01_Week 1 Notes (Foundations).pdf

September 6, 2024 Introduction Cognition: study of mental representations and processes involved in the acquisition, storage, retrieval, and use of knowledge enabling infinite tasks (attention, perception, speaking, reading, interpre...

September 6, 2024 Introduction Cognition: study of mental representations and processes involved in the acquisition, storage, retrieval, and use of knowledge enabling infinite tasks (attention, perception, speaking, reading, interpretation, etc.) You cannot infer the contents of the mind (thoughts, motivations, feelings, etc.) You can infer the mechanisms used Walking = navigational processes Listening = processing own language and tuning out foreign ones Motor and sensory controls Perception of smells Decision making **non-observable behaviours Knowledge: what the brain holds to enable our essentially human behaviours of encoding memories, recognizing familiar faces, and speaking sentences no one ever heard before Explicit: declarative memory (semantic/facts, episodic/autobiographical) Implicit: non-declarative memory (conditioning, sensory memory, habits and skills, procedural) Representation: how we encode information in the brain—and processes, that is, how we use that information in a variety of activities ranging from seeing colours and recognizing objects to understanding sentences, formulating complex thoughts, and making decisions “intelligent” activity (conscious or not) Retrieval of the information in a book = processes Conscious Behaviour is determined by intentions (explicit knowledge) Every book in a library = representations Unconscious/covert (implicit knowledge) Propositions: thoughts about states and events out there in the world lines of codes, complex structuring of concepts, translated into a the creating of a thought Mechanisms: akin to what computers do when they perform certain procedures. Computers follow rules (algorithms that run “behind the scenes”) to perform their functions i.e. 3D from 2D rule: if lines meet at a given point, it interprets as an a 3D object i.e. Grouping by similarity rule: interpret similar objects as having the same orientation in 3D  September 6, 2024 Noam Chomsky’s proposal: much of our linguistic capacities are part of our biological endowment, developed in the species rather recently, not as a gradual evolution of our ancestors’ communication systems the most important hypothesis about the nature and evolution of our cognitive systems Perception & Cognition VISUAL ILLUSIONS knowing that visual illusions are in fact illusions does not change what you perceive visual system needs to rely on what it knows about potentially incoming objects so that your eyes can blink fast often a conflict between what we perceive and what we know to be true about the world Perception: the earliest form of information processing by the organism information that is being captured by our senses how this information is organized early on in the brain Cognition: all processes that are not perceptual including the last stage of “thinking” or “cognizing” about what is perceived and about what we bring to fore from our past experiences, our beliefs and so forth constant dispute between sources of knowledge: 1. what we know—or believe—about the world 2. the information or knowledge that our perceptual systems keep feeding us with INFORMATION PROCESSING Two major stages: 1. Perception: early and fast, automatic, and independent of what you believe 2. Cognition or “higher cognition”: relatively slow, not necessarily automatic, and is closer to the common-sense idea of “thinking” EXPLICIT & IMPLICIT KNOWLEDGE Explicit: common-sense idea of knowledge Driven by own intentions Declarative Semantic (facts of life) Episodic (autobiographical)  September 6, 2024 Implicit: the codes and rules that your mind relies on to perform the most basic functions mental mechanisms that enable your comprehension and production of sentences, the recognition of objects, faces, and scenes procedures involved in the way you acquire information, how you go about storing this information how you go about retrieving and using it Automatic tasks, unconscious Not available for introspection Non-declarative Unobservable behaviours BEHAVIOUR DETERMINED BY KNOWLEDGE Behaviour: manifestations of internal dispositions relying on knowledge covert, cognitive behaviour: what your brain does “behind the scenes” while you perform overt behaviours, even when you appear to be doing nothing Explicit knowledge allows to determine your actions at will: you can make decisions, justify/explain about your actions by knowing goals, causes, consequences, etc. by having intentions, beliefs, and expectations Implicit knowledge seems to be a closed system cognitive processes based on implicit knowledge are autonomous and non- declarative STUDYING KNOWLEDGE (IN THE BRAIN) 1. Physics or chemistry: we could look into neurotransmitters and what they do in different processes, as we could also look into how neurons are organized into networks and how they fire as cognitive events unfold the most natural way of looking into the “particles” of cognition assuming all cognitive events correspond to neuronal events we should be looking into neurons, their structure, physiological functions, and synaptic processes for our explanation of cognitive phenomena Limitations although physiological processes ultimately underlie cognitive ones, we have to consider cognitive representations and processes to be autonomous as far as explanations go understanding the neurophysiological properties of neurons and how they are organized into layers and networks does not give us direct knowledge about what neurons represent nor how they come to represent knowledge  September 6, 2024 2. Representations: Cognitive processes run on representations and it is because we encode properties of the world in our brains that we can look into those to explain cognitive processes how representations (conceptual representations in particular) might be encoded and used in cognitive processes Concepts: the units of meaning; they enter into many processes involved in higher cognition in order to understand cognition, we appeal to their “particles” (representations) and how they are used in processes, all of which cannot be understood directly by looking into how neurons work at the physiological level LEVELS OF ANALYSIS: COMPUTATION, ALGORITHMS, AND THE BRAIN Computational Theory - Marr Goal of the computation: what it is supposed to be about/to solve Define our object of study, as precise as possible, and devise a general functional architecture for that particular process We need to know the nature of the task, and put it in computational terms between input and output functions Involves translating a general problem int information processing problem (computation) identifying component process Representation & Algorithm Level Representations that the system manipulates and how the system manipulates them Representations refer to the nature of the input and output of a given process, and how these input and output representations are actually encoded in the brain the rules for manipulating representations output would be the ultimate representation achieved Algorithms are the actual principles by which the transformations are realized the rules for manipulating representations Implementation level The physical installation: the hardware where processes are realized (functional anatomy) How processes are realized (neurophysiology); How representations and processes are mapped onto neurological processes fit the functions (computations, representations, algorithms) employed by a given process to the neurological apparatus of the human brain algorithms must interface with the machine (the brain) for a full account of the process to be valid we cannot devise a program (the set of algorithms) for a device that can’t run it  September 6, 2024 These three levels of analysis are intertwined: They constrain each other but it is at the level of implementation where, to a large extent, particular empirical proposals for cognitive processes meet validity