In Extended Rationality: A Hinge Epistemology, Annalisa Coliva aims to by-pass traditional sceptical challenges to the possibility of knowledge by arguing that all thinking and knowing ultimately rely on hinge assumptions which are immune from doubt because of their foundational role in the very framework that makes knowledge and rational thought possible. In defending her position Coliva also rejects the relativist challenge that there could be incompatible but equally plausible systems of justification relying on alternative hinges or assumptions. In this response to Coliva, I argue that even if we accept that we need to rely on some core assumptions in order to get the process of rational thought going, the question of the uniqueness of these assumptions remains open. I maintain that Coliva’s two argumentative strategies against the possibility of relativism, one based on empirical considerations and a second relying on considerations from logic do not guarantee the uniqueness of hinge assumptions and the possibility of at least a moderate form of relativism looms large.
In Extended Rationality: A Hinge Epistemology (2015) Annalisa Coliva develops a sophisticated tightly-argued account of knowledge and rationality that would allow us to by-pass some well-known sceptical challenges, in particular, those focusing on the perceptual justification of knowledge or as she puts it, “the kind of justification we have, based on current sense experience, for propositions about mid-size objects in our environment, which contribute to the content of our beliefs” (Coliva 2015: 3). When it comes to specific empirical beliefs, Coliva argues, perceptual justification is possible only on basis of a prior “system of assumptions” (2015: 3). The assumptions are needed in order to overcome “our cognitive locality,” they enable us to go beyond our personal or subjective phenomenal experiences and to connect such experiences to a “universe populated by physical objects” (2015: 4). Such background assumptions, however, are deemed immune from sceptical doubt because they are part of the framework that has to be in place in order for knowledge and rational thought to be possible.
Sceptics, ancient and modern, on a variety of grounds, have cast doubt on the legitimacy of using foundational assumptions to justify our claims to knowledge. How could we, as critical and fallible thinkers, allow for truths that are beyond any challenge? Is this not just a form of parochial epistemic arrogance or the very dogmatism that as rational thinkers we wish to avoid? If, on the other hand, we try to provide further evidence for our foundational beliefs, then we face the dual dangers of infinite regress in our justificatory attempts or a question-begging form of circularity. Coliva’s book is an attempt to overcome such worries while securing an adequate defence-line against scepticism.
As the book’s subtitle signals, Coliva takes her lead from On Certainty, where Wittgenstein famously claimed that it is a part of “the logic of our scientific investigations that certain things are not doubted” (Wittgenstein 1969: §342). This Wittgensteinian idea is core to what Coliva calls “moderatism”—which she contrasts with the “liberal position” of Pryor and the “conservatism” of Crispin Wright. In this approach, perceptual justifications take place within a background of prior assumptions and at least some of these assumptions, while justified by empirical evidence, receive their legitimacy from the role they play as constitutive elements of epistemic rationality itself. Such background assumptions, however, are not arbitrary, rather, they are the requirements for rational discourse because without them we cannot even begin the process of acquiring evidence and justification for any of our empirical knowledge claims. They are the hinges that make the revolving door of empirical investigation and hoped-for knowledge possible; like a mechanical hinge, they stay put to enable us to open the doors of fruitful empirical enquiry. Broad propositions such as “There is an external world,” “My sense organs work mostly reliably,” “I am not a victim of massive perceptual and cognitive deception,” “the Earth has existed for a very long time,” “there are physical objects” (Coliva 2015: 4ff) are examples of such hinge propositions. Hinge propositions are minimally truth-apt, but they also have a normative role. They are truth-apt because they do have a descriptive content, they are part of our ‘world-picture;’ they are normative because, in certain contexts, “they play the role of norms of evidential significance, they determine what can count as evidence for our ordinary empirical beliefs” (Coliva 2015: 41). The acceptance of such propositions is a pre-condition of epistemic rationality because they are mandated by the very idea of epistemic rationality. They are in that sense, an extension of epistemic rationality or an “extended rationality,” hence, the title of the book. Just as rules of a game, such as football, are part of that game, even if they aren’t moves within a game, the assumptions that are constitutive of epistemic rationality are part of it, even if they are not moves within it. Crucially, assumptions or beliefs that are constitutive of rationality do not require or admit of warrant (Coliva 2015: 129) and therefore the type of sceptical challenge under discussion in this book does not apply to them.
One of my concerns about the book is that even if we accept Coliva’s claim that at least the strongest forms of the sceptic’s challenge are undermined by the hinge epistemological approach, her arguments continue to face a distinct but equally threatening challenge posed by epistemic relativism. Coliva covers the topic of relativism in Chapter 4 of the book, but in what follows I shall argue that her line of defence underestimates the force and the scope of the relativist challenge.
The relativist challenge, as Coliva herself explains, amounts to the claim that there could be, either in practice or in principle, incompatible but equally plausible systems of justification and epistemic practices, each based on different basic assumptions or hinges (Coliva 2015: 140 ff). Rationality, in other words, may be extended but the extensions can take different directions and result in equally incontestable but incompatible knowledge claims. The challenge could be expressed in terms of a “uniqueness requirement.” In a nut shell, the problem is that even if we accept that rational thought is made possible through the unquestioning acceptance of some core principles, or hinges, we don’t have any good reasons for rejecting the possibility that there could be alternative, incompatible, systems of rationality, each with its own starting points or hinges.
The relativist challenge to Coliva’s account, as well as to Wittgenstein’s, also arises from the analogy that Coliva draws between rules of games and the rule-like role that she assigns to ‘hinge propositions.’ Rules, as Coliva and before her Wittgenstein have pointed out, could be constitutive of some practices and in that sense, necessary for and indeed definitive of that practice (Coliva 2015: 131). A game of football, for instance, is defined by a set of rules and once some of the core rules are changed then we are engaging not in football but a different game altogether. After all, that’s how the game of rugby was invented. But the contingency and arbitrariness of rules of a game show something about their standing. Rules, at least rules for games, should be understood as sets of conventions subject to change by collective decision making. Could the same be said of the rules governing our rational discourse? Is there a degree of both variability and arbitrariness in the rules we deem as hinge-propositions for any particular language or mode of discourse? Coliva does not think so.
Suppose we met people who … instead of the physicist … consult an oracle. …—If we call this ‘wrong’ aren’t we using our language-game as a base from which to combat theirs? (Wittgenstein 1969: §612)
Coliva (2010) has argued against relativistic readings of Wittgenstein. In this book, however, she focuses on the charge that her own version of hinge epistemology may be susceptible to the relativist challenge. Chapter 4 of the book is a defence against such a charge and Coliva takes a variety of steps to address the threat. In the first instance, she argues that for the relativist challenge to be credible it must be “intelligible” and compatible with our experiences. The relativist challenge will be toothless, she tells us, if it did not relate to scenarios and beings much like us. And when it comes to such creatures, it is hard to see how else they could form beliefs about material objects if not through the deliverance of their senses. She then extends the claim to all framework assumptions or hinges by suggesting that the core assumptions she has in mind are fundamental to all thinking and that we cannot conceive of thinking creatures who would not subscribe to or utilise such assumptions (2015: ch. 4) or manage their thinking from radically different starting points. Coliva, thus, denies the possibility that there could be conceptual or cognitive frameworks other than our current one; in other words, epistemic principles and core beliefs are unique. Central to Coliva’s thesis is the idea that forming beliefs, based on one’s perceptual experiences, is a “base practice” constitutive of epistemic rationality. The relativist challenge can now be put in the following form ‘Can there be alternative basic epistemic practices of forming beliefs about objects in our surrounding with different characteristic assumptions?’
Coliva offers two different readings of this challenge:
- (i)Could there be other ways of forming warranted beliefs about physical mind-independent objects in our surroundings, “with different and incompatible characteristic assumptions?”
- (ii)Could observation “lead to warranted beliefs about objects in our surroundings, on an understanding of the notion of object different from ours, thus doing without the very assumption that we are interacting with a world of mind independent entities” (Coliva 2015: 142). This second reading, Coliva tells us, is tantamount to considering the perennial question of the tension between common sense and idealism (2015: 142).
Coliva, unsurprisingly, answers both questions in the negative, but surprisingly, couches her response to (ii) in terms drawn from empirical findings, particularly child psychology and some contemporary theories of vision. She argues that “developmental psychology and the science of vision have shown that newborn infants have visual experiences with representational contents of external objects with specific properties” (Coliva 2015: 142). And thereby science not only shows that “our perceptual experience is objective” (2015: 142) but it also provides us with representations of those objects and properties that are actually out there. She bases her views on Tyler Burge’s (2010) empirically informed arguments about perception as a source of objective knowledge. The crucial moves in her argument are (a) we form most of our beliefs on basis of the deliverances of our senses (b) an objectivist, non-idealist or non-phenomenalist view, as developed by Burge, delivers the most plausible current account and understanding of human perception and experience (c) any plausible account of relativism worth its salt has to be in line with “our current understanding of the human condition” (Coliva 2015: 144).
I have two concerns about Coliva’s strategy in countering this particular version of the relativist challenge. The first, and possibly the less important one, is that psychology is a science in flux with very little consensus about its findings. Just to take one example, by appealing to the Burgean objectivist views and the data he uses, Coliva ignores the sizeable and still growing literature on the cognitive penetrability of perception (e.g., Jenkin & Siegel 2015). Cognitive penetrability is the claim that the phenomenal character of our sense experiences, visual, auditory, olfactory etc., depend on and can vary with our background beliefs, desires, or other cognitive states (e.g., Stokes 2013). The approach calls into question the objectivist contention that perceptual judgements are theory-neutral. Instead, the theory of cognitive penetrability advances the claim that the distinction between conception and perception cannot be taken for granted, and that, at a minimum, there is some conceptual basis for all the perceptual data that we use as justificatory evidence in the process of belief formation. It thus calls into question the line pushed by Burge, and Coliva, that basic perceptual beliefs can act as uncomplicated hinges for our objective beliefs about how things stand in the world. The ‘hinges’ themselves, it seems, are contaminated by our beliefs, all the way down.
Coliva could, rightly, point out that the scientific standing of theories of penetrability is far from secure. My point here, however, is not to garner support for the view but to argue that appeal to one of the many empirical positions forwarded by psychologists may not be an adequate basis for dismissing the foundational relativist worries surrounding Coliva’s position. Coliva may be in a position to turn the argument from penetrability to her advantage by maintaining that the conceptual bases of our perceptual beliefs are those very hinge propositions that are constitutive of our extended rationality. But the strategy will not be of much help to her when what is at stake is evidence and justification for the uniqueness of the hinge propositions and there is no evidence of such uniqueness.
The hinge proposition “no one has travelled to the moon” was indeed in line with Wittgenstein’s understanding of the “human condition” at his time, an understanding that has become obsolete through changed circumstances. Faced with Wittgenstein’s ‘travel to the moon’ example and Coliva’s objectivist theory of vision, the relativist could argue for the radical contingency of “our current understanding of the human condition” and the need to relativise what follows from such understanding to a time, place and context. But once that concession to relativism is made, then it is not very clear how the floodgates of relativism could be closed.
Suppose some adult had told a child that he had been on the moon. The child tells me the story, and I say it was only a joke, the man hadn’t been on the moon; no one has ever been on the moon; the moon is a long way off and it is impossible to climb up there or fly there. —If now the child insists, saying perhaps there is a way of getting there which I don’t know, etc. what reply could I make to him? What reply could I make to the adults of a tribe who believe that people sometimes go to the moon (perhaps that is how they interpret their dreams), and who indeed grant that there are no ordinary means of climbing up to it or flying there?—But a child will not ordinarily stick to such a belief and will soon be convinced by what we tell him seriously. (Wittgenstein 1969: §106)
Coliva also dismisses the relativist challenge that there could be “other ways of forming warranted beliefs about physical objects in our surroundings … by means of methods other than observation, with different and incompatible characteristic assumptions” (Coliva 2015: 142). She suggests that what is common to all systems of rationality and indeed to the basic notion of rationality is the practice of producing, assessing and withdrawing from ordinary empirical beliefs, such as the Moorean “here is a hand” example. She also singles out some basic epistemic practices, e.g. modus ponens, as essential to what she calls “logical or deductive rationality” (Coliva 2015: 141). A possible relativist challenge to Coliva’s point here comes from what has been called ‘argument from logical aliens’. The argument does not deal with the observational basis of belief formation but with the possibility, in Coliva’s words, of “different and incompatible characteristic assumptions.” (Coliva 2015: 141). Coliva when introducing and defending her position against the charge of relativism had emphasised that she is not dealing with the sort of relativism that could ensue from adherence to different logical principles, because she has dealt with the issue in previous publications. But the particular variety of relativism I discuss below is not covered in her earlier work.
A variant of the argument from “logical aliens” is raised, rather tendentiously, in Wittgenstein’s description of the “wood-sellers” who sell wood by area covered by their goods, without taking into account the height of the piles of wood they are selling. Coliva (2010) has denied that the example describes a coherent form of logical thinking. But a moderate and more plausible example of logical aliens is found in Hilary Putnam’s mereological sum thought experiment. To remind ourselves, Putnam invites us to imagine a world with three individuals a, b, c. He then asks: “How many objects are there in this world?” One possible and to most of us obvious answer is 3 (a, b, c). But in the world of Polish mereologists the natural response to the question “How many?” is 7 (a, b, c, a+b, a+c, b+c, a+b+c) (Putnam 1994). The thought, following Putnam, is that how we identify “individuals” and “objects” does not depend on perception and concepts only, but also on the logical tools and system of categorisation we bring to bear on these observations. Even when we grant to Coliva that belief formation relies heavily on observation, there could still be a strong disagreement on what counts as an object. Nothing in Coliva’s account, other than Burgean objectivism, stops the logically alien mereologist to individuate and count ‘objects’ in ways that are not compatible with ‘our’ ways of counting or individuating objects. In Chapter 5 of the book, Coliva advances arguments for the necessity of relying on basic rules of logic such as modus ponens (Coliva 2015: 161) in forming any rational belief, but the tribe of mereologists would be happy to go along with Coliva’s arguments about the basic moves of deductive logic and the reliance on modus ponens while delivering answers to the question ‘how many objects?’ that are incompatible with our verdicts. It is not clear how hinge epistemology will stop this weaker form of relativism about our basic concepts.
Relativism threatens in at least one other way. The threat arises from the contingent truth of supposedly hinge propositions such as “my sense organs work mostly reliably,” “the Earth has existed for a very long time,” etc. Such propositions are contingently true in the sense that we could imagine circumstances where their truth would not hold, e.g., in circumstance when our sense organs do not work reliably or in a scenario where an earth-like planet, unbeknownst to its would-be inhabitants, is artificially created to accommodate the remnants of humans. In such circumstances, it would be true to say that “I cannot trust my senses” or that “the earth has not existed for a long time.” Coliva would, rightly, respond that false propositions cannot be the hinges on which our knowledge claims turn, and that our acceptance of the truth of hinge propositions depend on some ceteris paribus clauses. But both the sceptic and the relativist could point out that their very starting point is that we are not in a position to know which of the two worlds we inhabit or whether we are in the relevant world or circumstances where the hinge propositions advanced by Coliva are true. We simply do not know whether the relevant ceteris paribus clauses hold. We grant to Coliva that, in order to get the claims to our empirical knowledge off the ground, we need to have some core assumptions in place, but the question about the uniqueness of these assumptions remains open. Thus, while she may be correct in arguing for the form of rational thought and beliefs, the content of our beliefs remains open to relativistic, and also sceptical, doubt.
Coliva’s book is a wonderful example of how to discuss some of the most fundamental questions of epistemology with a fresh eye on older approaches while creating a new tool-kit for solving these ancient problems. The problem is that two of the “most fundamental problems” of epistemology, scepticism and relativism, have acquired the honorific title “most fundamental” because they keep eluding even the most imaginative of solutions.
Jenkin Z. & Siegel S. (2015). “Cognitive Penetrability, Modularity, Epistemology, and Ethics,” Review of Philosophy and Psychology 6: 531–545.