1 Introduction

This paper revisits my paper, ‘Grace de Laguna’s 1909 Critique of Analytic Philosophy’ (henceforth, Critique) and responds to the much-appreciated commentary on it. In response to James Chase and Jack Reynolds, I clarify the disagreement de Laguna had with analytic philosophy about the extent of human knowledge, including her rejection of the assumption that we have knowledge conceived of as factive. I also clarify how this disagreement underpins her critique of analytic philosophy’s epistemic conservativism. Finally, I further examine the extent to which analytic philosophy has been conservative, suggest that de Laguna’s critique of analytic philosophy applies to recent forms of naturalism, including methodological and liberal naturalism, and explain how philosophical criticism of established opinion can be justified.

In response to Andreas Vrahimis, I contextualise my discussion of de Laguna’s work in 1909. I suggest that, contrary to Vrahimis, de Laguna had a fully developed critique of epistemically conservative philosophy before that year. Indeed, we will see that she will have already identified early analytic philosophy as the primary target of this critique. Also contrary to Vrahimis, I explain why de Laguna’s critique is effective against Bertrand Russell’s 1914 work.

Finally, I examine Cheryl Misak’s commentary. Contrary to Misak’s proposal, I argue that de Laguna’s philosophical roots are found in the work of Herbert Spencer and her absolute idealist teacher, James Edwin Creighton. I also explain why Misak’s suggestion that de Laguna was wrong to reject the label ‘pragmatist’ is based on a misunderstanding of the differences between pragmatism and idealism, one that reflects a problematic approach to historiography. Not unrelated, I argue that Misak fails to recognise the force of de Laguna’s critique of pragmatism.

2 The disagreement between speculative and analytic philosophy

Critique reconstructs and supports an argument by Grace de Laguna against epistemically conservative philosophy construed as the philosophy that, ‘in answering philosophical questions, tends and aims to avoid going beyond or critiquing (at least some part of) established opinion’ (Katzav, 2023a, p. 4). The argument is based on two named premises, Partiality and Purpose Relativity. These premises are as follows (Katzav, 2023a, p. 7):

(Partiality) Everyday and scientific knowledge are partially true in both recognised and unrecognised ways.

(Purpose Relativity) Which partial truths we accept depends on our purposes.

In clarifying Partiality, I noted that it should be understood to be telling us that ‘in everyday and scientific knowledge, partial truth is pervasive in recognised and unrecognised ways’ (Katzav, 2023a, p. 10). The argument against epistemic conservatism is roughly that, if partial truth pervades established opinion in recognised and unrecognised ways, philosophy ought not to assume the unqualified truth of the body of judgements in any part of established opinion but ought rather to engage in an examination of the extent to which they are partially true. Such examination is needed in order to determine whether the judgements are true enough for philosophy’s distinctive purposes. In illustrating my argument, I focused on three exemplars of the analytic approach–namely the approaches of Russell, Willard V. Quine, and David Lewis (henceforth, the trio).

Chase and Reynolds (2023, pp. 4–6) very usefully suggest that epistemic conservatism includes three threads. First, there is Starting Place, the assumption that philosophy should start to answer its questions by assuming the unqualified truth of a body of established opinion. Second, there is Path Dependency, according to which answers to philosophical questions should be arrived at by somehow unpacking the assumed to be true claims of established opinion. Third, there is what Chase and Reynolds call Modesty and I, for reasons that will become clear, prefer to call Inferiority, which is the assumption that philosophical justification is, at least traditionally, weaker than the justification available to the assumed to be true part of established opinion.

According to Chase and Reynolds, the disagreement between de Laguna and the trio about philosophical approach really concerns Path Dependency and Inferiority, with the latter being the most significant element of the disagreement. Moreover, Chase and Reynolds add, the real issue driving the critique of the trio’s approaches is not Partiality or Purpose Relativity but the assumption that philosophy has its own aims, especially the aim of criticism from its own independent perspective. Chase and Reynolds say Quine was aware that claims to knowledge are fallible and holistically evaluated, and Lewis can acknowledge that standards of accuracy vary with context. So, in Chase and Reynolds’ view, it is not clear to what extent Quine and Lewis disagree with Partiality or Purpose Relativity. At the same time, it seems that one can insist on Starting Place and Path Dependency while accepting de Laguna’s view that philosophy ought to include a critique of all established opinion. For one could argue for scepticism via an immanent critique of established opinion. Thus, Chase and Reynolds suggest that Inferiority and the desire to avoid critiquing established opinion must drive the trio’s conservativism (Chase and Reynolds, 2023, pp. 6–7).

Details of formulation aside, I agree that epistemic conservatism includes Starting Place, Path dependency, and Inferiority. However, I do not agree that the disagreement between de Laguna and epistemically conservative philosophy only concerns Path Dependency and Inferiority, nor do I agree that the desire to avoid critiquing established opinion drives epistemic conservatism. On my view, the disagreement about all three components of epistemic conservatism is underpinned by disagreement about the nature of knowledge and reality. Particularly relevant here, epistemically conservative philosophy is committed to Factivity, the assumption that knowledge is, as such, of unqualified truths, and Privilege, the assumption that established opinion includes, and is known to include, a body of knowledge conceived of as factive. Privilege underpins the assumption that philosophy can, in accord with Starting Point and Path Dependency, start with a body of established opinion that is assumed to be true without qualification. But Privilege also underpins Inferiority. Optimism about the epistemic status of established opinion is needed if we are to justify the view that it is epistemically superior to traditional philosophy.

Further, Partiality is what undermines Privilege and thus undermines the support for the various aspects of epistemic conservatism. Purpose Relativity can then be used to support the idea that, in order to attain philosophy’s aims, it needs a speculative side. Here, the assumption that philosophy is to use its own independent standpoint is argued for rather than assumed. It is worth adding that Partiality also puts pressure on Factivity. The threat that Partiality will lead to scepticism about knowledge that is factive will push us to consider rejecting Factivity.

Perhaps Chase and Reynolds miss all this because they do not see how strong Partiality is. They do not, to begin with, note that it is incompatible with Starting Point irrespective of where Starting Point leads. More importantly, they assume, as is made clear by their comments about Quine, that Partiality can easily be accommodated by epistemic holism and fallibilism. However, de Laguna’s argument as I have reconstructed it is not focused on these two theses. It focuses on how much truth we have. Her argument is that the extent to which partial truth is found in knowledge means that there is no instance of knowledge that can be assumed to be unqualifiedly true. There is reason to suppose that commonsense, the best systems of formal logic, mathematics, and the special sciences are pervaded by partial truth in recognised and unrecognised ways. As a result, if we are to make claims to unqualified truth, extensive investigation and argument are required.

Let me illustrate how the views of the trio support my construal of the underpinnings of epistemic conservatism. As Critique points out, Russell’s commitment to Privilege is exemplified in his assumptions about our knowledge of sense data and logic. He assumes that judgements about sense data and what he calls ‘the laws of logic’ are true without qualification, and he assumes that we have available to us a body of highly certain common knowledge that can be logically analysed in terms of judgements about sense data (Katzav, 2023a, pp. 4–5; Russell, 1914, pp. 75–80). Russell, apparently oblivious to the idealised nature of the structure of science, adds that philosophical analysis must preserve the structural claims of science (Russell, 1924, p. 378). This version of Privilege allows him to adopt a version of Inferiority according to which ‘science has a much greater likelihood of being true in the main than any philosophy hitherto advanced’ and he uses Inferiority to justify the rest of his epistemically conservative approach to philosophy (Russell, 1924, pp. 377–378).

At the end of the 1940s, Quine moves away from the kind of sense-data foundationalism put forward by Russell as well as from the view that science can provide us with the high degree of certainty Russell thought it could. Oblivious to Partiality, however, Quine still believed that to test scientific statements is to test whether they are unqualifiedly true (Quine, 1951, p. 40; 1969). He could accordingly still adopt a version of Privilege, along with an implicit commitment to Factivity, by equating scientific knowledge with the body of statements, theoretical and otherwise, that have survived testing and are believed, albeit fallibly, to be true (Quine, 1951, pp. 39–40). Quine does not even see a difference between this view of science and his epistemically conservative philosophy, along with its implicit commitment to Starting Point, Path Dependency, and Inferiority. He writes that, on his view of science, ontological questions ‘are on a par with questions of natural science’ (Quine, 1951, pp. 42–43). Metaphysics, accordingly, must answer its questions by analysing our best science and doing so on the assumption that it is true (Quine, 1951, p. 43). More broadly, Quine believed that.

[t]he naturalistic philosopher begins his reasoning with the inherited world theory as a going concern. He tentatively believes all of it, but believes also that some unidentified portions are wrong. He tries to improve, clarify and understand the system from within (Quine, 1981, p. 72).

Not recognising the centrality of Privilege to epistemic conservatism goes along with the supposition that speculative, but not epistemically conservative, philosophy is epistemically immodest. The converse is the case. While Critique accepted, for the sake of argument, that we might have some unqualified truths, it also made clear that de Laguna takes the identification of such truths, including by philosophy, to be an unattainable ideal. Further, she nowhere supposes that philosophy provides us with something like scientific knowledge. Indeed, she is explicit that philosophy does not do so (see, e.g. De Laguna, 1927, p. 207). The trio, as we have seen, assume that unqualified truth is easy to evaluate and attain. But they also assume that philosophical knowledge is, like scientific knowledge, of unqualified truth and that epistemically conservative philosophy will provide us with knowledge.

Russel, to begin with, holds that speculative philosophy is not feasible (Russell, 1917, p. 113). However, he is confident enough to dub his approach to philosophy scientific and to tell us that ‘[t]he failure of philosophy hitherto has been due in the main to haste and ambition: patience and modesty, here as in other sciences, will open the road to solid and durable progress’ (Russell, 1917, p. 124). Indeed, Russell’s epistemic immodesty is such that he thinks that his new philosophy can approximate the truth in a way that assures us that its insights, just like those of science, will survive when its errors are corrected (Russell, 1917, p. 113).

Quine, of course, supposed not only that philosophers should start by believing key scientific theories but also that philosophy makes use of the methods of natural science. When properly done, philosophy is scientific in the same way as natural science. Thus, philosophy is concerned with what we should believe, just like natural science (Quine, 1981, p. 72). Quine is not, as far as I can tell, explicit about how confident this should make us about his scientific philosophy. But the implication of his equation of properly done philosophy with natural science is that our uncertainty about the results of scientific philosophy is like our uncertainty about the results of natural science. Our philosophy, on his view, is just part of what is supposedly our shared world theory. This conclusion fits well with Quine’s actual claims about the results of scientific philosophy. For instance, he says that the theory of sense data is not merely an oversimplification but ‘a basic falsification’ and thus far less certain than an alternative based on natural science (Quine, 1953, p. 201).

Recognising the limitations of human knowledge also bears on Chase and Reynolds’ suggestion that epistemically conservative philosophy leaves philosophers with much work to do. As indicated in Critique, I think this is correct. However, I also point out that the limits of epistemically conservative projects need to be kept in mind and that ultimately it is the speculative philosopher who must judge the value of conservative work (Katzav, 2023a, p. 16). Chase and Reynolds suggest, for example, that epistemically conservative philosophers can still pursue explanations of the experience of the passage of time in light of four-dimensionalist theories of physics (Chase & Reynolds, 2023, p. 7). However, if de Laguna’s argument holds, we cannot appropriately account for time merely through an uncritical examination of some part of physics, say of general relativity. We must first critique physics.

To be sure, how epistemically conservative a philosophy is is a matter of degree, and this might lead one to wonder, with Chase and Reynolds, whether analytic philosophy itself includes various speculative projects (Chase & Reynolds, 2023, p. 8). My view is that analytic philosophy was characterised as epistemically conservative at the institutional level during much of the last century (Katzav, 2023b). Such a view is compatible with supposing that there was some speculative philosophy within analytic philosophy. Indeed, my view is that, even during its most conservative period, it included more and less epistemically conservative approaches.

That said, when I consider key individual analytic thinkers from the previous century, I find that even those with speculative reputations were hardly speculative philosophers. Quine was, as far as I can tell, one of the more speculative philosophers of mid-twentieth-century analytic philosophy. He did, after all, argue that philosophy was, to a limited extent, speculative (Quine, 1951, p. 20). Nevertheless, as I have suggested, Quine is an epistemically conservative philosopher. Other candidate speculative thinkers from this period are similar. Wilfred Sellars, one of the candidate speculative, analytic thinkers mentioned by Chase and Reynolds, is clear that his vision of philosophy as making sense of how all things hang together in the most general way is strongly constrained by the assumption that philosophy does not contribute substantively to what we know (Katzav, 2018, p.15). Chase and Reynolds also suggest that Iris Murdoch may have been a speculative, analytic philosopher. On my reading, she is a follower of Wittgenstein, arguably the most conservative of the major analytic thinkers. Murdoch thinks of herself as offering a metaphysical framework for morals, but she also insists that, in doing so, she is uncovering our deep common sense by examining how we use language (see, e.g. Forsberg, 2018).

I agree, however, that analytic philosophy has grown more speculative over time. A key question here is whether it has done so to a degree that adequately responds to critiques like de Laguna’s. I do not think so. Epistemically conservative attitudes do, in my view, still dominate despite a growing number of more speculative approaches. Moreover, related features of analytic philosophy remain, such as its relative lack of systematicity when compared to speculative philosophy. Critique began to illustrate this answer, and I will illustrate it further now in the case of the naturalistic approaches Chase and Reynolds propose as examples of recent, more speculative approaches.

One such example is methodological naturalism, the view that philosophy ought to be meaningfully constrained by the results of the sciences. Brian Leiter (2013) notes that such a view gives the philosopher a more speculative role. The naturalist philosopher, on this proposal, can adopt key ideas of an existing research programme in science and apply them in new domains to guide new research. Liberal naturalism, which is adopted by Reynolds (2018 and 2022), is another form of naturalism that can be more speculative than Quinean naturalism. Liberal naturalism aims to reconcile the diverse compartments of knowledge in a single understanding of reality. In doing so, it recognises and draws on non-scientific knowledge, especially regarding the everyday world.

De Laguna would not object to some of the tasks of methodological and liberal naturalism. She agrees that philosophy should propose speculative research projects for the special sciences. She also accepts the existence of non-scientific knowledge (Katzav, submitted for publication). However, de Laguna would insist, and I would agree, that these forms of naturalism inappropriately constrain philosophy. Restricting the speculative side of philosophy to using methods already found in science, to claims in areas where science has yet to develop, or to reconciling available bodies of knowledge is too epistemically conservative for de Laguna.Footnote 1 Her critique of epistemically conservative philosophy suggests that it is only through an examination of the extent to which science offers us abstract and perspectival judgements that we can determine to what extent philosophy ought to be constrained by science’s judgements and, by implication, methods. As a result, we cannot require of philosophy that it generally be continuous with science or that philosophy merely reconcile rather than propose more substantial revisions of accepted bodies of judgement. If, for example, a critique of established opinion shows that conceptualisation is merely a perspective on reality and that some philosophical questions require transcending what concepts can tell us, then philosophy that is discontinuous with science both in method and ideas will be appropriate.

The recognition that it is only through a critical, epistemological engagement with science that the extent to which science and philosophy ought to constrain each other is determined also helps to answer Chase and Reynolds’ query about how speculative philosophy justifies itself. Their worry is that de Laguna rejects not only rationalism but also the authority of empirical science and thus leaves philosophy with no source upon which to base its judgements (Chase & Reynolds, 2023, pp. 9–10). In response, philosophy can partly base its claims on its epistemic critique of established opinion. Philosophy, on this view, has as a focus the empirical investigation of the nature and limitations of judgement, especially of scientific judgement (Katzav, 2023a, p. 11–17). Philosophy thus has the philosophy of science, appropriately construed, as an essential part. As such, it becomes plausible that philosophy can legitimately claim to provide us with our best understanding of knowledge. To be sure, philosophical knowledge of this kind is limited–it is tentative, partial, and does not typically deserve the title ‘scientific’–but that is just the way things are when it comes to our knowledge of our knowledge.

What, however, about the evaluation of philosophy’s own epistemic standards? De Laguna brings metaphysics into the picture here. Her view, as Critique points out, is that an epistemology is to be tested in part by determining how well it meshes with an appropriate metaphysics (Katzav, 2023a, pp. 2–3). This is a substantive constraint because a central task of metaphysics is not only to explain the possibility of knowledge but also its limitations, including those of philosophical knowledge. Our vision of reality needs to include in it a place for the human knower. Metaphysics appropriately conceived thus includes the empirically informed completion of an optimal, though no doubt not unique, estimate of the human epistemic situation.

In concluding my response to Chase and Reynolds, let me consider their claim that Lewis, despite his explicit commitment to epistemically conservative philosophy, is a speculative philosopher (Chase & Reynolds, 2023, p. 7). I take it that the proposal here is that key parts of Lewis’s philosophy do go beyond established opinion in ways that are not required by it despite his explicit commitment to epistemically conservative philosophy. Lewis was thus disingenuous (he recognised that some of his appeals to established opinion were spurious but did not admit this), misguided (he was overconfident about the support established opinion offered his philosophy), or some combination of these. Either way, it is unclear why making poorly founded philosophical claims should be a criterion for being a speculative philosopher. Criteria that are plausible, e.g. requiring with de Laguna that the speculative philosopher systematically critique the special sciences and develop a vision of reality in light of this (Katzav, 2023a, pp. 2–3), would have us classify Lewis as non-speculative.

3 A properly contextualised reading of de Laguna’s critique of analytic philosophy

Vrahimis’ discussion of Critique distinguishes between appropriationist and contextualist approaches to the history of philosophy. The appropriationist aims to reconstruct the philosophy of historical figures in order to address current philosophical problems, irrespective of the actual views of those figures. Contextualist readings aim to understand past philosophers on their own terms, irrespective of whether the resulting understanding remains philosophically interesting or relevant. According to Vrahimis, my extraction of an argument against analytic philosophy from de Laguna’s work in about 1909, especially from her paper ‘The Practical Character of Reality’ (PR), ‘appears manifestly appropriationist’ (Vrahimis, 2023, p. 2). This is supposedly seen in my suggestion that, although analytic philosophy was not de Laguna’s target in 1909, it was effectively so because of her general opposition to epistemically conservative philosophy (Vrahimis, 2023, p. 4). Vrahimis aims instead to offer a contextualist reading of de Laguna’s argument. On his view, such a reading provides a more charitable interpretation of her work (Vrahimis, 2023, p. 2).

On Vrahimis’ reading of PR, it could not have had analytic philosophy as a target because analytic philosophy did not exist in 1909. So too, Vrahimis points out that a critique of epistemic conservatism is much broader than one of analytic philosophy as such. PR’s real focus is a disagreement between absolute idealism and pragmatism, especially that of John Dewey. More specifically, PR is concerned with whether, in accord with absolute idealism, there are no fully determinate, finite experiences or whether, in accord with pragmatism, there are such experiences (Vrahimis, 2023, p. 7).

Vrahimis suggests that de Laguna is, like pragmatists, committed to instrumentalism, though instrumentalism leads her to an intermediate position between the absolute idealist view of experience and the pragmatist one. De Laguna characterises instrumentalism as the view that (a) thought is ‘a mode of organic adjustment to environment’ and ‘its whole development has been, and is, determined with reference to this function’, and (b) ‘all distinctions and terms of thought, that is to say, all meanings, are relative to the specific conditions which have called them forth and to the functions which they perform’ (De Laguna, 1909, p. 396). Vrahimis worries that it is not clear why the critique of Dewey’s position should be taken to be a critique of a broader one, never mind clear why it is a critique of his epistemic conservatism. That said, Vrahimis does acknowledge that de Laguna later voices her opposition to analytic philosophy and that she does develop a general critique of epistemic conservatism (Vrahimis, 2023, pp. 4–5).

I do not see my treatment of de Laguna’s work as either appropriationist or contextualist. My main aim in presenting de Laguna’s argument is not the appropriationist one of contributing to the examination of issues that are of concern to analytic philosophy. My main aim, rather, is to use her work to help transform our understanding of what issues are key to philosophy and of how these are to be addressed. Moreover, unlike the contextualist, doing philosophy is essential to my historical project. My view is that close attention to the historically situated meaning of philosophical texts is helpful in transforming philosophy. In these respects, my project is similar to others put forward by those sympathetic to the speculative tradition (e.g. Gare, 2017).

I also disagree with Vrahimis’ suggestion that de Laguna’s critique of epistemically conservative philosophy only becomes fully general over time. As Critique notes, de Laguna starts her metaphilosophical discussion in PR with a discussion of classical mechanics (De Laguna, 1909, p. 411). She endorses an argument that was at the time widely discussed, especially as a result of James Ward’s defence of it in Naturalism and Agnosticism (Ward, 1899). As de Laguna puts the argument, the abstract nature of mechanics and the special purposes of metaphysics indicate that a critique of mechanics is needed before it is used in drawing metaphysical conclusions. Similarly, as we will see, PR uses Partiality and Purpose Relativity to reject a pragmatist critique of absolute idealism. What motivates these arguments by de Laguna are just general considerations regarding the abstract nature of knowledge and the specific purposes of philosophy rather than specific considerations about mechanics or the pragmatist’s arguments. Thus, PR indicates that, for de Laguna, the philosophical implications of Partiality and Purpose Relativity are general ones.

De Laguna is more explicit about the generality of her argument in 1904, when critiquing Henry Heath Bawden’s pragmatist treatment of the mind–body problem. In that critique, she notes that, because all the sciences do no more than provide us with partial truths and their goals are not those of metaphysics, they cannot be a direct source of metaphysical conclusions. She writes that the laws and formulas of a science

are true so long as applied to the particular abstractions from concrete experience with which the science deals, but their application either to experience as a whole or to the subject matter of other sciences, is entirely illegitimate (Andrus, 1904a, p. 440).

Instead of extracting visions of reality from science, philosophers are to develop such visions by transforming what the sciences teach (Andrus, 1904a, p. 444; 1904b, p. 663). Indeed, de Laguna’s critique of Bawden is largely that he aims to address the mind–body problem without adequately taking the partial truth of scientific judgements and the special purposes of philosophy into account (Andrus, 1904a, pp. 430, 439–440). Here, the critique of epistemically conservative philosophy is restricted only in that it does not extend to conservatism about non-scientific knowledge, and this is merely because science alone is taken to be relevant to philosophy.

Turning to the connection between de Laguna’s critique of Dewey’s pragmatism and her critique of epistemically conservative philosophy, it will help to see how de Laguna thinks of Dewey’s philosophy. In 1951, she recognises that he did not think that philosophy is primarily speculative. Nevertheless, she insists that he was a speculative philosopher, noting that he aimed to provide a general insight into existence (De Laguna, 1951, p. 16). We can add that, unlike Lewis, Dewey developed his naturalistic view of knowledge and reality through a critical engagement with key sciences, especially biology and psychology (De Laguna & De Laguna, 1910, pp. 117–134). Thus his vision, unlike that of Lewis, is a speculative one in that it involves a critique of established opinion and develops out of this critique. Indeed, de Laguna argues that Dewey at least implicitly provides a critique of all conceptual knowledge. She writes that.

[k]nowing, according to Dewey, is a natural event and one that makes a difference. Real existence is no fixed system, if for no other reason than that in becoming known it undergoes a real change and suffers a new growth. Reality is more than it can be “known as” (De Laguna, 1951, p. 16).

That de Laguna thought of Dewey as a speculative philosopher suggests that her critique of his views in PR is not that his philosophy is epistemically conservative. Indeed, a closer look at PR confirms that her specific point about Dewey’s self-conception was that he was not a sufficiently consistent speculative philosopher.

PR is a critique of what de Laguna takes to be the most distinctive claims of pragmatism as well as of some of their key implications. The first distinctive claim is instrumentalism, and the relevant implication of instrumentalism is that reality has a practical character. Instrumentalism tells us that judgement is correct only for local purposes and so implies that reality has a practical character in the sense that what is real, and indeed the nature of reality, is a local affair (De Laguna, 1909, pp. 396–398). The second distinctive claim is immediatism, especially Dewey’s variant of it. Immediatism is the view that things are as they are experienced to be (De Laguna, 1909, p. 398). Immediatism implies that, when we learn something new about an object from a changed experience of it, e.g. when we come to judge that a horse previously experienced as dangerous is not dangerous after all, reality has changed. Thus, immediatism implies that knowledge comes with a change in reality (De Laguna, 1909, pp. 398–400).

Pragmatism’s claims are undermined by analyses of what it is like to experience something and of the use of the concept of reality in practice. These analyses suggest that.

[t]he experience of a thing as anything is always an interpretation, an assumption on which we act in our dealings with it; and the question as to the real nature of the thing refers to the verification of the assumption (De Laguna, 1909, p. 406).

The first part of this quote is an epistemic claim and is that claims about experience are interpretations that might turn out to be wrong. The second part is the crux of de Laguna’s criticism of immediatism and is an ontological claim about what it is to be real. According to her analyses, any claim about the real must be verifiable from multiple perspectives and so cannot be about what is immediately given. Another way she puts her ontological claim is similar to Vrahimis’ way of stating it: the real ‘is never immediately experienced at all; it is always ideal’ (De Laguna, 1909, p. 405).

Regarding the immediatist claim that knowledge comes with a change in reality, de Laguna notes that it can be understood as the truism that ‘knowing is a change in reality’, or as a deeply paradoxical position (De Laguna, 1909, p. 407). The paradoxical position she has in mind fits her 1951 characterisation of Dewey’s position well. Paradoxically interpreted, the view that knowledge comes with a change in reality is the view that the result of inquiry, i.e. the object known, is the real and yet, since knowing is a change, it cannot be the real that is changed by knowing (De Laguna, 1909, p. 408).

Against instrumentalism and the claim that reality is practical, de Laguna argues that the sense in which reality can be termed practical is a sense ‘almost directly opposed to that in which Professor Dewey has employed the phrase’ and one according to which reality is practical because of ‘its stability throughout the changes of our attitudes’ (De Laguna, 1909, p. 409). This thesis nicely meshes with her view that the real is the ideal.

Epistemic conservatism only appears in PR after de Laguna’s critique of pragmatism, when she responds to the pragmatist critique of absolute idealism and, in doing so, provides a (qualified) defence of this form of idealism. The pragmatist critique of absolute idealism is that its view of reality as what conforms to the ideal of unqualified, systematic knowledge or of an all-embracing experience is irrelevant to actual experience. No actual judgement is or can be judged by absolute idealist standards (De Laguna, 1909, pp. 410–411). De Laguna responds by drawing an analogy between the absolute idealist view of knowledge and reality and the mechanistic one. Given Partiality and Purpose Relativity, it would be a mistake to accept either of these views without proper critique. However, that is hardly a reason to reject them; it merely means that, like all knowledge, they are abstractions and accordingly need to be appropriately qualified. Regarding the ideal of absolute knowledge, we need, for example, to point out that the claim that the true judgement is one that will withstand all tests is subject to the caveat that, because the evolution of knowledge involves conceptual change, knowledge is strictly speaking not cumulative. Still, given the view that the real is the ideal and the rejection of immediatism, the door is open to accepting the absolute idealist conception of knowledge and reality when these are appropriately qualified (De Laguna, 1909, pp. 411–414).

De Laguna’s critique of immediatism, then, is not part of her critique of epistemic conservatism. The latter critique is articulated in her statement about how to react to attempts to read off mechanistic ontologies from science and in her response to the pragmatist critique of absolute idealism. Moreover, de Laguna’s response to pragmatists here is that they mistakenly assume that noting that absolute idealism is abstract suffices as a critique of it while, in accord with Purpose Relativity, such a critique also requires a consideration of the purposes to which absolute idealism is being put. The pragmatist is being epistemically conservative merely in taking absolute idealism to be a view about the unqualified truth.

In what sense, however, can de Laguna be thought of as offering a critique of analytic philosophy? I agree with Vrahimis that analytic philosophy did not exist in 1909 and thus that it is a mistake to think that de Laguna then had analytic philosophy in mind as a target of her critique. Indeed, as Vrahimis recognises, I only claim that de Laguna’s, 1909 argument is in effect a critique of analytic philosophy (Katzav, 2023a, p. 2). It is much later that de Laguna explicitly states her opposition to analytic philosophy (Katzav, 2023a, p. 3).

Nevertheless, by 1909, early analytic philosophy was developing alongside American new realism. This development is usually characterised as the development of a realist opposition to an entrenched idealism and the beginnings of a broader analytic tradition. The development, however, was also recognised by key figures at the time to be an epistemically conservative reaction to speculative philosophy. It is notable that the British philosopher Thomas Case, who already in 1888 describes his philosophy as an analytic one, makes epistemic conservatism about science the touchstone of good philosophy, including adequate epistemology and metaphysics (Case, 1888, pp. 3–13). Not much later, new realists such as Ralph Barton Perry (1904) and Walter T. Marvin (1912) respond to the speculative philosophy of science of Ward and others, and in doing so identify the new realists’ dispute with idealism as not being primarily about idealism but about the appropriate attitude philosophers should adopt to science. According to Perry and Marvin, it is only an epistemically conservative attitude that is appropriate. Marvin notes that they are not alone but part of a large, emerging epistemically conservative movement (Marvin, 1912, p. 309).

To be sure, there was at the time no sharp dividing line between conservative and speculative philosophers; there was an entire spectrum of approaches. Even among the new realists, we find William Pepperrell Montague, who rejected critiques of science as radical as de Laguna’s but went on to speculate boldly where scientific evidence was as yet unavailable; while not a member of de Laguna’s speculative school, his philosophy is a kind of speculative philosophy (Sheldon, 1954). Still, the emerging epistemic conservatism, along with its tendency to deny philosophy’s epistemic independence, was explicitly recognised as such by speculative philosophers within de Laguna’s camp, including, for example, by Creighton (1904), Ward (1904), and Morris Raphael Cohen (1913). De Laguna could not but have been aware, with those around her, of the coalescing wave of epistemically conservative philosophy that was going to become analytic philosophy. And she could hardly have been unaware that this wave was the perfect target of arguments such as hers.

Vrahimis further worries that my reading of PR is uncharitable. More specifically, he thinks that it is uncharitable to apply the argument I find in PR to the sense data theory Russell developed in 1914 in Our Knowledge of the External World. First, according to Vrahimis, Russell’s theory is atomistic. It assumes the existence of external relations between sense data. Because of this, de Laguna’s argument would beg the question against Russell (Vrahimis, 2023, p. 5). Her argument, as I present it, supports Partiality by arguing that conceptualisation of phenomena, including sense data, distorts their nature because it involves considering them apart from the context in which they are found. Second, Russell views philosophy as a value-free, theoretical discipline that has no practical aims. This would set him against Purpose Relativity. Finally, according to Vrahimis, de Laguna’s critique of immediatism and support for Purpose Relativity rests on instrumentalism and evolutionist premises. Russell’s, 1914 work, however, criticised evolutionist philosophy (Vrahimis, 2023, p. 5). Russell’s critique is of Henri Bergson and is that Bergson’s evolutionism is not rendered likely by evolutionary biology and cannot be argued for on practical grounds (Russell, 1914, p. 25).

As Critique points out, however, Partiality was already well supported in 1909 (Katzav, 2023a, p. 8). This support included substantial cases against atomistic accounts of experience such as Russell’s, including lines of argument that eventually made his 1914 view that knowledge is founded on immediate certainty about sense data the quaint view it is (e.g. Bergson (1889), James (1890), De Laguna and De Laguna (1910)). PR’s own examination of the use of the concept of the real in judgement suggests, recall, that when we consider whether something really is as it appears to be, our question is not about the appearance itself but points beyond it to other experiences. Judging that apparently converging lines really are convergent, for example, must refer to potential further experiences of where the lines meet (De Laguna, 1909, pp. 404–406). Thus, according to de Laguna, the idea that sense data, conceived of as immediately knowable entities, are real is incoherent. At the end of PR, she intimates that the best that might be hoped for regarding the idea of a given is that it will turn out to be a useful abstraction (De Laguna, 1909, p. 415).

As for de Laguna’s case for Purpose Relativity, it is made empirically by considering judgement in a variety of contexts, especially in science. For example, the claim that the purpose relativity of judgement extends to logic is based on a consideration of the history of mathematics and logic (Katzav, 2023a, pp. 10–13). She appeals neither to the biological theory of evolution nor to practical considerations, so that Russell’s arguments against Bergson are irrelevant to her. Similarly, as we have seen, her case against immediatism rests on phenomenological and conceptual considerations rather than on instrumentalism. We have also seen that she in fact rejects instrumentalism, a point on which I will elaborate in the next section.

Finally, Vrahimis is unsure whether de Laguna explicitly says that abstraction inevitably involves distortion of abstracted phenomena (Vrahimis, 2023, pp. 7–8). However, she does so as early as 1904. In her discussion of Bawden’s work, she notes that each science deals with ‘the reality of concrete experience’ but that ‘in no case does it remain unchanged experience’ (Andrus, 1904a, p. 440). De Laguna’s use of mechanics to illustrate the requirement that philosophers must come to terms with the idealisations of science indicates that here too de Laguna takes abstraction to involve distortion. In 1910, when arguing for Partiality, de Laguna and her husband note that considering characteristics involves regarding them ‘out of connection with others which are equally constitutive of the subject in other relations’ (De Laguna & De Laguna, 1910, p. 153).

4 Beyond pragmatism and analytic philosophy

According to Misak, de Laguna was exposed to pragmatism at the early age of 20, when William James first presented it to the world in 1898 at Berkeley. As a result, de Laguna adopted two key pragmatist positions from the beginning of her career. These are the view that the categories of experience are to be understood in terms of their functions and ‘[p]ragmatism’s primary insight that the philosopher must begin with human beings and their practices’ (Misak, 2023, pp. 2–3). At the same time, Misak acknowledges that de Laguna starts out by rejecting pragmatism’s immediatism and insisting that pragmatism does not sufficiently overcome dogmatism, including the view that some of our knowledge is true without qualification (Misak, 2023, p. 3). Misak also thinks that de Laguna disagreed with pragmatism’s methodology. De Laguna, unlike pragmatists, supposedly thought that philosophy should adopt a standpoint distinct from those of the special sciences and, in doing so, scrupulously avoid using their concepts (Misak, 2023, p. 3).

Turning to consider early analytic philosophy, Misak claims that de Laguna and I seriously misunderstand this school when we take it to be epistemically conservative. According to Misak, de Laguna, like me, fails to realise that early analytic philosophy ‘starts with logic and experience, both supposedly the engines of objectivity. Russell, Moore, and Wittgenstein were trying to reduce or logically analyse all meaningful propositions to simple propositions that correspond to objects in the world. The result looked nothing like science’ (Misak, 2023, p. 3).

Misak, however, mistakenly assumes that de Laguna was exposed to pragmatism before she had developed her own perspective on philosophy. It was de Laguna’s husband, Theodore, who was present at James’ early lectures (Pearce, 2023). She only became a student in 1900 and only met Theodore in 1903 (Montoya, 2000a, b). As for her own views about the nature of philosophy, she had already outlined them in 1899 (Andrus, 1899; Katzav, 2023c).

Further, taking de Laguna to be a pragmatist on the grounds that she preferred functionalist views of the categories and thought that philosophy must start with established opinion involves a serious misunderstanding of pragmatism’s distinctive contributions to philosophy. Functionalist accounts of the categories as well as the idea that philosophy must start with established opinion were widespread in America throughout much of the nineteenth century while pragmatism only emerges as a distinct position at the century’s end (see, e.g. Schneider, 1946, chs. VI, VII and VII).

De Laguna’s own relevant commitments come from Herbert Spencer and her idealist teachers. Spencer was her first philosophical influence (Andrus, 1903; Montoya, 2000a), and he, as de Laguna herself explains, thought that philosophy ought to offer functionalist accounts of key categories of experience and to start its investigations with science (Andrus, 1903). The same commitments are to be found in the work of de Laguna’s teacher, James Edwin Creighton, and his supervisor, Jacob Gould Schurman. In explaining how philosophy should be done, Schurman writes that ‘[t]he evolutionary or historical method…makes science a reproduction in thought of the successive phases of objective reality’ and that ‘the full nature of any reality reveals itself only in the totality of its development’ (Schurman, 1892, pp. 74–75). Elsewhere Schurman examines which methods are appropriate to ethics and writes that he is ‘sure facts and science must precede theories and philosophy’ and that ‘a philosophy without science is as empty as theory without fact, as unconvincing as reason without the voucher of sensuous experience’ (1887, p. 34). For this reason, Schurman believes developing an ethics ought to wait until the relevant science is sufficiently developed (Schurman, 1887, pp. 35–37). Similarly, Creighton emphasizes the need to understand thought and experience in terms of their functions (Creighton, 1898, pp. 260–261) and argues at length that scientific results ‘possess a real objective value which must be reckoned with in our philosophy’ (Creighton, 1901, p. 54). De Laguna herself recognises that her functionalist commitments are found in broader and earlier currents than those of pragmatism, notably in the Hegelian tradition (Andrus, 1904b, p. 660).

Misak’s failure to recognise early analytic philosophers as defenders of established opinion rests on a misunderstanding of what epistemically conservative philosophy can be. It need not be akin to empirical science or epistemically conservative about it. Being conservative about logic and judgements about immediate experience, as Misak observes Moore, Russell, and Wittgenstein were, is also a form of epistemic conservatism. Further, their conservatism does extend to other parts of established opinion, as illustrated above in my discussion of Russell.

Misak is no less mistaken about the ways in which de Laguna’s views diverge from pragmatism and, as a result, Misak’s defence of pragmatism fails. On her view, de Laguna’s criticism of immediatism is a critique of the idea that beliefs are, or at least can be, infallible representations of experience. Supposedly, de Laguna thought that pragmatists disagreed with her view that all experience is interpreted. Yet, Misak tells us, pragmatists did think that ‘all beliefs are fallible interpretations of experience’ (Missak, 2023, p. 5). De Laguna, however, was aware that pragmatists thought this. In criticising James’ immediatism in her ‘The Pragmatic Method, The Will-To-Believe, Humanism, and Immediatism’, de Laguna states that the pragmatist thinks that ‘[t]he diversified character of our purest sense-experience is … attributable in an indefinite degree to the work of past thought’ (De Laguna & De Laguna, 1910, p. 233). She, accordingly, was not criticising pragmatists on the grounds that they failed to endorse this epistemic position. She was, as we have seen, criticising an ontological thesis. She was rejecting the idea that, in experience, there is something that is immediately given which is then interpreted. For her, experience just is an interpretation. Her targets, Dewey and James, are both committed to the idea that experience involves an immediately given something that is not truth evaluable (Dewey, 1905; O’Shea, 2018).

Similarly, Misak mistakenly takes de Laguna to object to pragmatism on the grounds that it starts with the results of current science. Misak responds that, while pragmatists do assume that we must start with what we believe to be true and thus with current science, they recognise that our beliefs are fallible and are willing to revise them (Misak, 2023, p. 5). But de Laguna’s charge that pragmatists are too epistemically conservative only concerns, in addition to the critique of idealism discussed in the previous section, their treatment of logic. According to de Laguna, pragmatists still treat logic, e.g. the principle of non-contradiction, as self-evident. Such treatment is inconsistent with the pragmatist commitment to instrumentalism and is undermined by the historical evidence for the fallibility of formal systems (Katzav, 2022, pp. 205–206). De Laguna could add, of course, that the pragmatists’ view of the starting point of inquiry as a point at which our collected judgements might be true ignores Partiality. The starting point of inquiry is a point at which our judgements are recognised to be partially false.

De Laguna, we have seen, also rejects the pragmatist’s view of the practical character of reality and, as a result, rejects instrumentalism. However, de Laguna rejects instrumentalism for further reasons. Recall that instrumentalism includes the view that the meaning of a judgement is analysable in terms of the conditions that elicit it and the function it performs. The function here is the behaviour to which the judgement gives rise (e.g. Dewey, 1903, pp. 65–85). Such analyses are feasible if, with Dewey and James, we identify the real with what is immediately given. This identification allows us to suppose that the meaning of a judgement is fully analysed in terms of how it brings us into contact with the given and thus in terms of how it relates stimulus conditions to responses. But if, with de Laguna, we identify the real with the ideal conceived of as what cannot be immediately given, we also need to modify our view of judgement. The meaning of a judgement is then, as de Laguna recognises, plausibly thought of partly in terms of how it guides reasoning about the ideal and thus at least partly as a function of the inferences and values appealed to in deciding the adequacy of systems of judgements (De Laguna & De Laguna, 1910, pp. 197–198, pp. 205–206 and p. 210; Katzav, 2022).

Further, also contrary to instrumentalism, since a judgement’s meaning depends on the structure of the system to which it belongs, judgement is not thought of as a biological adaptation. Instead, it is the result of social development. In general, for de Laguna, human ‘mental development is a social phenomenon’ (De Laguna & De Laguna, 1910, p. 138). For the same reason, meanings are not conceived of as adjustments to specific conditions. Indeed, de Laguna argues that, despite Purpose Relativity, history indicates that the evolution of judgement is partly in the direction of increasing independence from specific circumstances and purposes. For some kinds of judgement, e.g. for laws of nature, the idea that judgement is relative to circumstances and purposes ‘becomes exceedingly false’ (De Laguna & De Laguna, 1910, p. 160). De Laguna’s view thus contrasts with the pragmatist view that all judgement is of a single kind—the practical one—and recognises diverse kinds of judgement.

This rejection of instrumentalism allows de Laguna to claim that she, unlike the pragmatists, properly recognises the ‘essentially social nature of thought’ and the associated ‘independence and autonomy of thought’ (De Laguna, 1927, p. 353). For the instrumentalist, thought is just a tool for dealing with experience. Accordingly, it does not come with its own standards of correctness; experience alone is the standard by which we can evaluate the correctness of judgements. De Laguna, however, argues that the increasing independence of judgement requires the socially underpinned independence and autonomy of thought. For thought, conceived of as regulated by social pressure and having its own standards of correctness, is what allows our judgements to be increasingly independent of experience (De Laguna, 1927, pp. 343–354).

Misak, perplexed by de Laguna’s claim that pragmatists fail to recognise the social nature of thought, protests that they do recognise it and offers George Herbert Mead’s work as an illustration of this (Misak, 2023, pp. 7–8). De Laguna was well aware of Mead’s work. She discusses it at length, explaining his views about the social nature of cognition and arguing that her key objections to pragmatism apply to these views (De Laguna, 1946). Mead, like other pragmatists, cannot really recognise the social nature of thought. This is because doing so would lead him to reject the pragmatist view of judgement.

Importantly, de Laguna’s views about judgement and thought transform how we can think of philosophy. The pragmatist thinks that philosophy needs to conform to the general, practical nature of judgement and thus is constrained by this nature. De Laguna, by contrast, thinks such constraints are flexible. Philosophy aims, in developing a vision of reality that accounts for the perspectives of all the sciences, to develop in thought a further independence from specific conditions than what is found in the sciences. In her terms, philosophy has as an aim the completion of thought (De Laguna, 1942, pp. 165–166).

I want to bring out one final way in which Misak misunderstands de Laguna. Misak suggests that de Laguna’s approach to philosophy is inconsistent. According to Misak, de Laguna’s 1904 discussion of Bawden’s work includes the commitment not to use the conceptions of the special sciences in answering philosophy’s questions. Yet, in 1927, de Laguna uses evolutionary conceptions to address philosophical questions (Misak, 2023, p. 8). As we have seen, however, de Laguna’s 1904 discussion is about developing a vision of reality. She is thus concerned with metaphysics as she understands it. In 1927, however, de Laguna aims at a more limited philosophical goal, namely the provision of methodological advice to psychologists. In this context, she is explicitly avoiding metaphysics (De Laguna, 1927, pp. 127–128).

With the disagreement between de Laguna and the pragmatists more explicitly outlined, we can see that her disagreement with them is deep, extending to the role and approach of philosophy, how to analyse meaning, the nature of knowledge, and metaphysics. We can also see how close she is to those immediately preceding her in the speculative tradition, that is, to her Hegelian teachers. Creighton too rejects the existence of immediate experience. His view is that ‘[a]n experience that is “pure” in the sense of reine Erfahrung, something free from all introjections of thought, is not only practically, but logically an impossible ideal’ (Creighton, 1901, p. 49). Further, Creighton, like de Laguna, thinks that the real is the ideal (Creighton, 1901, p. 49). Indeed, the similarity between Creighton and de Laguna’s visions of reality extends much further than this general agreement. Notably, both take the mental, the physical, and the social to be equally fundamental (Creighton, 1917; Katzav, 2023d). Creighton’s views about the relationship between science and philosophy are also close to those of de Laguna. To his view that scientific judgements provide us with knowledge of objective reality, he adds, much like de Laguna, that the knowledge provided is incomplete, so that, while philosophy needs to start with science, it cannot simply accept scientific methods or results as they stand but must critique these in light of the goals of philosophy (Creighton, 1901, pp. 49–56).

It is, to an extent, understandable that Misak should get de Laguna and the context in which she worked so wrong. De Laguna was hitherto entirely unfamiliar to analytic philosophers. Yet, Misak’s misreading of de Laguna does reflect a deeper historiographical issue. Misak only considers de Laguna’s work in light of an internal history of classical pragmatism. Relevant idealist and broader speculative contexts are entirely absent. Yet these were important within American philosophy during the early decades of the twentieth century (Schneider, 1946, p. 471), and it is easy to see that de Laguna’s thought is close to that of her Hegelian teachers.

This shortcoming, note, is not limited to the reading of de Laguna. Misak, like many writing the history of analytic philosophy, tends to pay limited attention to idealist and non-pragmatist, speculative context in writing about early twentieth-century philosophy more broadly. Doing so is liable to result in a distorted understanding of the significance of individuals, schools, and traditions. For example, Misak offers a largely internal history of pragmatism (Misak, 2013) but, as we saw in considering de Laguna’s relationship with pragmatism, we cannot understand what was distinctive about pragmatism without a proper understanding of the idealist context in which it developed.

The shortcoming at issue here is also metaphysical, epistemological, and ethical. It is not possible to understand who or what we are as (Anglophone) philosophers, including the extent to which we are rational and have made progress, unless we have an inclusive understanding of the history of twentieth-century philosophy. Additionally, the neglect of de Laguna’s work and of the work of others in her tradition involves an injustice. Merely correcting the record in de Laguna’s case by dubbing her ‘pragmatist’ aims to legitimise her but also contributes to the maintenance of familiar boundaries concerning who deserves inclusion in our histories.

Let me, in concluding, consider why de Laguna has been forgotten. Misak suggests that the reason Dewey never responded to de Laguna’s work is that he recognised her as a fellow pragmatist. This is an odd suggestion. First, one would expect that Dewey would still at least have wanted to clarify that his position is not affected by de Laguna’s arguments. Second, even if Dewey failed to realise that de Laguna presented a real challenge to pragmatist views of judgement, he could not possibly have failed to see that his positions were at odds with hers in many other ways. For example, he could not have failed to see that de Laguna rejected his immediatist ontology. Third, Dewey was not averse to commenting positively on the work of those with whom he was sympathetic, as he did, for example, in the case of Mead.

Why, then, did Dewey and other pragmatists not respond to de Laguna’s work? Part of my answer is that they responded indirectly. This involved claims to have been misunderstood, critiques of and responses to her philosophical school, and developing ideas in directions that aimed to address concerns that she and other like-minded philosophers raised. Documenting this answer is beyond the scope of the present paper.

The above answer is, in any case, incomplete. It remains puzzling that there is extremely limited, direct engagement with de Laguna’s work and that she was completely forgotten. Moreover, it is likely that there are multiple factors behind a complex social phenomenon such as the treatment of her work. I will mention three more of these. She was one of many American speculative philosophers who started their careers in the early decades of the twentieth century and have subsequently been forgotten. It is now well documented that this was due to a deliberate campaign by analytic philosophers (see, e.g. Katzav, 2023b). So too, de Laguna was a woman. While de Laguna’s speculative tradition aimed to find a place for women in philosophy, the tradition that became dominant in America, the analytic one, did not (Katzav et al. 2023). Finally, it is plausible that de Laguna’s status suffered because she was married. Here is Stephen C. Pepper, describing a visit to the college where de Laguna worked with her husband:

There was a husband and a wife there, the de Lagunas. She [Grace] was better than he was, though he had the higher rank, that was not supposed to be said out loud, but the whole philosophy world knew it. They were both good, but she was definitely better (Pepper, 1963, p. 83).