Search Results

This chapter discusses a vision of computational neuroscience, and also addresses what has happened since the publication of David Marr’s Vision. It argues that it is now time to reemphasize the links between levels of understanding, if progress in computational neuroscience is wanted. It is believed that computational neuroscience over the past thirty years can be described as mostly exploring each level of understanding independent of the others. Some of the past research trends in computational neuroscience are explained. This brief look at computational neuroscience in the last decades indicates that significant progress has been made at each of the levels of understanding—in a sense following Marr’s prescription—though the problem is far from being solved. It argues that neuroscience can help computational theory and even computer science.

In the first section of this chapter, I follow David’s early intellectual and scientific trajectory. I then describe his passions other than science: music and flying. The coherent thread running through these two sections and the later section, “The last challenge,” is of David Marr as a young man who never gave up in the face of any challenge, no matter how big, how hard. He was so determined, so intense and full of passion that he rose victorious above every challenge and this made him really happy. One could see it in his unique smile, a mixture of softness, warmth, and pride in victory, and in the indescribable, also unique, twinkle in his eyes.

The idea that a deductive competence theory is central to human cognition both has a long pedigree and is widely held by many leading figures in the psychology of reasoning. This chapter argues against this tradition in the psychology of reasoning. It claims that almost no everyday human reasoning can be characterized deductively, or has any significant deductive component. Although many theorists have argued that deduction is at the core of cognition, it is argued that it is at the periphery. The chapter begins by clarifying the meaning of the claim that human reasoning involves deduction by introducing Marr’s levels of description of a computational process.

Interpretation processes are necessary, whether one then applies probability theory or some other logic in reasoning from the resulting interpretations. In the case of suppression, understood in probabilistic terms, interpretation shows up as the necessity to change one's probabilities in ways not sanctioned by Bayesianism. This chapter argues that a computational level analysis, in the sense of Marr, must also incorporate the interpretation process, not only the reasoning, once the interpretation is chosen. This is not to deny the role of Bayesian probability in a characterization of the computational level. If a subject construes the task as involving uncertain conditionals, in the sense of positive probability of exceptions, principles like Bayesian conditionalization may well form part of the computational level. In this case competence theory is needed of how judgements of probabilities can change in non-Bayesian ways. This is one of the most interesting technical challenges issuing from the present analysis.

The chapter discusses David Marr's theory of vision, which likens the visual system to an information-processing system with three levels: the topmost “theory of computation,” the algorithmic level, and the implementation level. Marr's work, which is based on computational theory, has been assumed by many acolytes of this field of study to be “intentional.” This chapter aims to refute this assumption utilizing the broad tenets of computational methodology. It argues that, in utilizing the formal, mathematical paradigms of computational theory, Marr's theory is rendered essentially neutral, with no accompanying interpretations. Of course, interpretations based on underlying, external factors can also be valid, but are not essential, even if assigning intentional content can have practical uses for the researcher. The chapter ends the discourse by countering the proposed argument of “narrow content,” which posits that the primitives of Marr's vision theory are intentional, while their content is independent of external, causal factors.

The chapter revolves around Tyler Burge's interpretation of David Marr's theory of vision, and the chapter's arguments against such construal. Burge asserts that through evolution, our sensory systems have become adapted to our current environment, and can be assumed to be “successful” in that we are able to generate veridical perceptions of the same. According to Burge, this ascribes a “success-orientation” to Marr's theory. This chapter argues that Marr's own assumption of success in his work is merely a methodological dictate necessary to solve or understand the initial query and purpose of the study. In the same way that artificial intelligence is developed, one has to make intelligent choices on which types of human cognitive abilities should the project be patterned after. Thus, Marr's employment of this practice does not automatically warrant the assumption of veridical perception in his work.

This chapter looks at changes and developments in the study of object recognition during the past thirty years. It discusses the Marrian revolution attributed to David Marr, who took ideas and concepts from psychophysics with the aim of translating them into working computer models. Another major development during this period was the recognition and serious research on the neural basis of visual perception and object recognition.

In the study of neural networks, there have been very few attempts to examine in detail how a particular part of the brain might compute a particular function. This chapter examines one theory of learning and memory that does consider seriously the constraints imposed by the available computing machinery of the brain. This is the proposal, due to the late David Marr (1971), that the mammalian hippocampus acts as a temporary content-addressable memory store. This study remains a potential source of inspiration for those interested in the theory of the nervous system.

This first chapter introduces the main argument of the book and locates my position relative to existing approaches in the philosophy of psychology. The chapter includes a discussion of mental architecture and a discussion of earlier attempts to link visual experience with models from cognitive neuroscience. The conclusion of the main argument is that visual perception is an ongoing process of anticipation and fulfilment.

This book has tried to lay out in some detail an approach that treats visual perception primarily as an information-processing problem. This chapter presents a conversation between an imaginary skeptic and an imaginary defender of the information-processing point of view in order to answer the objections most commonly raised in this field. The dialogue is based on lunchtime conversations at the Salk Institute between Francis Crick, Tomaso Poggio, and David Marr. However, it does not follow those conversations very closely, and the imaginary objector is a combination of many real-life people.

This chapter attempts to traverse the difficulties that narrative machinery encounters in blackness-in/as-abjection by visiting a kind of text that generically aims to work with (and to work) psychic/body responses: pornographic writing. In Samuel Delany's The Mad Man (1994), a literary pornographic work, the protagonist John Marr is a black gay male character who feverishly seeks out the pleasure of sexual acts that involve some form of apparent humiliation or degradation. It is argued that what is represented in The Mad Man is something in the nature of a rough model of working with the legacies of a history of conquest and enslavement (which is to say, with blackness, with having-been-blackened) through the transformation provided by erotic/sexual fantasies. Delany imagines a position that takes on board race without having at the same time to take up its fellow traveler, so often mistaken for the thing itself, ego. Is it possible to have race without ego, without defensive postures, without boundaries to police and ramparts on which to stand watch? The character of John Marr tries to model for us this position. Delany imagines him living his black body in its collective, sociogenic dimension, in which the demand to self-protection of that seductive individual I is refused in favor of one's becoming immersed in, lost in what it is to be the race, precisely as to be black means to have-been-blackened, to have been rendered abject.

How do people recognize an object in different orientations? One theory is that the visual system describes the object relative to a reference frame centered on the object, resulting in a representation that is invariant across orientations. Chronometric data show that this is true only when an object can be identified uniquely by the arrangement of its parts along a single dimension. When an object can only be distinguished by an arrangement of its parts along more than one dimension, people mentally rotate it to a familiar orientation. This finding suggests that the human visual reference frame is tied to egocentric coordinates.

This chapter examines the influence of Nicholas Yakovlevich Marr, a historian and linguist from Georgia, on Soviet archaeology. It first provides an overview of Marr's educational background in linguistics and his rise as the recognised leader of Caucasian studies in Russia before turning to his scholarly work in archaeology. It then considers Marr's collections of archaeological finds from Armenia, along with his activities during and after the Russian Revolution of 1917. It also discusses Marr's research on the Japhetites and his ‘Japhetic theory’, which became ‘the new doctrine of language’.

Marr’s Motor Learning Theory (MLT) theory (Marr, 1969) represents a seminal example of how the structure/function relationship in a neural circuit can be conceptualized. It has laid the basis for cerebellar investigations for almost half a century. The foundations of MLT are that (1) the cerebellum granular layer generates a sparse representation of mossy fiber inputs that is then conveyed to Purkinje cells through the parallel fibers; and that (2) the connections between parallel fibers and Purkinje cells adjust their weights depending on teaching signals coming from the inferior olive through the climbing fibers. The MLT elegantly solved the problem of how the cerebellum might act as an adaptable controller and operate in several forms of behavior. Nonetheless, the MLT was based on connection statistics and lacked biological details about spatiotemporal dynamics and circuit geometry, so that it has been subsequently challenged by experimental findings related to neuronal and synaptic time-dependent properties, mechanisms and forms of long-term plasticity, and local connectivity rules. Interestingly, the fundamental intuition about the relationship between structure and function was supported by experimental findings on granular layer processing, Purkinje cell integration and parallel fiber plasticity. At present, the need to account for critical experimental observations has promoted the development of new biophysically-grounded models, which may lack the elegance of the initial MLT intuition but provide insight on how the neuronal circuit could operate dynamically during signal processing in the spatio-temporal domain.

In this chapter we outline and discuss the key aspects of Marr’s seminal work on a theory of neocortex, published 45 years ago in the Proceedings of the Royal Society, B. Compared to his later work on computational models of vision, where he was principally concerned with developing algorithms, it is interesting to see how carefully Marr maps his early computational approach onto a neural implementation. Inevitably Marr had to make many guesses about the specifics of the cortical circuits, because so many key principles of organization were still missing. Here we review how the concepts introduced by Marr have evolved, driven by a richer understanding the structure, function and development of cortical circuits and their computational functions.

Cognitive Science has come a long way – even the old Greeks have put forward intricate thoughts about what our minds can know and what seems inaccessible. This chapter addresses the historical development of cognitive science. The fundamental homunculus and qualia problems are introduced, and tentative answers from embodied cognitive science are sketched-out. A look at biology emphasizes that our bodies and brains are evolutionarily shaped, making us ready to develop our minds. Psychologically, minds develop and adapt given environmental feedback, striving to optimize behavior. Only by means of constructive, generative processes, however, can a behavior-oriented “understanding” of the environment develop. These “understanding”-oriented structures also make our mind language ready. Finally, in order to comprehend how the human mind actually comes into being, based on the considered principles, it is necessary to pursue functional, computational, and where possible, algorithmic and hardware perspectives in the sense of David Marr’s three levels of understanding.

Sander Gilman argued that the consensus of nineteenth century ethnography was that the Jews “were ‘black’ or, at least, ‘swarthy.’ ” In fact, there was no such consensus. Monogenist ethnographers had to show that European Jews were white in order to be able to show that black Jews had changed from white as a result of climate determinism. The increased racialization of Jews and blacks around the time of the American Civil War was accompanied by a renewal of the idea that the Jews and blacks had much in common, including color. Robert Knox was one of the first anthropologists to argue that the Jews were negroid. He was followed by many others, most of them polygenists. Other conflations of Jews and blacks included features such as smell, ears, eyes, hair, sexuality, etc. Jews were constructed as negroes, and at the same time blacks were constructed as Jews. They were both considered to be uniquely ugly. The ugliest Africans were the so-called Hottentots, who themselves had been deemed to be of Jewish origin. African tribes throughout the continent were constructed as Jews. The founder of political antisemitism, Wilhelm Marr, came to his hatred of Jews through a hatred of blacks in America. He thought Jews had negroid features and antecedents. From the time of the German Enlightenment Jews were inserted into the category of black slaves. For the influential racial theorist Houston Stewart Chamberlain, black blood flowed in the veins of Jews. The same goes for the important theorist Friedrich Ratzel. In the Americas, out-and-out anti-black racists were often equally anti-Jewish.

Andy Clark has defended a view of the body’s relationship to the mind that he calls the Larger Mechanism Story (LMS). At the same time, he has criticized an alternative account, the Special Contribution Story (SC). After first clarifying the commitments of these two stories, I argue that Clark’s favored LMS is of less psychological interest than SC, and that SC offers an interesting and viable research program.

This chapter begins with a discussion of the arguments and concepts of mathematician turned theoretical neurophysiologist, David Marr. It then covers the rejection of the Marrian approach; modules, isolable subsystems, and functional specialisation; the anatomical bases of modularity; information-processing models; the COGENT formalism; connectionist models; and more advanced symbolic architectures.