used in practice, or with solving logical puzzles that need not refer to anything outside themselves. I do not know if this is a genuine worry for the proposed principle so much as it is a statement of a fundamental metaphilosophical disagreement, where the metaphilosophical stance presupposed in the worry – that philosophy is essentially and only a pursuit of ‘pure theory’– is not one that would see the desirability of practice being a touchstone for philosophical theorizing in the first place, so anyone who truly held this view of philosophy would be equally unsympathetic to the motivation for Davies’s pragmatic constraint and my pragmatist principle.

It is beyond the scope of this article to argue for one metaphilosophical view over another; however, I would note that this objection would leave philosophy of art to be concerned with what was characterized above as the mere performance of logical operations on a variable called ‘art’ that might just as well be called ‘X’. This would undermine the very idea of philosophy of art as a distinct subfield of the discipline, with this outlook effectively reducing all ‘philosophies of’ to applied logic. I take it that the majority of philosophers of art, if not all, would reject this view. If that is right, it would mean that most philosophers of art will not seriously raise this worry but instead will have an interest in the pragmatist principle as a way to guard against their field of philosophy becoming an insular theory that refers only to itself, or merely a form of applied logic.

Competing Interests

The author has no competing interests to declare.