Jesse Lopes

The paper demonstrates that the phenomenology of logical experience, or Husserlian phenomenology, and the computational theory of mind, in the form of a Fodorian language of thought, necessarily constitute a single theoretical science. We first show that Husserl, like Fodor, argues that symbolic mental representations themselves must have syntactic (language-like) form as part of an unconscious causal mechanism of symbolic transformation, thereby anticipating Fodor’s application of Turing to cognitive psychology for a language of thought. We then argue, based on these facts, for a theoretically necessary transformation of the pure description of phenomenology into a complete nomological-explanatory theory with a computational (not necessarily biological or neurological) implementation as its causal basis.

Exploring the cognitively internalist roots of transcendental philosophy, this book sheds new light on the predictive processing theory of perception and other externalist theories. The rise of neo-empiricist approaches employing associationism and iconic representations—together with appeals to causality within perception by these same approaches—have inadvertently revived the transcendental arguments of Immanuel Kant against David Hume’s imagistic associationism. While they invoke causality, they fail to provide a viable account of how the representation of objective distal causes, as opposed to statistical half-objects, could be generated within perception. This book demonstrates the ongoing relevance of transcendental arguments by building on anti-reductive accounts of causality in relation to associative (Bayesian & Deep Learning) networks, arguing that perception must contain an a priori synthetic (Kantian) rule of causality (e.g., if X, then Y). Otherwise perception is impossible. This necessarily entails either a Language of Thought in perception thesis or a non-iconic sui generis schema of causality thesis. In either case, it demonstrates the inadequacy of empiricist, associative-networks, as well as externalist, anti-individualist theories of perception and instead supports a transcendental-computational theory of perception. The cognitive psychology of Jerry Fodor and methodological apparatus of Edmund Husserl are combined to show that Kant's transcendental argument for cognition of causality against Humean associationism must finally be assimilated by cognitive science. The epistemological foundation for this assimilation is found in transcendental phenomenology, which is shown to support a computational theory of mind within perception.

The evidence shows that productivity is fundamental to cognition and phenomenology. Productivity requires an expandable memory, a necessary condition on Turing-completeness (or universal computation). Neural nets and dynamic systems are thought to be necessarily incapable of Turing-completeness due to structural instability, so that they are impossible to build or find in nature. Therefore if phenomenology is evidence for anything in nature, outside of the "epoché", it is evidence, first and foremost, for the language of thought. That is because the language of thought can trivially recruit memory as needed for productivity, which is an eidetic phenomenological law. Neuroscience is starting to gather evidence for the possibility that a productive symbolic system exists, not between associated neurons, but inside of them, via molecular machinery. This machinery and its symbolic operations have the potential to explain transcendental phenomenology, as it goes beyond associationism, empiricism and their resultant psychologism. Just as Mendel explained Darwin’s descent taxonomics, so emerging neuroscience might finally explain Husserl’s cognitive taxonomics, setting the stage for a new synthesis in cognitive science.

Hume once argued the basic science to be not physics but “the science of man” and the foundation of this science to be the empiricist mechanism of association governed by the law of similarity in appearance—now more popular than ever in the form of artificial neural networks. I update Hume’s picture by showing phenomenology to be centrally concerned with providing a unifying basis for all the sciences (including physics) by going beyond the psychology of associationism (passive synthesis) to reveal phenomena that are irreducibly syntactic (not associative) in structure. I therefore argue that the language of thought (LOT) is the necessary mechanism at the basis of these descriptive phenomena. I conclude by sketching a new picture of all the sciences unified by LOT based on Husserl’s opposition to Galilean “physicomathematical” science vis-à-vis the life-world (Lebenswelt).

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).

Yoshimi has attempted to defuse my argument concerning the identification of network abstraction with empiricist abstraction - thus entailing psychologism - by claiming that the argument does not generalize from the example of simple feed-forward networks. I show that the particular details of networks are logically irrelevant to the nature of the abstractive process they employ. This is ultimately because deep artificial neural networks (ANNs) and dynamical systems theory applied to the mind (DST) are both associationisms - that is, empiricist theories that derive the principles of thought from the causal history of the organism/system. On this basis, I put forward a new aspect of the old argument by noting that ANNs & DST are the causal bases of the phenomena of passive synthesis, whereas the language of thought hypothesis (LOT) and the symbolic computational theory of mind (CTM) are the causal bases of the phenomena of active synthesis. If the phenomena of active synthesis are not distinct in kind from and are thus reducible to those of passive synthesis, psychologism results. Yoshimi’s program, insofar as it denies this fundamental phenomenological distinction, is revealed to be the true anti-pluralist program, by essentially denying the causal efficacy of the mechanistic foundations of active synthesis by referring phenomenology exclusively to associationism for its causal foundation.

This essay examines the possibility that phenomenological laws might be implemented by a computational mechanism by carefully analyzing key passages from the Prolegomena to Pure Logic. Part I examines the famous Denkmaschine passage as evidence for the view that intuitions of evidence are causally produced by computational means. Part II connects the less famous criticism of Avenarius & Mach on thought-economy with Husserl's 1891 essay 'On the Logic of Signs (Semiotic).' Husserl is shown to reaffirm his earlier opposition to associationist (Humean) and adaptationist (Darwinian) explanations of our thought-machine on the ground that they cannot reconstruct the notion of truth and its syntactic mental preservation in symbolic thought-trains. Part III reveals Husserl's interesting commitment to the idea that descriptive sciences necessarily transform into explanatory sciences on pain of contradicting the essence of science as a rational ideal. Since explanation in relation to phenomenology means causal explanation, and causal explanation in the form of the Denkmaschine accounts for phenomenological intuition, it is inferred that the rationally compelling ideal of explanatory science requires a computationalist reading of the subsequent logical investigations.

We argue in this article that Marx’s scientific method coupled with his analysis of the phenomenological consciousness of agents trapped within the capitalist mode of production provides a sufficient solution to the transformation problem. That is, Marx needs no amending – mathematical, philosophical, or otherwise – and the tools he uses to demonstrate and resolve the problem – science and phenomenology – were already clearly spelled out in his texts. Critics of Marx either fail to understand his scientific method, or are themselves trapped within a non-scientific capitalist phenomenology. Similarly, Marxists that mathematically resolve the transformation problem fail to realise that Marx’s scientific analysis alone demonstrates that a mathematical solution to the transformation problem is a misapprehension of the relation between Marx’s abstract theory and concrete phenomena. Consequently, we also criticise the monetary theorists who try to dismiss the problem as pointless by claiming that Marx was not a pre-monetary theorist.

Hubert Dreyfus once noted that it would be difficult to ascertain whether Edmund Husserl had a computational theory of mind. I provide evidence that he had one. Both Steven Pinker and Steven Horst think that the computational theory of mind must have two components: a representational-symbolic component and a causal component. Bearing this in mind, we proceed to a close-reading of the sections of “On the Logic of Signs” wherein Husserl presents, if I’m correct, his computational theory of mind embedded in a language of thought. My argument goes like this: the computational theory of mind is the idea, following Haugeland, that the mind comes prepackaged as, or is endogenously constrained to be, an automatic formal system; this explains, according to Husserl, why automatic trains of thought without logical intent resemble arguments exhibiting deductive structure with logical intent. In general, an automatic formal system yields true results provided that the syntactic symbols with which they compute are univocal and are semantically evaluable, and the mechanized inferences they perform are valid and preserve truth. These two conditions describe a computational cognitive process: the first condition connects representations to syntax, and the second condition uses the syntax, in inauthentic judging, to arrive at true conclusions through blind causality. Now, in point of textual fact, these are the conditions which Husserl attributes to our “natural psychological mechanism of symbolic inference” which typically yields true results. Since a formal system attributed to the “internal structure” of the mind, and guided by blind causality, just is the computational theory of mind, it follows, I think, that Husserl had a computational theory of mind. This computational theory is, moreover, embedded in a language of thought, since Husserl attributes a language-like form to our thoughts so that they may be mechanically processed. I conclude with a discussion of my results.

In current debates concerning atheism, two positions are considered possible: naturalistic atheism or anti-naturalistic theism. Anti-naturalistic theism is motivated by the failure of naturalism to explain the fundamental nature of reality. We, however, endorse anti-naturalistic atheism by reviving the ‘anthropomorphic critique’, arguing that theism misattributes human traits to the deity. Anti-naturalistic atheism is better suited to refute theists, since it undercuts their appeal to science's inadequacies. We trace the anthropomorphic critique from Hume's Dialogues, through Kant's epistemology, and conclude with its reception in Feuerbach. The anthropomorphic critique is an epistemological – not metaphysical – thesis, hence, it is agnostic about the fundamental nature of reality. Yet it convincingly shows that theism is not a tenable position. In essence, we aim to decouple atheism from naturalism, on the basis of a salient critique in the history of modern philosophy, in order to ameliorate current debates between atheists and theists.Export citation.

Psychologism is defined as “the doctrine that the laws of mathematics and logic can be reduced to or depend on the laws governing thinking” (Moran & Cohen, 2012 266). And for Husserl, the laws of logic include the laws of meaning: “logic evidently is the science of meanings as such [Wissenschaft von Bedeutungen als solchen]” (Husserl ( 1975 ) 98/2001 225). I argue that, since it is sufficient for a theory to be psychologistic if the empiricistic theory of abstraction is employed, it follows that neural networks are psychologistic insofar as they use this theory of abstraction, which I demonstrate is the case (Husserl ( 1975 ) 191/2001 120). It’s sufficient for psychologism because, according to Husserl, the theory in question reduces one’s phenomenological ability to intend types (or universals) to one’s past history of intending tokens (or particulars), usually amalgamated in some fashion (classically via associations; recently via autoencoders) (ibid; Kelleher, 2019 ). Similarly dynamical systems theory entails psychologism. For dynamical systems theory ties content to the temporal evolution of a system, which, according to Husserl, violates the fact that intentionality toward validities and objectivities does not pertain to "particular temporal experience[s]" (Husserl ( 1975 ) 194/2001 121). It follows that neither the species (neural networks), nor the genus (dynamical systems), can avoid psychologism and intend objects "in specie" (ibid). After critiquing these two approaches, I proceed to give an account based on the essentialist school of cognitive psychology of how we may intend objects "in specie" while avoiding the empiricistic theory of abstraction (Keil, 1989, Carey, 2009, Marcus & Davis, 2019 ). Such an account preserves the type-token distinction without psychologistic reduction to the temporal evolution of a dynamical system (Hinzen, 2006 ). This opens the way toward a truly unifying account of Husserlian phenomenology in league with cognitive science that avoids Yoshimi's ( 2016 ) and neurophenomenology's psychologistic foundation (herein demonstrated) and builds upon Sokolowski's ( 2003 ) syntactic account of Husserlian phenomenology.