EXTENDED KNOWLEDGE

The aim of the 1st International Conference on the topic of 'Extended Knowledge' was to bring together leading scholars from around the world in order to critically examine the ramifications of the extended cognition programme (and related views) to contemporary epistemology, and to the theory of knowledge in particular.

The main event took place on Wednesday and Thursday, followed by two satellite workshops on Friday, April 24. The first satellite (morning session) was dedicated to the Epistemology of Education, and the second to Group Knowledge. Adam Toon (University of Exeter) joined to kick-off the event on Tuesday, April 21, with an 'Extended Knowledge' reading group dedicated to his paper 'Where Is Understanding' (7 George Square, room G. 32, at 15:30).

The following is an indicative list of the more specific questions that the above events addressed:

Is extended epistemology compatible with epistemic internalism?
Does extended epistemology entail that we should treat epistemic virtues as environmentally extended?
Are standard views about the nature of epistemic virtues—such as those offered by virtue epistemology—compatible with the idea that our epistemic characters can be extended?
What kind of cognitive responsibility is involved in extended epistemology?
Does extended epistemology have implications for epistemic value? Is extended epistemology compatible with the idea that knowledge has a distinctive value?
What are the connections between the notions of extended and distributed cognition as pursued in the Western intellectual tradition and related notions as approached from within Eastern (e.g., Chinese) thinking?
To what extent can the epistemic virtuesor even cognitive abilities more generallybe thought of as distributed across several individual agents?
Are the questions posed by the extended and distributed cognition hypotheses the same for all epistemic standings, or do they differ depending on the kind of epistemic standing in play (e.g., whether it is understanding rather than knowledge we are focusing on)?
What can the answers to the above set of questions tell us about the epistemology of education? Can they provide significant information for the design of future educational equipment?
Is it possible for groups to acquire knowledge in the same way individuals do? Are there any such real-life examples and if so, what insights can extended epistemology provide for their more efficient design in the future?

Below is the programme as it happened:

WEDNESDAY, APRIL 22

09:30 - 10:00: Tea/Coffee

10:00 - 11: 10: ‘Extended Knowledge, the Recognition Heuristic, and Epistemic Injustice’

Mark Alfano & Gus Skorburg (Philosophy, University of Oregon)

Abstract: Following Palermos’s (forthcoming) pioneering application of dynamical systems theory to the extended mind debate, in recent work (Alfano & Skorburg forthcoming), we have developed a principled way to distinguish between cases of cognitive embedding and cases of outright cognitive extension. We here deploy that distinction to argue that the effectiveness of common heuristics (availability, representativeness, recognition) is an example of embedding. We focus in particular on the recognition heuristic. Recognition is the thinnest and most basic form of memory: to recognize something, all you need to do is be aware that you’ve encountered it before. You needn’t remember any particular information about it, how you came to be acquainted with it, or what its nature is. Gigerenzer and his colleagues have shown that people often use their recognitional capacities in decision-making: if you recognize only one of two cities, you will tend to infer that therecognized city is more populous than the unrecognized city, and if you recognize only one of two universities, you will tend to infer that the recognized university is more prestigious than the unrecognized university. This heuristic works, when it does, because newspapers and other media tend to discuss bigger cities and prestigious universities more than they discuss smaller cities and less prestigious universities. Moreover, people know this, if only implicitly, and therefore rely on the “if that were true, I would have heard by now” inference (Goldberg 2011). The reliability of the recognition heuristic for a particular criterion variable (e.g., population) in a given context (e.g., Western European cities) depends on the extent to which intensity of media coverage correlates with the criterion. Using data mining and a new experimental results, we show that there are systematic biases in the US press against reporting on cities outside the United States and Western Europe. We conclude by placing these biases in the context of epistemic injustice.

11:10 - 12: 20: Emerging Digital Technologies: Implications for Extended Conceptions of Cognition and Knowledge’

Paul Smart (Computer Science, University of Southampton)

Abstract: A recent focus of attention in both the philosophy of mind and contemporary epistemology is the notion of extended cognition: the idea that the material mechanisms associated with the realization of certain kinds of mental states and processes extend beyond the biological realm to encompass aspects of the surrounding technological and/or social environment. Emerging digital technologies have an important role to play in terms of shaping the philosophical debate in this area. Firstly, new technologies can highlight specific areas of interest and attention for philosophical discussion. The advent of new technologies, such as wearable computing devices, ambient computing, augmented reality, linked data, and brain-machine interfaces all have the potential to transform the kind of technological landscape against which philosophical debates and scientific interests are formulated. The ethical debates concerning privacy, for example, have gained a new momentum as a result of the growing awareness associated with the novel surveillance opportunities that the Internet provides to both government agencies and criminal fraternities. Secondly, emerging digital technologies provide a focus for philosophical debates relating to the future transformation of our cognitive abilities. The reliability, portability, and accessibility of the technologies, as well as the extent to which they can be deemed to be an integrated part of our cognitive characters are features that are of interest to those seeking to isolate concrete instances of extended cognition and extended knowledge. By understanding some of the capabilities and limitations of a range of new technologies, the philosophical community is better enabled to evaluate whether these technologies satisfy the conditions for cognitive extension. In addition, the feedback gleaned from this philosophical debate serves as a potential driver for future forms of technological innovation and development by the engineering community. Thirdly, the social dimension of emerging digital technologies is relevant to discussions concerning distributed cognition and social epistemology. New technologies are, typically, embedded within the broader socio-technical ecology of the World Wide Web, and this has provided a wealth of opportunities for novel kinds of information processing and information distribution, as well as arguably transforming the nature of many different kinds of social processes. In view of the extended cognition thesis, it is natural to wonder whether this reshaping of our social interactions and problem-solving processes has any bearing on the philosophical debates relating to the technological transformation of our cognitive and epistemic capabilities. It is also important to consider the way in which the advent of the Web is giving rise to new forms of machine-based intelligence; ones in which human contributions play an important role in terms of scaffolding and shaping the nature of machine-based processing. This potentially introduces a new dimension in the philosophical debate relating to extended cognition. Whereas much of the current debate is concerned with the extent to which new technologies will impact the cognitive and epistemic profile of human subjects, we can now also begin to ask whether the advent of the Web enables collections of human agents to participate in the extended processing loops that realize the next generation of intelligent machines.

12:20 - 13:30 ‘Explanatory Virtues, Extended Minds, and the Primeness of Knowledge’

Zoe Drayson (Philosophy, University of Stirling)

Abstract: Two radically externalist approaches to the mental have been proposed that go far beyond the content externalism of Putnam and Burge. Both Williamson (1995, 2000) and Clark & Chalmers (1998) argue that environmental factors are relevant to more than just the contents of mental states. Williamson argues that a factive state like knowing, which has traditionally been considered a composite of mental and non-mental environmental conditions, is in fact a ‘prime’ (non-composite) genuinely mental state⎯with important implications for epistemology. Clark and Chalmers argue that the physically-implemented vehicles of our mental states needn’t be confined to the brain, but can extend into the extra-bodily environment⎯with important implications for the methodology and practice of cognitive science. These two radically externalist approaches differ in their metaphysical presuppositions and their implications, but they appeal to similar explanatory considerations to justify their externalist claims about mental states: Williamson argues that there are cases where knowledge-citing psychological explanations are better explanations than their belief-citing alternatives, while Clark and Chalmers claim that their extended notion of belief is more useful than a non-extended notion in psychological explanations. Both parties have had the specific details of their appeals to explanatory virtue challenged, but in this paper I focus on the more general form of these arguments. Drawing on work by Potochnik (2010), Weslake (2010), and Strevens (2008), I examine the value of explanatory generality and its relation to other explanatory virtues, in order to reassess Williamson’s argument for the primeness of knowledge and Clark and Chalmers’s argument for extended minds.

13:30 - 14:15: Lunch

  14:15 - 15:25: ‘Extended Cognition, Distributed Knowledge and (Liberal) Education’

Ben Kotzee (Education, University of Birmingham)

Abstract: In educational thought, what one can call traditional school knowledge (such as factual knowledge of subjects, skill at mental calculation, etc.) is just emerging from a long losing streak. Progressive educational thinkers have long attacked traditional school knowledge as stultifying, oppressive and insensitive to students' own needs or interests. In its own way, the idea of extended cognition (and perhaps also of distributed cognition) has played a part in this attack on traditional school knowledge. Briefly put, if improving technology makes it possible for children to rely more and more on external aids (pocket calculators, Google, etc.) to bolster and extend their own mental resources, is it still necessary to teach traditional school knowledge (such as mental arithmetic, recall of facts, etc.)? And if knowledge is distributed between whole groups of people, what happens to the educational ideal that each child should have a broad, liberal education? Could we not teach individual children each part of the sum of human knowledge (perhaps what they are particularly interested in or talented at), confident that, when these children make up a group of knowers, the group as a whole will possess all the knowledge needed? In this paper, I consider the problems and prospects that the ideas of extended and distributed cognition pose for education. I hold that the challenge is not as straightforward as traditionally conceived. To make proper use of extended knowledge aids (such as Google or the pocket calculator), I hold, one still needs much traditional school knowledge. Moreover, to play a constructive role in a group of knowers, individual knowers with different specialisms need to know enough about one another’s concerns to be able to communicate⎯comprehensive knowledge helps one to be a good player in a system where there are lots of different knowers with their own specialism. This aside, considering extended and distributed cognition in educational thought helps to illuminate the real value of a traditional, broad theoretical education in the arts and sciences. Other proponents of the ideal of a liberal education (e.g., RS Peters) hold that its value is intrinsic or definitional⎯it follows from an analysis of the idea of the educated person, that that person must have had a liberal education. Rather than that, I will hold that the value of a liberal education lies in how it makes available central concepts and methods of thinking to the student that allows her to make sense of a large number of different specialist fields and thereby makes her fit for epistemic cooperation with others. The value of liberal education, I will hold, can therefore be located through consideration of distributed cognition.

15:25 - 16:35: ‘What is Extended Knowledge?’
  Fred Adams (Philosophy/Cognitive Science, University of Delaware)
Abstract: TBC

16:35 - 16:50: Tea/Coffee

   16:50 - 18:00: ‘Tracking in Encultured Cognitive Systems’

Richard Menary (Cognitive Science/Philosophy, Macquarie University)

Abstract: There is no doubt that biological evolution has produced different kinds of tracking mechanisms across different kinds of species suited to their different environmental niches (Sterelny 2003). However, cultural evolution has produced tools and artefacts for tracking particular environmental variables⎯such as: maps, telescopes, sextants and so on. There are also abstract systems of representation that track numbers, sets, sounds and phonemes. The first part of this paper will look at recent cultural evolution and its effect on human cognition. This paper argues that the development of artefacts and systems of representation lead humans to develop ‘enculturated’ cognitive mechanisms in ontogeny. Phenotypic and neural plasticity are the domain general mechanisms by which existing neural and cognitive systems become redeployed to perform new cultural functions (Dehaene 2009). The final part of the paper will concern the truth tracking function of extended cultural systems, in relation to recent work on extended cognition and epistemology.

18:00: Dinner Buffet

THURSDAY, APRIL 23

  09:30 - 10:00: Tea/Coffee

10:00 - 11:10: ‘Extended Cognition, Trust and Glue and Knowledge’

Kenneth Aizawa (Philosophy/Cognitive Science, Rutgers University)

Abstract: Some recent work in Extended Epistemology has looked not at the relationship between extended cognition per se and epistemology. Instead, it has examined the relationship between the version of extended cognition that relies on conditions of “trust and glue” and epistemology. This paper begins with this clarificatory point, then sounds a cautionary note. The cautionary note is a new argument for the view that the conditions of “trust and glue” commit a version of the infamous “coupling-constitution” fallacy.

11:10 - 12:20: 'Reflection and Extended Knowledge: Confucius and Ernest Sosa'

Michael Mi (Philosophy, University of Soochow)

Abstract: In this article I examine the role that reflection plays in knowledge. I argue that a notion of reflection grounded in ancient Chinese philosophy can help us understand extended knowledge and reflective knowledge in both the accounts of Confucius and Ernest Sosa. I also argue that reflection can help us understand the most ideal kind of knowledge or knowing full well. The article begins by laying out Confucius's and Sosa's accounts of knowledge, while at the same time drawing the reader's attention to their common concern with reflection and extended knowledge. Next I draw on an account of reflection from Confucius and elaborate on it. With this account of reflection in hand, I return to Confucius's and Sosa's accounts of knowledge and show how this account of reflection can help to explain their accounts of extended knowledge.

12:20 - 13:30: 'What 'Extended Me' Knows'

Andy Clark (Philosophy, University of Edinburgh)

Abstract: Arguments for the ‘extended mind’ seem to suggest the possibility of ‘extended knowers’ – agents whose specifically epistemic virtues may depend on systems whose boundaries are not those of the brain or the biological organism. Recent discussions of this possibility invoke insights from virtue epistemology, according to which knowledge is the result of the application of some kind of cognitive skill or ability on the part of the agent. In this talk, I argue that there is a fundamental tension in these appeals to cognitive virtue. The tension centers on the presence of a tool or technology as an object of awareness, hence something apt for epistemically virtuous engagement on the part of the agent. I highlight a dilemma: the better something looks as a non-biological element of the machinery of mind, the worse it looks as a potential object of any specifically epistemic skill or ability on the part of the agent. This tension has practical implications for both the design and the ‘epistemically secure’ adoption of new technologies.

13:30 - 14:30: Lunch


14:30 - 15:40: ‘Representations and Group Cognitive States’

Jeroen de Ridder (Philosophy, Free University Amsterdam)

Abstract: One argument against the existence of robustly collective cognitive states such as group belief or group knowledge is that there are no collective representations, i.e., representations held by groups rather than individuals. Since belief requires representation, so the argument goes, there can be no collective belief. In this paper, I reply to this argument. First, I’ll scrutinize the assumption that belief requires representation and point out that it is in fact a substantive and controversial issue whether belief indeed requires representation and, if it does, how so. Secondly, I’ll argue that even if we grant the above assumption, the argument can be resisted, since there is a natural way to make sense of collective representations. By drawing on the ideas of the extended mind and distributed cognition, I’ll outline how we can conceive of collective representations and thereby undermine the argument against group cognitive states.

  15:40 - 17:00: 'Location, Location, Location: Can Epistemic Virtues be Distributed?’

Heather Battaly (Philosophy, California State University, Fullerton)

Abstract: Thus far, virtue epistemology has been largely concerned with individual agents. It has focused, in part, on the conditions under which an individual agent has an epistemic virtue. But, epistemic virtues may not be restricted to individuals. We sometimes attribute epistemic virtues to groups; as when we say that a research team was open-minded, or a committee was tenacious, or a police investigative unit was conscientious. Recent work on distributed cognition may enable us to explain how a group can have an epistemic virtue. This paper explores whether epistemic virtues, as they are described by virtue epistemologists, can be distributed across groups. Virtue epistemologists have developed two key analyses of epistemic virtue. Virtue-reliabilists, like Ernest Sosa and John Greco, have argued that epistemic virtues are (roughly) stable qualities that reliably produce true beliefs. These qualities range from hard-wired capacities like vision, to acquired skills like using a microscope to diagnose a patient, to character traits like openmindedness. Virtue-reliabilists think that we share some of our epistemic virtues, like vision and memory, with other animals. In contrast, virtue-responsibilists, like Linda Zagzebski and Jason Baehr, restrict epistemic virtues to acquired character traits for which the individual agent can be praised. Such traits include: open-mindedness, intellectual tenacity, intellectual humility, and conscientiousness. Zagzebski and Baehr agree that epistemic virtues are personal qualities which require good motivations, but they disagree about whether epistemic virtues require reliability (Baehr thinks they do not). Despite their differences, virtue-reliabilists and virtue-responsibilists have all been thinking of epistemic virtues as attributable to individual agents. They have made the natural assumption that epistemic virtues are dispositions that are located entirely within the head of the individual agent who possesses them. But recent work on extended and distributed cognition challenges this assumption. It argues that cognitive processes need not be entirely located in an individual agent’s head—they can extend beyond a single agent’s head to include objects in the environment, and even other people. Such external objects and people are taken to be constituent parts (rather than independent causes) of the cognitive process itself. In short, cognitive processes may be located partly outside of an individual agent’s head, or may even reside in a group of people rather than a single agent. Advocates of extended cognition typically motivate their views via functionalism and parity: roughly, they argue that if a part of the world instantiates a functional role that we would recognize to be part of a cognitive process, were we to find it inside an agent’s head, then that part of the world is part of a cognitive process (Clark & Chalmers). At least it is part of a cognitive process, provided there is mutual interaction and interdependence between it and other parts of the process. In a similar vein, advocates of distributed cognition typically argue that for a cognitive process to be distributed across a group of people, function must be preserved, and the group must exhibit cognitive interdependence, interaction, and (perhaps) convergent expectations (Theiner). For advocates of extended and distribution cognition, location doesn’t matter. What matters is function, and interdependence. Does location matter for epistemic virtues, as they are described by virtue-reliabilists and virtue-responsibilists? Granted, virtue epistemologists have assumed that epistemic virtues are located in individual agents, rather than groups. But, must they make this assumption? Is there anything to prevent epistemic virtues from being cognitively distributed? This paper suggests that the features of reliabilist virtues do not prevent them from being located in groups. Indeed, memory is one of the cognitive processes already being studied by advocates of distributed cognition. But, responsibilist virtues are another matter. Responsibilists argue that reliability is not enough (or, in some cases, is not even required) for epistemic virtue because: (1) reliability doesn’t tell us anything about the agent as a person—it doesn’t tell us what the agent values or cares about; and (2) reliability is subject to luck, and so the agent need not be praised for it. Responsibilists conceive of epistemic virtues as traits that express who an individual agent is as a person, and traits for which that individual agent can be praised. So, on the face of it, location in the individual agent matters more for responsibilism than reliabilism. There is thus a disanalogy between the two sorts of virtues; a disanalogy that results from the key differences in the ways these virtues are conceived. But, could these features of responsibilist virtues nevertheless be extended to a group? That will depend on whether groups have values and motives. It will also depend on whether groups can be praised for possessing traits like openminddness. (Are groups partly responsible for acquiring such traits? What happens when a group possesses a trait that none of its individual members possesses? Is the group still partly responsible for acquiring the trait?) Perhaps we can look to collective epistemology (and collective responsibility) for help. In sum, while there are no features of reliabilist virtues that prevent them from being located in groups, responsibilist virtues are more complicated. If we can’t make sense of group responsibility for a trait, or of group motives and values, then the location of responsibilist virtues will be restricted to individuals. And if that happens, we might wonder which we should revise: the notion of responsibilist virtues, or the notion that groups can have traits like open-mindedness?

17:00: Dinner Buffet

FRIDAY, APRIL 24

  MORNING SESSION: EXTENDED KNOWLEDGE AND EDUCATION SATELLITE


09:30 - 10:00: Tea/Coffee

10:00 - 11:00: 'Factitious Intellectual Virtues'
Mark Alfano (Philosophy, University of Oregon)
Abstract: TBC

11:00 - 12: 00: 'Skillful Reflection as an Epistemic Virtue'
Chienkuo Mi (Philosophy, Soochow University)
Abstract: TBC

12:00 - 13:00: 'Responsibilist Virtues in Reliabilist Classrooms'
Heather Battaly (Philosophy, UC-Fullerton)
FULL PAPER here

13:00 - 14:00: Lunch

  AFTERNOON SESSION: GROUP KNOWLEDGE SATELLITE

14:00 - 15:00: 'Social Machines'

Paul Smart (Computer Science, University of Southampton)

Abstract: A key trend in the recent technological evolution of the Web has been the development of applications and services that support greater levels of user participation in the generation and management of online content. This has given rise to what has been referred to as the ‘Social Web’: a suite of applications, services, technologies, formats, protocols and other resources, all united in their attempt to both foster and support social interaction and information exchange. At the heart of this Social Web are systems, such as YouTube, Wikipedia, Facebook and Twitter. These are among some of the most popular systems on the Web today. The rise of the Social Web has led to renewed interest in large-scale socio-technical systems, and a panoply of new terms have appeared to refer to parts of the emerging conceptual landscape. We thus have terms such as crowdsourcing, human computation, collective intelligence, social computing, the social operating system, the global brain, and social machines. The current talk will focus on the notion of social machines, although the relationship between social machines and other kinds of technology-mediated social participation system will be covered. Social machines can be seen as integrated systems that combine the representational and computational resources of both humans and machines, often at large scale. They provide a means for computing technologies to play an ever greater role in the moulding and (perhaps) the mechanistic realization of a variety of social processes, with significant implications for the shape of human society and the social agents that live within them. In addition, social machines act as a mutually-permeable cognitive frontier: they sit ready to re-define our collective cognitive and epistemic capabilities; however, they also enable machine-based processes to draw on human capabilities in new and interesting ways. The result is a potential transformation of our usual notions of machine intelligence, with human society serving to extend the cognitive reach of a variety of machine-based processes.

  15:00 - 16:00: 'On Human becoming: Creative thinging and incompleteness'

Lambros Malafouris (Anthropology, University of Oxford)

Abstract: Humans can be described as deep time creative thingers. That is, we are embodied and self-conscious beings able to influence our developmental paths by changing our means of material engagement. In short, we create material things and assemblages which scaffold the ecology of our minds and shape the boundaries of our thinking. This openness of the human mind to creative evolution is one of the distinctive features of our species. It also indicates that the nature of human intelligence is to remain amenable to deep reorganization and reconstitution, thus, to remain incomplete. This paper will explore what exactly are the major implications of this ongoing dialectic of co-constitution of people and things for the traditional (neo-evolutionary) ways we understand the importance of embodiment, plasticity, and material culture in human evolution.

16:00 - 16: 15: Tea/Coffee


  16:15 - 17:15: 'Discovering our Blind Spots: Dialogic Teaching and Group Learning'

Andrea English (Education, University of Edinburgh)

Abstract: When we imagine the traditional classroom with the teacher as a transmitter of knowledge, and each student sitting silently at a desk as a passive recipient, it may not be immediately clear that there is a connection between teaching and group knowledge. This model of the “teacher-as-transmitter” and “learner-as-recipient” is part of a deeply ingrained commonsense understanding that is reinforced by trends in international educational policy on high-stakes testing and accountability. Such policies imply that knowledge is deposited as “input” into the individual learner’s mind and can be readily tested as “output”. However, these ideas of the teacher and learner are countered by research on dialogic forms of education, which are rooted in an understanding of the human being as one who learns in interaction with others and with the environment. From a dialogic perspective, the teacher’s task is viewed as facilitating such interaction.

In this paper, I argue that dialogic teaching is indispensable to education, because it initiates and draws out learners’ engagement with educative forms of uncertainty and unsettledness, moments that I have elsewhere termed “discontinuities” in learning (English, 2013). Discontinuity points to a moment of confrontation with the limits of knowledge and ability that occurs when we encounter something unexpected — when we discover a “blind spot” — and in turn become uncertain. In the first section, I draw upon pragmatism and phenomenology to clarify the concept of learning, with a focus on the role of discontinuity. In the second section, I examine the concept of dialogic teaching and its role in facilitating group learning. Using a classroom example, I illustrate dialogic interaction between a teacher and learners with the aim of pointing out certain conditions that make learning in groups possible, including the subsequent cognitive gains in knowledge by group members. A key point I seek to draw out is the interconnection between cognitive and moral development in group learning. In closing, I underscore significant challenges for the practice of dialogic teaching and consider how future interdisciplinary research on dialogue, group knowledge, and distributed cognition might serve to address these challenges


17: 15 - 18:15: 'Two (or More) Senses of Group Knowledge'

Jeroen de Ridder (Philosophy, VU University Amsterdam)

Abstract: Talk about group knowledge is ubiquitous. We ascribe knowledge to various kinds of collectives: committees, boards, organizations, research teams, scientists, countries, and even humankind. Various philosophers have argued that at least some of these group knowledge ascriptions must be construed realistically — as referring to genuinely collective epistemic states — and have proposed an analysis of group knowledge that can account for such group knowledge ascriptions. Such accounts, however, have been plagued by systematic objections and counterexamples.

In this paper, I propose a different strategy and argue for pluralism about group knowledge: group knowledge is not one unique thing, but there are a number of collective epistemic states that all deserve the label of group knowledge. Extant accounts home in on some of these states, mistakenly to the exclusion of others. Group knowledge, then, has a number of senses.

While this may seem like a theoretical cop-out, I will also argue that there are in fact good systematic reasons to expect that there is more than one sense of group knowledge. The epistemologist’s paradigm case of knowledge is individual knowledge. Because, trivially, groups aren’t individuals, group knowledge is going to differ in some respects from individual knowledge and since there are different respects in which group knowledge can differ from individual knowledge, we should expect there to be different senses of group knowledge.: