The representations of deep convolutional neural networks (CNNs) are formed from generalizing similarities and abstracting from differences in the manner of the empiricist theory of abstraction (Buckner, Synthese 195:5339–5372, 2018). The empiricist theory of abstraction is well understood to entail infinite regress and circularity in content constitution (Husserl, Logical Investigations. Routledge, 2001). This paper argues these entailments hold a fortiori for deep CNNs. Two theses result: deep CNNs require supplementation by Quine’s “apparatus of identity and quantification” in order to (1) (...) achieve concepts, and (2) represent objects, as opposed to “half-entities” corresponding to similarity amalgams (Quine, Quintessence, Cambridge, 2004, p. 107). Similarity amalgams are also called “approximate meaning[s]” (Marcus & Davis, Rebooting AI, Pantheon, 2019, p. 132). Although Husserl inferred the “complete abandonment of the empiricist theory of abstraction” (a fortiori deep CNNs) due to the infinite regress and circularity arguments examined in this paper, I argue that the statistical learning of deep CNNs may be incorporated into a Fodorian hybrid account that supports Quine’s “sortal predicates, negation, plurals, identity, pronouns, and quantifiers” which are representationally necessary to overcome the regress/circularity in content constitution and achieve objective (as opposed to similarity-subjective) representation (Burge, Origins of Objectivity. Oxford, 2010, p. 238). I base myself initially on Yoshimi’s (New Frontiers in Psychology, 2011) attempt to explain Husserlian phenomenology with neural networks but depart from him due to the arguments and consequently propose a two-system view which converges with Weiskopf’s proposal (“Observational Concepts.” The Conceptual Mind. MIT, 2015. 223–248). (shrink)

The cognitive scientist Donald Hoffman argues that we don't perceive reality: spacetime, objects, colors, sounds, tastes, and so forth, are all merely an interface that we evolved to track evolutionary fitness rather than to perceive truths about external reality. In this paper, I expound on his argument, then I extend it, primarily, by looking at key ideas in physics that are quite germane to it. Among the topics in physics that I discuss are black holes, the holographic principle, string theory, (...) duality, quantum gravity, and special relativity. I discuss these ideas from physics with an eye to their relevance for Hoffman's view. (shrink)

I argue that Sellars’s distinction between signifying and picturing should be taken seriously by philosophers of mind, language, and cognition. I begin with interpretations of key Sellarsian texts in order to show that picturing is best understood as a theory of non-linguistic cognitive representations through which animals navigate their environments. This is distinct from the kind of discursive cognition that Sellars called ‘signifying’ and which is best understood in terms of socio-linguistic inferences. I argue that picturing is required because reflection (...) on signifying cannot adequately explain our need for cognitive friction. I then show how the idea of picturing is further developed by Paul Churchland, Ruth Garrett Millikan, and Huw Price. I finally turn to predictive processing as a theory of cognitive representation, and in particular Andy Clark’s ‘radical predictive processing’, to further characterize picturing. However, doing so has the cost of pushing picturing and signifying further apart than Sellars intended. (shrink)

How can we think about things in the outside world? There is still no widely accepted theory of how mental representations get their meaning. In light of pioneering research, Nicholas Shea develops a naturalistic account of the nature of mental representation with a firm focus on the subpersonal representations that pervade the cognitive sciences.

The historical debate on representation in cognitive science and neuroscience construes representations as theoretical posits and discusses the degree to which we have reason to posit them. We reject the premise of that debate. We argue that experimental neuroscientists routinely observe and manipulate neural representations in their laboratory. Therefore, neural representations are as real as neurons, action potentials, or any other well-established entities in our ontology.

Implicit and explicit attitudes manifest themselves as distinct and partly dissociable behavioral dispositions. It is natural to think that these differences reflect differing underlying representations. The present article argues that this may be a mistake. Although non-verbal and verbal measures of attitudes often dissociate, this may be because the two types of outcome-measure are differentially impacted by other factors, not because they are tapping into distinct kinds of representation or distinct storage systems. I arrive at this view through closer consideration (...) than is usual of the mechanisms and processes that underlie overt behavior. (shrink)

Edgington has proposed a solution to the sorites paradox in terms of ‘verities’, which she defines as degrees of closeness to clear truth. Central to her solution is the assumption that verities are formally probabilities. She is silent on what verities might derive from and on why they should be probabilities. This paper places Edgington’s solution in the framework of a spatial approach to conceptualization, arguing that verities may be conceived of as deriving from how our concepts relate to each (...) other. Building on work by Kamp and Partee, this paper further shows how verities, thus conceived of, may plausibly be assumed to have probabilistic structure. The new interpretation of verities is argued to also help answer the question of what the verities of indicative conditionals are, a question which Edgington leaves open. Finally, the question of how to accommodate higher-order vagueness, given this interpretation, is addressed. (shrink)

My dissertation establishes the basis for a systematic outlook on the role language plays in human cognition. It is an investigation based on a cognitive conception of language, as opposed to communicative conceptions, viz. those that suppose that language plays no role in cognition. I focus, in Chapter 2, on three paradigmatic theories adopting this perspective, each offering different views on how language contributes to or changes cognition. -/- In Chapter 3, I criticize current views held by dual-process theorists, and (...) I develop a picture of the complex interaction between language and cognition that I deem more plausible by using resources from the literature on the evolution of the faculty of language. Rather than trying to find one general explanation for all cognitive processes, I take seriously the idea that our mind is composed of many subsystems, and that language can interact and modify each in different ways. There is no reason offered in the empirical literature—besides maybe parsimony—that suggest that language has to interact in the same ways with all cognitive processes. Yet, this is seemingly taken for granted, especially within dual-process approaches. -/- On my view, it is a central requirement for a theory of the role of language in cognition to explain how language might have effects, at once, on and within various parts of cognition. In Chapter 4, I explore how this framework can modify how we think about some experiments in psychology, specifically in research on categorization. My idea is that language, once it evolved, changed how some cognitive capacities worked and interacted with each other, but did so in more than one or two ways. Cognitive systems are changed in very different ways—sometimes the transformation is very subtle, such as our way of forming categories by using how similar objects are, while other times it is deep and changes the very way the system works. (shrink)

Colombo (Phenomenology and the Cognitive Sciences, 2012) argues that we have compelling reasons to posit neural representations because doing so yields unique explanatory purchase in central cases of social norm compliance. We aim to show that there is no positive substance to Colombo’s plea—nothing that ought to move us to endorse representationalism in this domain, on any level. We point out that exposing the vices of the phenomenological arguments against representationalism does not, on its own, advance the case for representationalism (...) one inch—beyond establishing its mere possibility. We criticize the continual confounding of constitutive and explanatory claims and the lack of recognition of a Hard Problem of having to provide a naturalistic account of content, coupled with an inability to face up to it. We point at the inadequacy of various deflationary moves that end up driving representationalists towards the idea of neural representations with non-standard contents or without content altogether, both of which either render neural representationalism unfit for purpose or vacuous. Referring to possibilities for neural manipulation and control, or to established scientific practice does not help representationalism either. (shrink)

According to the deliberative view of democracy, the legitimacy of democratic politics is closely tied to whether the use of political power is accompanied by a process of rational deliberation among the citizenry and their representatives. Critics have questioned whether this level of deliberative capacity is even possible among modern citizenries--due to limitations of time, energy, and differential backgrounds--which therefore calls into question the very possibility of this type of democracy. In my dissertation, I counter this line of criticism, arguing (...) that deliberative democrats and their critics have both idealized the wrong kind of citizen deliberation. Citizen deliberation should not be concerned with the indeterminate project of "translating" abstract democratic principles and values into everyday cases of political problem-solving. Instead, deliberation should take the form of analogy, just as we already find it in everyday politics and affairs. When ordinary citizens use analogies, they do not derive decisions from general principles or values, but they still reason nonetheless. Seen from this analogical perspective, deliberative democracy is already a practical reality to a large degree. When an election is on the horizon, a campaign season arises in which debates, forums, and "barstool" dialogues exponentially increase the amount of citizen deliberation. In these settings, citizens can readily be seen to be mapping analogous past candidates, elections, issues, and problems onto those currently on the ballot so as to reason about them. Consequently, analogical reasoning allows citizens to treat the majority rule mechanisms that proliferate in real politics as "deliberative outlets," which is to say, as catalysts of deliberation akin to the "creative outlets" that catalyze self-expression in the arts. While citizens may recognize majority rule mechanisms as catalysts of deliberation, many democratic theorists will hesitate to embrace this vision of the practical reality of deliberative politics. Isn't analogical reasoning too low in rigor to be placed at the heart of the deliberative ideal? I develop two arguments to explain the foundational role analogy plays in deliberation and to counter such critics. First, I draw on the explosion of research on analogical reasoning over the past two decades to show that it is far more rigorous and systematic than many suppose. Second, I argue that to the extent that citizen deliberation is concerned with rational planning, rather than just reasoning in general, analogical reasoning is logically superior. When we reason about what to do, we make plans that incorporate predictions about what is likely to ensue when a given course of action is selected. However, as soon as predictions enter into deliberation, its underlying logic changes as well. The reason for this change in logic is that as our probabilistic reasoning expands, the probability of its conclusions degenerates. Therefore, when assessing probabilities, we no longer should seek decisions derived from long, elegant chains of reasoning that connect our various options to generalities like values and principles. Instead, what we need is "short and sweet," or terse, humble lines of reasoning, which are more congruent with this form of deliberation. Thus, to the extent that democratic deliberation is involved in rational planning, it calls not for the elegant, deductive kind of reasoning idealized by proponents and critics of deliberative democracy alike. Instead, democratic deliberation calls for the "short and sweet," analogical kind of decision-making we associate with ordinary citizens already. After all, as research has shown, analogies are a much preferred and rigorous way by which even experts engage in probabilistic reasoning. By focusing on analogical reasoning, I therefore conclude that the practical reality of deliberative democracy should be recognized in ways that might ordinarily be dismissed. (shrink)

Representationalism (aka intentionalism) has been the most significant weapon in the late twentieth century defence of direct realism. However, although the representationalist objection to the Phenomenal Principle might provide an effective response to the arguments from illusion and hallucination, plausible representationalist theories of perception are, when fleshed-out, incompatible with metaphysical direct realism’s directness-claim. Indeed within cognitive science, direct perception is the avowedly-radical anti-representationalist heterodoxy. Drawing on both the philosophy and cognitive science, we develop a robust argument against representationalist direct realism (...) which does not rest on illusory cases or the notion of an object of immediate awareness. None of the representations so far posited, or in prospect, can supply the content required of a representation in a direct realist theory of veridical perception. (shrink)

‘Predictive Processing’ is an emerging paradigm in cognitive neuroscience that depicts the human mind as an uncertainty management system that constructs probabilistic predictions of sensory s...

The target article sketched and explored a mechanism (action-oriented predictive processing) most plausibly associated with core forms of cortical processing. In assessing the attractions and pitfalls of the proposal we should keep that element distinct from larger, though interlocking, issues concerning the nature of adaptive organization in general.

An influential thesis in contemporary philosophy of mind is that subjectivity is best conceived as inner awareness of qualia. has argued that this unique subjective awareness generates a paradox which resists empirical explanation. On account of this “paradox of subjective duality,” Levine concludes that the hardest part of the hard problem of consciousness is to explain how anything like a subjective point of view could arise in the world. Against this, I argue that the nature of subjective thought is not (...) correctly characterized as inner awareness, that a non-paradoxical approach to the first-person perspective is available, and that the problem about subjectivity should be distinguished from the perennial problem of qualia or phenomenal properties. (shrink)

One major fault line in foundational theories of cognition is between the so-called “representational” and “non-representational” theories. Is it possible to formulate an intermediate approach for a foundational theory of cognition by defining a conception of representation that may bridge the fault line? Such an account of representation, as well as an account of correspondence semantics, is offered here. The account extends previously developed agent-based pragmatic theories of semantic information, where meaning of an information state is defined by its interface (...) role, to a theory that accommodates a notion of representation and correspondence semantics. It is argued that the account can be used to develop an intermediate approach to cognition, by showing that the major sources of tension between “representational” and “non-representational” theories may be eased. (shrink)

In their target article, Hutto and Satne eloquently articulate the failings of most current attempts to naturalize mental content. Furthermore, we think they are correct in their insistence that the only way forward is by drawing a distinction between two kinds of intentionality, one of which is considerably weaker than—and should be deployed to explain—the propositional variety most philosophers take for granted. The problem is that their own rendering of this weaker form of intentionality—contentless intentionality—is too weak. What’s needed is (...) a species of intentionality distinct from both the industrial-strength version beloved by philosophers and the intentionality lite recommended by Hutto and Satne. We briefly motivate and sketch this alternative, and say a few words about the account of cognition that it spawns. (shrink)

Reading fiction for pleasurable is robustly correlated with improved cognitive attainment and other benefits. It is also in decline among young people in developed nations, in part because of competition from moving image fiction. We review existing research on the differences between reading/hearing verbal fiction and watching moving image fiction, as well as looking more broadly at research on image/text interactions and visual versus verbal processing. We conclude that verbal narrative generates more diverse responses than moving image narrative., We note (...) that reading and viewing narrative are different tasks, with different cognitive loads. Viewing moving image narrative mostly involves visual processing with some working memory engagement, whereas reading narrative involves verbal processing, visual imagery and personal memory (Xu et al 2005). but Attempts to compare the two by creating suggest that existing research is flawed by attempts to create equivalent stimuli and task demands face a number of challenges, and we discuss these difficulties in comparative approaches. We then investigate the possibility of identifying lower level processing mechanisms that might distinguish cognition of the two media, and propose a focus on internal scene construction and on working memory as foci for future research. Although many of the sources we draw on are focussed on English speaking participants in European or North American settings, we also cover material relating to speakers of Dutch, German, Hebrew and Japanese in their respective countries, and studies of a remote Turkish mountain community. (shrink)

David Haig propounds and illustrates the unity of a radically revised set of definitions of the family of terms at the heart of philosophy of cognitive science and mind: information, meaning, interpretation, text, choice, possibility, cause. This biological re-grounding of much-debated concepts yields a bounty of insights into the nature of meaning and life. An interpreter is a mechanism that uses information in choice. The capabilities of the interpreter couple an entropy of inputs to an entropy of outputs is dispelled (...) by observation. The second entropy is dispelled. I propose that an interpreter’s response to inputs meaning of the information for the interpreter. In this conceptual framework, the mechanisms of interpreters provide the much-debated link between Shannon information and semantics. (shrink)

Provided we agree about the thing, it is needless to dispute about the terms. —David Hume, A treatise of human nature, Book 1, section VIIMap-like representations are frequently invoked as an alternative type of representational vehicle to a language of thought. This view presupposes that map-systems and languages form legitimate natural kinds of cognitive representational systems. I argue that they do not, because the collections of features that might be taken as characteristic of maps or languages do not themselves provide (...) scientifically useful information above and beyond what the individual features provide. To bring this out, I sketch several allegedly distinctive features of maps, and show how they can easily be grafted onto a simple logical language, resulting in a hybrid “manguage.” The ease with which these linguistic and map-like properties can be integrated into a single representational system raises the question of what work broader categories like language and map are doing. I maintain that, as an empirical matter of fact, they serve no particular purpose in cognitive science, and thus do not characterize distinct kinds of cognitive architectures. (shrink)

This paper proposes a model for implementation of intrinsic natural language sentence meaning in a physical language understanding system, where 'intrinsic' is understood as 'independent of meaning ascription by system-external observers'. The proposal is that intrinsic meaning can be implemented as a point attractor in the state space of a nonlinear dynamical system with feedback which is generated by temporally sequenced inputs. It is motivated by John Searle's well known (Behavioral and Brain Sciences, 3: 417–57, 1980) critique of the then-standard (...) and currently still influential computational theory of mind (CTM), the essence of which was that CTM representations lack intrinsic meaning because that meaning is dependent on ascription by an observer_._ The proposed dynamical model comprises a collection of interacting artificial neural networks, and constitutes a radical simplification of the principle of compositional phrase structure which is at the heart of the current standard view of sentence semantics because it is computationally interpretable as a finite state machine. (shrink)

Arithmetical cognition is the result of enculturation. On a personal level of analysis, enculturation is a process of structured cultural learning that leads to the acquisition of evolutionarily recent, socio-culturally shaped arithmetical practices. On a sub-personal level, enculturation is realized by learning driven plasticity and learning driven bodily adaptability, which leads to the emergence of new neural circuitry and bodily action patterns. While learning driven plasticity in the case of arithmetical practices is not consistent with modularist theories of mental architecture, (...) it can be enriched by the theory of neural reuse. According to neural reuse, cerebral regions are reused to contribute to multiple neural circuits in functionally constrained ways throughout ontogeny. By hypothesis, learning driven plasticity is complemented by learning driven bodily adaptability, which suggests that there is an interesting functional relationship between finger gnosis, finger counting, and arithmetical practices. The emerging perspective on enculturated arithmetical cognition will be complemented by considerations on associated developmental and acquired disorders, namely developmental dyscalculia and acquired acalculia. The upshot is that we need to take the cerebral, extra-cerebral bodily, and socio-cultural dimensions of enculturation into account in order to arrive at a better understanding of the phylogenetic and ontogenetic conditions of arithmetical cognition. (shrink)

Representationalism is today the leading physicalist theory of the phenomenal character of perceptual experience. And Russellian representationalism, which identifies contents with extensions, is the leading iteration of that theory. If there exist phenomenally distinct experiences as of the impossible, then these would _prima facie_ serve as counterexamples to the theory. In order that they definitively serve as counterexamples, it needs to be that there is no plausible account of the experiences on which they decompose into constituent elements each of which (...) is unproblematic from the perspective of the theory. The contention of this paper is that the _stygian color experiences_, afterimage-experiences as of maximally dark, hued surfaces, of Churchland (Churchland, Philosophical Psychology 18:527–560, 2005) serve as counterexamples to Russellian representationalism. (shrink)