The Standards Problem in Conceptual Engineering

1. Introduction

Surely, concepts are our concepts, inherited, changeable and open to challenge – definitions and ideas that we human beings have forged to make sense of what William James called the ‘blooming, buzzing confusion’ of experience (1890: vol. I, 488). As Matthieu Queloz, in his superb The Practical Origins of Ideas, puts it:

Trying to understand the ideas we live by in isolation from the circumstances in which they are felicitously deployed is like studying a shoal of beached fish as if they were in their natural habitat. (Queloz 2021)

I say ‘surely’ with a note of irony, for the traditional account of concepts in analytic philosophy has been that they are somehow part of the furniture of the world, or exist in some abstract realm or are everlasting essences. In the late 1880s, Frege held that concepts exist objectively and we discover their content. Russell and Moore in the early 1900s established the method of conceptual analysis, where we are to take a concept and analyse it into its elementary components until we have unpacked its meaning and achieved clarity. The early Wittgenstein and the early Carnap argued in the 1920s (in the Tractatus and the Aufbau) that the philosopher can build a logically perfect structure of concepts. On all these approaches, our concepts are as they are – our job is to try to understand them, not to change them.

Since the late 1880s, pragmatists have been arguing that these approaches are wrong. The founder of pragmatism, C.S. Peirce put it thus: we ‘must look to the upshot of our concepts in order to rightly apprehend them’ (CP: 5.3, 1902); to get a complete grasp of a concept, we must connect it to that with which we have ‘dealings’ (CP: 5.416, 1906); ‘We must not begin by talking of pure ideas – vagabond thoughts that tramp the public roads without any human habitation – but must begin with men and their conversation’ (CP: 8.112, 1900). That self-consciously pragmatist approach has continued in the work of C.I. Lewis, Frank Ramsey, David Wiggins and Huw Price. I count myself as part of this pragmatist tradition which says that we can and should keep working on our concepts, including the scientific such as disease, the mathematical such as probability and the philosophical such as truth or justice.

The last decade has seen a heartening surge within the analytic philosophy of self-styled conceptual engineers, who say we should revise certain of our concepts in order to make them more useful or fit for their purpose (e.g. Brun 2016, Cappelen 2018, Chalmers 2020). They usually take their lineage to be the later Carnap and the later Wittgenstein. I have recently argued that they should turn rather to the pragmatists to see the earliest and best origins of conceptual engineering and to get some insights into how they should be moving forward (Misak 2022a).

Other philosophers, such as Nietzsche and Foucault and more recently Edward Craig, Ian Hacking, Sally Haslanger, Miranda Fricker, Martin Kusch, Amia Srinivasan and Bernard Williams have pursued an allied approach, which says we must look to the history or genealogy of our concepts to understand them. Queloz suggests that it wasn’t until the work of Sally Haslanger (her Resisting Reality: Social Construction and Social Critique (2012)) and Mirada Fricker that we would have an explicit discussion of ‘the possibility of using genealogy for ameliorative purposes’ (201).

We pragmatists have always wanted to bring the genealogists into our camp, with some success. For instance, Craig attended a pragmatist reading group that Huw Price and I organized in Cambridge in 2015 and was happy to be swept up in the pragmatist tradition. Queloz has now deftly merged the two approaches, arriving at a highly promising pragmatist genealogy. He proposes a normative, pragmatist genealogy that doesn’t compete with historiographical or Foucauldian forms of genealogy. It feeds off them and should be informed by them. In trying to discover the ‘naturalistic credentials or the point and value of certain concepts’ we will need to rely on social and other sciences (244).

I will be a terrible critic, for I have very little by way of negative things to say about this splendid book. My aim here will be to invite Queloz to bring more clearly into light his position with respect to a pressing problem for conceptual engineering. His style is the charitable one which finds insights in predecessors (including on occasion my own). This approach is commendable but has the potential downside to put Queloz’s contributions in the shadows. I will present a challenge that all pragmatists and genealogists must meet and then try to draw out Queloz’s solution to it. I will do that by setting out some pragmatist responses to the challenge and asking Queloz where he stands with respect to them.

2. The practical point of view

Queloz argues that we must examine concepts ‘from a practical point of view – to look at what ideas do rather than at whether the judgements they figure in are true – in order to see how exactly our ideas are bound up with our needs and concerns’ (3). That practical point of view allows us to evaluate our concepts, weakening or strengthening our confidence in the ideas we live by and revising them if necessary.

One of Queloz’s most fundamental insights is that if an idea persists, the reason may be that it fills a need, or that it earns its keep through serving some kind of concern or interest (2). He moves from there to a reverse engineering methodology, which looks into the origins of a community’s conceptual practices in an attempt to untangle the complex needs to which our concepts answer. This is the method he calls pragmatic genealogy. The philosopher is to tell ‘partly fictional, partly historical narratives’ which explore what might have driven us to develop certain ideas and discover what these ideas do for us. What point do they serve? What is the salient useful difference these concepts make to the lives of those who live by them (3)? We are to reconstruct the practical problems to which our concepts offer practical solutions. And if it turns out that those problems have changed, our concepts should also change. Queloz is not interested in a genealogy that tells us merely how we have actually used a concept.

With respect to the concept of truth, for instance, we must ask, as Williams put it, what exactly is it we value when we value the truth? Peirce also asked this question, as did Huw Price in his Facts and the Function of Truth (1988), both predecessors to William’s Truth and Truthfulness (2002). Once we discover the function of truth, we will be well-placed to criticize those concepts of truth that do not fulfil the function and work towards a concept that does.

Here is how this conceptual engineering project works in Peirce’s hands. Not only is it a good example of the methodology, but it will come into play when we try to meet our challenge. Peirce argued that ‘transcendental’ accounts of truth, such as the correspondence concept, on which a true belief is one that corresponds to, or gets right, or mirrors the believer-independent world, serve no function in inquiry and belief. The very idea of the believer-independent world, and the items within it to which beliefs might correspond, seems graspable only if we could somehow step outside our corpus of belief, our practices, or that with which we have dealings. The correspondence concept fails to make ‘readily comprehensible’ the fact that we aim at the truth or at getting things right (CP: 1.578, 1902). How could anyone aim for a truth that goes beyond what we can experience or beyond the best that inquiry could do? How could an inquirer adopt a methodology that might achieve that aim? The correspondence concept of truth is missing the dimension that makes it suitable for inquiry. It makes truth ‘a useless word’ and, ‘having no use for this meaning of the word “truth”, we had better use the word in another sense’ (CP: 5.553, 1905). A concept of truth that is useful in inquiry and fulfils the functions it actually plays in our lives (such as being that at which we aim when we investigate or deliberate), is indefeasibility, or beliefs that would stand up to all evidence and argument.

Peirce did not think that any concept that happens to work for us is thereby legitimated. He insisted that a concept must not be ‘determined by any circumstance extraneous to the facts’ (W: 3253, 1877). The brute impinging of experience, which we take to be an indicator of reality, is a check on our concepts. Other pragmatists, notably Lewis, Ramsey and Wiggins, have followed him in seeking an account of conceptual change that is answerable to human needs but also something more objective. Ramsey, for instance, said that definitions ‘show how we intend to use [concepts] in the future’; they ‘fix our future meaning’ ([1929] 1990: 1). But he also insisted that our concepts must also fit with the facts, otherwise they won’t really be useful. The belief in a certain concept of hell, for instance, had better attend to the object of the belief – whether there is really a hell. Otherwise, we will find ourselves with the ‘ludicrous’ pragmatism of Williams James, who seemed to think that a belief in hell need not take into account whether hell existed, but only whether the belief was satisfying ([1929] 1991: 91). Our concept of hell had better be connected to whether there is in fact a hell, for ‘if there is no such place it will be a mere waste of opportunities for enjoyment’ ([1929] 1991: 91–92).

Queloz is part of this best tradition of pragmatist conceptual engineering – one that tries to be true to both the objective and subjective nature of our concepts, one which sees that concepts are our concepts and that they develop through an engagement with the world we inhabit.

3. The standards problem

The most significant challenge for pragmatist and genealogist accounts of concepts is to say how we can assess proposals about concept revision. What is it for our concepts to fare or not fare well under genealogical reflection? How do we vindicate and criticize concepts and how do we do so in the face of disagreement? What are the standards by which we determine whether a concept plays its rightful place in our lives? How do we determine which concepts merit our confidence and when to abandon those that do not?

We can find examples of disagreement about concepts in all domains of inquiry and life. Science might be divided over whether the need for simplicity in a theory outweighs the need to respect the complexity of the phenomena. Medicine might be divided about the concept of evidence – whether only randomized control trials count or whether adaptive evidence is suitable, for instance, during a pandemic. Some say our ethical and political needs include substantive homogeneity in a population; others say this is not a need, but a dangerous mistake. The concepts in the vicinities of these questions will be those of truth, evidence, warrant, community, race, nation, justice, rule of law etc. Each of these concepts is shaped by our needs, but it is hard to say what those needs really are and how we should adjudicate between competing needs and the competing concepts that arise from divergent needs.

Some (but not all) conceptual engineers see and try to meet the standards problem. Some of those conceptual engineering answers to the standards problem aren’t squarely placeable in the genealogical or pragmatist traditions. Amie Thomasson (2022), for instance, argues that we must begin with the idea that parts of our linguistic or conceptual scheme, like other cultural artefacts, serve certain functions, and in assessing how well concepts fulfil their functions, we assess the rightness of our concepts. That is, she looks to work in linguistics to solve the standards problem.

A more pragmatist approach says that our concepts must meet the need to predict and control experience. As Peirce put it, experience ‘jabs you in the ribs’ and we have no choice but to attend to it (CP: 6.95, 1903). Our concepts are shaped by experience and are evaluable in terms of whether they are successful in experience and action.¹ Lewis tells us how that might work. He argued in his ‘A Pragmatic Conception of the A Priori’ (1923) and Mind and the World Order (1929) that our body of beliefs and concepts forms a pyramid, with the most comprehensive, such as those of logic, at the top, and the least general at the bottom. Or, to use his student Quine’s better metaphor, we operate within a web of beliefs with some at the core and some at the periphery.² What we think of as the a priori and analytic are in fact disconfirmable by experience – not directly, as is The swan coming around the corner will be white or All swans are white. But they can be shown to be mistaken if the tide of experience makes us question them. The a priori is the revisable ‘uncompelled initiative of human thought’ – the human-made net of categories and definitions without which we cannot ‘interrogate’ or ‘capture’ experience (Lewis 1923: 237–8, 1929: 307). We hold such ‘categorical principles’ – the principles of ‘definition, classification, and inference’ firm, unless experience speaks against them (Lewis 1929: 247). Any concept or belief, including those of logic, such as the principle of bivalence, can fall to experience, as long as we are willing to make the requisite revisions elsewhere in our interconnected corpus of belief. When we encounter a surprising experience, we attempt to fit it into our human, preformed patterns. Persistent failure leads to readjustment in the web, but the closer to the core a concept stands, the more reluctant we are to disturb it because the more radical and far-reaching the results will be.

Another pragmatist answer (not incompatible with Lewis’s) to the standards problem holds that we are justified in taking some things to be regulative assumptions or hopes. There are a handful of preconditions for our central needs and capacities, but these preconditions are not necessary, as Kant thought they were. They are simple principles and concepts we have to hold in place if we are to carry on in the way it seems that we must carry on. They are essential to the very activity of inquiry, belief or assertion. Peirce thought the following are such regulative assumptions or hopes: there are real things and something about them can be discovered through investigation; there are explanations for what we observe and answers to the questions into which we are inquiring; and inquiry will extend into the indefinite future. We can use such regulative assumptions as fallible principles in evaluating concepts.

A third, related, pragmatist answer to the standards problem also derives from Peirce and is more explicitly made by Wiggins (2001) and Misak (1991). Truth is indispensable to belief and assertion – it is what we aim at when we believe and assert. Hard on the heels of the idea that truth is internally connected to inquiry and assertion comes the thought that truth is also internally related to reasons, evidence and standards of good belief. If we unpack the commitments we incur when we assert, we find that we have imported all these notions. Those evolving standards are employable when we ask how concepts should be revised.

Finally, there is another related answer I have offered, building on Peirce’s theory of truth as that which would be indefeasible – that which would stand up to all evidence, experience and argument (Misak 1991, 2000). An important and concrete standard arises from this theory of truth. Since the right concept would take all experience seriously, we have an epistemic (in addition to a moral) reason to pay attention to the experience of all, especially those whose experience has in the past been ignored or denigrated. Those who ignore or denigrate the experience of others show themselves to not be engaged in the search for truth, for that is a search for belief that stands up to the experience of all. To take a current example, anti-vaccine Covid deniers have concepts of research, evidence and much else that does not stand up to the experience of all.

4. Queloz’s solution

Queloz is fully aware of the standards problem and he aims to solve it: ‘Far from ignoring the fact that need ascriptions are contestable’, pragmatic genealogies give us the tools to confront such contestations and to ascertain whether we rightly treat something as need (236).

One route to an answer, which he finds in many of the genealogies he examines, is similar to the pragmatist argument about regulative assumption or core concepts. There are ‘some important needs’ or ‘timeless human problems’, ‘be it the need to avoid conflicts over external goods (Hume), the need to avoid deception within the community (Nietzsche), the need to flag good informants (Craig), the need to gain and share information effectively (Williams), or the need to neutralize prejudice (Fricker)’ (224–5, 231). Once an important need is identified, it can ‘act as a basis for evaluations of the extent to which we have reason to continue to engage in the practice’ (225). He also mentions needs such as being hungry or needing to replenish energy reserves (145). He seems to endorse Craig’s idea that genealogies are ‘at their strongest when the human needs from which they start are the most practical, hence the most undeniable ones’ (236, Craig 1990: 89).

But it is not clear to me what kind of argument Queloz is offering as a justification for employing such needs in our concept assessment. Is it that some of them are impossible to argue against, for life literally depends on them? Is it an argument that the flat earther, unless she stays close to home, won’t make her way around the world successfully and neither will the white supremacist? Or is along the lines suggested above, that there is something about these principles such that we need to assume them, on pain of not holding genuine beliefs that aim at the truth?

If there is no such further argument, then the standards worry persists, as we can push the question back so that it asks how we adjudicate disagreement about whether or not we continue to engage in a practice. The current sorry state of disagreement about vaccines is a case in point. An alarming number of people think that scientific practices, so central to our way of thinking about the world, are part of a conspiracy to put a micro-chip in the arms of people so that they can be controlled for nefarious purposes, including population control and the sexual abuse of children. How do we argue against those who would toss in the rubbish bin concepts so central not only to our way of thinking but to human flourishing? What we tend to do is point to inconsistencies in the anti-vaxxers view (how they rely on the concepts of medical science for so many other things), to evidence that the vaccines work, and to evidence against their conspiracy theories. Those arguments have tended to not work. What, then, do we fall back on? I suggest that it has to be along the lines of the pragmatist arguments I set out above. We can dismiss the anti-vaxxers (while never giving up on them or ceasing to try to convince them) because they betray what it is to aim at the truth – their assertions fail to fit with experience or with what succeeds.

Queloz raises a different worry about the ‘timeless human problem’ solution: it severely restricts the scope of concept assessment. For it looks like assessment is ‘appropriate only when dealing with anthropologically necessary conceptual practices – and surely the greater part of human thought is not necessary in that way’ (231). It seems that he concludes that ‘pragmatic genealogy is neither constitutively committed to there being an enduring core at the centre of the practices it investigates nor restricted to considering only universal needs’; it is ‘not in principle committed to there being an unchanging, timeless core at the centre of the practice it investigates’ (231–2).

But notice that Lewis offers a position on which we can have an enduring (but fallible) core of needs and concepts without restricting assessment to those very needs and concepts. For once we have a core in place, we find that we are committed to many more specific things since our beliefs are inter-connected. And since we are also fundamentally committed to taking experience seriously on Lewis’s pragmatism, we are committed also to countless more peripheral beliefs and concepts. That is, we have a fine example in Lewis on how the core need not be timeless and unchanging, allows us to not be restricted to considering only universal needs and allows us to have something to say to those who deny the principles that underpin success in the world.

Queloz also seems to accept an argument he finds in Williams and Craig, and which I have traced back to Peirce. We aim at shared concepts that are insensitive to the differences between concept-users, concepts which ‘track the objective properties of things that render them suitable to certain uses, irrespective of whether anyone in particular has the need or capacity to use them’ (145). To meet the ‘practical demands on inquirers’ we need to tap into others’ stock of information (194). Something like the following is at the heart of what it is to function as a human inquirer:

Since we are not all at the same place at the same time, but continually take up more or less different points of view, we acquire different information over time. Someone sitting atop a tree may see approaching predators that I cannot see, while I may know what happened here yesterday when others do not. Already in virtue of the fact that we occupy different points of view, therefore, we sometimes need information that others are better placed to acquire than we are, and we therefore have a need for informants. (162)

Similarly, he appeals to Miranda Fricker’s idea (also put forcefully by Peirce) that no one can really be a sceptic – they ‘must be inquirer first, and sceptic second; someone committed to the practical possibility of knowledge first, and committed to undermining that possibility second’ (150, Fricker 2008: 46). This is a fundamental pragmatist argument: what ‘we are all committed to already in virtue of our most basic human needs is to function well’ (202). Some things follow, according to Fricker, for instance, that our system of epistemic division of labour needs to be free of the distorting influence of prejudice. Queloz agrees: ‘It makes good naturalistic sense that testimonial justice should make sense to us as a virtue’ (200–10). We need to ask what can go wrong in our system of epistemic labour or concept formation, in terms of confounding or corrupting influences and epistemic injustices, such as social categorization.

But the standards question continues to press in on this version of the argument. What do we say to say to those who do not want to engage in a cooperative inquiry in a socially heterogeneous community? Perhaps Queloz wants to avail himself of the argument that says that if we aim at truth, we have a built-in requirement to take all voices, especially those that have not been heard, seriously. At times, he seems attracted to it. For instance, he endorses Williams’s point that truth is ‘internally related’ to belief and assertion and that ‘whatever else we take humans to need, it will – as a matter of structural necessity – be the case that each individual has a need for ‘information about the environment, its risks and opportunities’ (162, Williams 2002: 58).

My uncertainty about Queloz’s position on the standards question is amplified by the fact that there are a number of other standards in his mix. He seems to follow Craig in citing the following as the kind of thing that ‘informs our take on extant conceptual analyses’: a concept which makes it so demanding that its extension would effectively be empty is pointless, since nothing would ever satisfy the concept; a concept should stand in the right causal relation to the state of affairs in question; a concept should have proven reliable in the past; it should be able to support justifications; and it should be maximally cost-effective (150–1). Craig’s position, and Queloz’s as well, with respect to these ways of measuring is that there are no necessary conditions that our concepts must meet, only ‘typical conditions: conditions worth tracking because doing so typically pays off’ (150).

Queloz also seems to accept, via Williams, some local standards. Routes to evaluation come from a ‘comprehensive view of our conceptual practices as rooted in a complex historical accumulation of both generic and socio-historically local needs’ (241). Williams’s justification of the concept of liberty, for instance, ‘presents our special concern with liberty as a local manifestation of a near-universal predicament, a manifestation reflecting practical pressures that are distinctive of our situation; and it simultaneously presents the fact that we brought ourselves into this situation as an achievement’ (241). One notes that the ‘near-universal’ is nearly taking us back to the core needs position, but I think that Queloz and Williams do in fact want to suggest that some perfectly legitimate standards can be local to a particular conceptual scheme or form of life.

The standards problem reappears yet again. This time it comes in the form of the question of whether there is one community or many. If the latter, then we will need ways of adjudicating between ways of life that tell us to believe contradictory things or hold competing concepts. I happen to think that the only way to meet this version of the problem is to adopt Davidson’s argument that if we can translate across different so-called conceptual schemes, that shows that there is only one scheme. But I would be interested in hearing what Queloz thinks.

5. Conclusion

I take my comments to be an invitation to Queloz to make his answer to the standards challenge, as I have presented it here, a little more explicit. What concepts can we have or assume, without adopting standards that pretend, impossibly, to be above the fray of inquiry and deliberation? What standards or methodological principles can we help ourselves to – not in a definitive or locked-in way, not written in stone – but which we can justify as employable as we go about our conceptual engineering projects? Since Queloz’s The Practical Origins of Ideas will stand as one of the most important pragmatist treatises on conceptual engineering, it would be good to have his clear answer close to hand so that we can build upon his impressive work.

References

Footnotes