COGS 1 HW.docx

Week 1 Day 2 HW Carandini // From circuits to behavior: a bridge too far? Gap between neural circuits and behavior is too wide, need to understand intermediate neural computations Metaphor → how would people try to understand a laptop computer with the knowledge and tools of a century ago? But...

Week 1 Day 2 HW Carandini // From circuits to behavior: a bridge too far? Gap between neural circuits and behavior is too wide, need to understand intermediate neural computations Metaphor → how would people try to understand a laptop computer with the knowledge and tools of a century ago? But each generation tends to compare the brain to a complex technology of their time: a loom, a telephone exchange, a chemical plant, etc General rule in science to seek an appropriate level of description Linear filtering vs divisive normalization Linear filtering: widespread computation in sensory systems. Neurons operate on sensory inputs by weighted summation in linear receptive fields Divisive normalization: neuronal responses are divided by a common factor, the summed activity of a pool of neurons. Odors, deployment of visual attention, etc Marr’s approach → “any particular biological neuron or network should be thought of as just one implementation of a more general computational algorithm” What methods to establish funcitonal connectivity between neurons? Connectome: full diagram of the circuits of the brain A map may not be helpful with connection strength info Simulome: effort to simulate brain circuits in all their complexity. “Understanding of the way nervous systems compute will be very closely dependent on understanding the full details of their structure” Crane // Computers and Thought A computer is a causal mechanism which contains representations Distinguish between 2 questions 1) Can a computer think? Can anything think simply by being a computer? 2) Is the human mind a computer? Are any actual mental states and processes computational? A computer is a device that processes representations in a systematic way Algorithm: a method for finding the value of a function Turing Machines What determines whether a certain method of calculation is adequate for the task at hand? → Turing asked what is the simplest possible device that could perform any computation whatsoever, no matter how complicated? → Turing machine Turing machines are an abstract, theoretical specification of a possible machine. The point is to illustrate some very general properties of algorithms and computations Machine states, reading tape symbols To perform all the operations that Turing machines can perform, we don’t need a separate machine for each operation. We need only one machine that is capable of mimicking every other machine. A universal Turing machine Any piece of text can be coded in terms of numbers → any algorithm that can be written in English can be run on a Turing machine (binary) Digital, not analog Some systems can be modelled on a computer when a theory of that system is computable. A system performs computations, however, when it processes representations by using an effective procedure Automatic Algorithms: Algorithms that are some way “embodied” in the physical structure of the computer Breaking down black boxes, input/output machines (mousetrap example, starting with live mouse/dead mouse input/output) Functional analysis Thinking Computers? The fact that the notion of representation can be used to define both thought and computers does not imply anything about whether computers can think It’s foolish to argue that computers can think simply because representation can be employed in defining thought and computers Leibniz’s universal character: a mathematically precise, unambiguous language into which ideas could be translated, and by means of which the solutions to intellectual disputes could be resolved by “calculation” Cobb: Why your brain is not a computer Olaf Sporns: “neuroscience still largely lacks organizing principles or a theoretical framework for converting brain data into fundamental knowledge and understanding” Big data is not knowledge By viewing the brain as a computer that passively responds to inputs and processes data, we forget that it is an active organ. Part of a body that is intervening in the world Widespread assumption that nervous systems can only be explained as emergent properties, unpredictable Overlooks different kinds of emergence, weak and strong Brains and minds are identical. Software and hardware are separate on computers, but in brains they’re the same. “Wetware.” what is happening and where it’s happening are completely intertwined The mind is NOT an immaterial thing that can be uploaded to a computer “Endlessly debating the truth or falsity of a metaphor like ‘the brain is a computer’ is a waste of time. The relationship proposed is metaphorical, and it is ordering us to do something, not trying to tell us the truth” Analyzing a retro arcade chip with neuroscientific techniques → “the problem is that they would not understand it given the approaches they are currently taking” Week 1 Day 3 Balint’s Syndrome: A Disorder of Visual Cognition (Rafal) RM suffered two strokes Right parietal lobe, right cerebellum → developed and recovered from transient left hemiparesis, left hemispatial neglect Left parietal lobe → functionally blind Normal visual acuity, but severe spatial disorientation. Easily lost When shown two objects he often saw only one Nearly symmetrical lesions in each parieto-occipital region Balint’s syndrome: visual acuity is preserved and patients can recognize objects placed directly in front of them, but they are unable to interact with/make sense of their visual environment. Lost in space 1) simultagnosia: constriction, not of the visual field, but of visual attention to one object at a time 2) spatial disorientation: a loss of all spatial reference and memory that leaves the patients lost in the world and unable to look at objects (“psychic paralysis of gaze”) or to reach for them (“optic ataxia”) Anatomy and Etiology of Balint’s Syndrome Bilateral lesions of the parieto-occipital junction Symptom Complex of Balint’s Syndrome Inability to perceive more than one object at a time (optic ataxia / simultagnosia) Ex. clinical test uses overlapping figures, observe the degree to which local detail can capture the patient’s attention and exclude all other objects “Discerned objects around him with difficulty, that they flashed before his eyes and sometimes disappeared from his field of vision” Spatial disorientation No recollection of spatial relationships of places he knew well before injury, no ability to learn new routes Impaired oculomotor behavior Striking disturbances of fixation, saccade initiation/accuracy, and smooth-pursuit eye movements Lack of space was specific to vision. Patient was able to localize sounds Impaired depth perception Consequence of loss of topographic perception Loss of blinking in response to a visual threat, inability to recognize the nearness of the threatening object Nosological Consideration More commonly the case that lesions will cause more extensive damage to the occipital, parietal, and temporal lobes A given patient can have optic ataxia, spatial disorientation, or simultagnosia without other elements of Balint’s May result from very different mechanisms than those in Balint’s Implications of Balint’s for understanding visual cognition 1) attention makes a selection from object-based representations of space 2) independent neural mechanisms that operate in parallel orient attention within objects and between objects 3) candidates objects on which attention operates are generated preattentively by early vision in the absence of explicit awareness 4) attention is involved in affording explicit access to the spatial representations needed for goal-directed action, binding features of an object Object- and space-based attention Does visual attention act by selecting locations or objects? → Balint’s demonstrates object-based model Neural representations of objects in space Humphreys posited that attention operates on spatial representations determined by objects, there are separate mechanisms operating inparallel for shifting attention within and between objects Visual processing outside of conscious awareness Preattentive processes Preattentive representation of space. Simultagnosia may also occur for nonspatial information, such as shifting between words and pictures Parietal damage does NOT eliminate representations of spatial information, but rather prevents explicit access to this information Preattentive grouping of features, alignment of principal axis Attention, spatial representation, feature integration: gluing the world together Conclusions Objects appear and disappear, their features become jumbled together Insight into neural basis of visual attention and perception, how they operate together normally to provide coherent perceptual experience/efficient goal-directed behavior Unsolved issues Are simultagnosia and spatial disorientation independent symptoms? Speed (Oliver Sacks) Speed of thought, perception of time Apparent slowing of time in emergencies may come from the power of intense attention to reduce the duration of individual frames With disorders of time scale, there seems no limit to the degree of speed or slowness Week 2 Day 2 HW: Evolution and the Nature vs. Nurture Debate “The Brain is a Bag of Tricks” Hunt-together ancestors, highly social “Bag of tricks that just happens to work, to allow us to survive” Reasonable Person The reasonable person will weigh the following factors before acting 1) the foreseeable risk of harm his actions create vs. their utility 2) the extent of the risk so created 3) the likelihood such risk will actually cause harm to others The brain doesn’t have that “reason” hard wired in Ex: corpus callosum cuts lead to split brain phenomenon Left side controls language brain/mind makes do with whatever it has Memory is also a Bag of Tricks Brenda Milner, amnesia studies (patient HM!) HM could retain skills, but not explicit memories Emotion is a Bag of Tricks Emotional parts of the brain Drive us to act, but not entirely uncontrolled. Like a horse Ex. anxiety and low mood in controlled students Anxious students have enlarged amygdala The upper fifth is most anxious The larger the emotional systems, the smaller the control systems Being Aware Isn’t Always Being in Control Ex. effortless driving The brain learns, adapts, habituates things that it doesn’t have to spend deliberate time on “Controlled Processing” vs. Habits Prefrontal cortex vs. striatum Face Perception in Monkeys Reared with No Exposure to Faces Infant monkeys reared with no face exposure for 6-24 months With facial exposure… Monkeys showed preference for human and monkey faces in photographs Experience-independent ability for face processing, apparent sensitive period during which a broad but flexible face prototype develops into a concrete one for efficient processing of familiar faces Big question: is face processing a special perceptual process? It is organized as such at birth? Face-deprivation period was 6 months for four monkeys, 12 months for another four monkeys, and 24 months for a final two monkeys Results: infant monkeys showed remarkable face-processing abilities even though they were never exposed to faces or face-like stimuli Control monkeys (no deprivation) could not discriminate between novel and familiar faces from the human face set, although they could from the monkey face set Preference for facial features to which they were exposed Socially meaningful stimuli may have prepotent and activating properties for socially naive infant monkeys The Fallacy of Biological Determinism (Lewontin) Inequality within societies and between nations/continents Some argue inequality is biologically determined “The assortment of persons into occupational roles simply is not ‘fair’ in any absolute sense. The best we can hope for is that true merit, given equality of opportunity, act as a basis for the natural assorting process” Idea that biological differences in ability between individuals must inevitably be turned into a hierarchical system of differential status “The members of human societies sometimes cooperate closely in insectan fashion, but more frequently they compete for the limited resources allocated to their role-sector” Aristocrats claiming superior “blood”, then anatomy, then DNA as basis of biological distinction “The gathering and presentation of the ‘facts’ from which the heritability of IQ has been calculated have been so scandalously bad as to constitute a veritable Watergate of human behavioral genetics” The main conceptual error in the biological determinism argument stems from an incorrect understanding of the relation of gene, environment, and organism It is totally incorrect to equate “inherited” or “genetic” with “unchangeable” Inherited disorders can be treated and corrected as easily (or with as much difficulty) as those arising from birth traumas, accidents or environmental insults “Genetic potentional,” incorrect assumption that the genes set the limit or the maximum of a character, while the environment determines how much of that “limit” will be realized/fulfilled No one has ever been able to characterize human norms of reaction for any behavioral trait (or for any human trait at all) It is not possible to ascribe a particular fraction of an individual’s phenotype to genes and the rest to environment There is neither a logical nor empirical relationship between the heritability of a trait within a population and the causes of differences between populations Week 2 Day 3 HW: How you see the world Marr: The Philosophy and the Approach 1.1 Background Gestalt school of psychology, concerned with describing the qualities of wholes by using terms like solidarity and distinctness Trying to formulate the “laws” that governed the creation of these wholes Binocular vision studies (digital electronic computers made it possible) Stereovision → independent modules of perception Adaptation and threshold detection studies → existence of independent, spatial-frequency-tuned channels in the early stage of our perceptual apparatus Character of a sensation depends on which fiber carried the message, not how the fiber was stimulated Led to the view that the peripheral nerve fibers could be thought of as a simple mapping supplying the sensorium with a copy of the physical events at the body surface Single neuron recording → receptive fields Barlow’s study of ganglion cells in frog retina Selectivity of retinal neurons → “bug detectors” perform a primitive but vitally important form of recognition The activities of neurons are thought processes → a revolution stemming from physiological work. The activity of each single neuron may play a significant role in perception Theory of the cerebellar cortex → simple and regular cortical structure is interpreted as a simple but powerful memorizing device for learning motor skills One could hope to understand the cortical structure in functional terms Even if the theory was correct, it did not much enlighten one about the motor system If we really had all these answers, we should be able to program them on a computer Three types of approach were taken to try and tackle edge detection 1) unashamedly empirical → take some new edge detection trick and run it on images, observe the result. Never accompanied by any serious assessment of how well the different algorithms performed 2) try for depth of analysis by restricting the scope to a world of single, illuminated, matte white toy blocks set against a black background. Binford-Horn line finder. Only had to do with straight edges 3) Horn’s analysis of shape from shading → analysis of the problem as an information-processing task 1.2 Understanding Complex Information-Processing Systems Need to contemplate different kinds of explanations at different levels of description that are linked, at least in principle, into a cohesive whole Representation and Description Representation: a formal system for making explicit certain entities or types of information Description: the result of using a representation to describe a given entity Process Processes revolving around a fundamental theory. Constraints Choose your level of analysis 1) computational theory, abstract mapping of one kind of info to another 2) representation for the input/output, algorithm by which the transformation its accomplished How can this computational theory by implemented? How is the input and output represented, and what is the algorithm for that transformation? 3) physical structure of the device, how can the representation and algorithm be realized physically? Importance of Computational Theory Not as empirically accessible, but nonetheless critical “Trying to understand perception by studying only neurons is like trying to understand bird flight by studying only feathers: it just cannot be done” Failure to recognize the theoretical distinction between what and how also greatly hampers communication between fields of AI and linguistics Intro to Vision Visual system constructs mental representation of the world around us Anatomy of the visual system Cornea: transparent covering over the eye Pupil: small opening in the eye through which light passes Iris: colored portion of the eye, has muscles that control size of the pupil Lens: curved, transparent structure that serves to provide additional focus Retina: light-sensitive lining of the eye Fovea: part of the retina, contains densely packed specialized photoreceptors (cones, perceive color) Rods: retinal periphery, work well in low light conditions and detect motion Optic nerve: carries visual info from the retina to the brain Blind spot: where the optic nerve joins the retina. No photoreceptors there Optic chiasm: x-shaped structure that sits just below the cerebral cortex at the front of the brain Amplitude and Wavelength Amplitude: the height of a wave from peak to trough Wavelength: the length of a wave from one peak to the next Related to frequency, number of waves that pass a given point in a given time period Electromagnetic spectrum Color Vision Trichromatic theory of color vision: all colors in our visual spectrum produced by combining red, green, and blue. Three types of cones, one for each color Opponent-process theory: color is coded in opponent pairs of black-white, yellow-blue, and green-red Afterimage: continuation of a visual sensation after removal of thes stimulus Trichromatic and opponent-process theories are not mutually exclusive Depth perception Binocular cues from binocular disparity Monocular cues (ex. Linear perspective) Integration with Other Modalities Vestibulo-ocular reflex: when you move your head and reflexively move your eyes in the opposite direction to retain focus on an object Synesthesia Stereo Sue: why two eyes are better than one (Oliver Sacks) Stereographs and the illusion of depth Effect produced by two slightly different images that get perceptually superimposed by each eye Loss of binocular vision = loss of depth perception But how do some stereo-blind people achieve feats of visuomotor coordination? One cannot convey to the stereo-blind what stereopsis is like Stereoscopy is highly conserved in predators depsites its costs (sacrifice of panoramic vision, need for special neural/muscular mechanisms, complex visual system in brain, etc) Perceptual Scotomas: A Functional Account of Motion-Induced Blindness Motion-induced blindness: when salient objects in full view repeatedly fluctuate into and out of conscious awareness when superimposed onto certain global moving patterns It may be a functional product of the visual system’s attempt to separate distal stimuli from artifacts of damage to the visual system itself Happens even when observers are fully knowledgeable about the relevant movements Dependence on additional types of lower-level visual properties Ex. MIB is enhanced when the target is placed stereoscopically behind the mask Explaining Motion-Induced Blindness Attentional competition: suggestion that MIB may reflect a disruption of the attentional competition that normally determines what observers are and are not aware of Interhemispheric rivalry: idea that interhemispheric competition may contribute to MIB like it does to binocular rivalry Boundary adaptation: suggestion that MIB is analogous to the perceptual “filling in” mechanism Surface completion: suggestion that there’s a link to visual surface processing. Perhaps the mask elements are integrated into a single visual surface, which is then taken to occlude the static target “underneath” A Perceptual-Scotoma Account Exploring the idea that MIB is a product of the visual system’s attempt to separate distal stimuli from artifacts of damage to the visual system itself Target stimulus may get filled in by the background. Artificial blind spot even though visual acuity is fine Explaining why MIB occurs rather than how it occurs Experiment 1: target-fixation motion congruence MIB can persist through slow movements of the target Typically eliminated by saccadic movements Testing the hypothesis that targets moving relative to the retina, compared with those moving in a retinal stable manner, should be less likely to undergo MIB. scotomas themselves are retinal stable Finding → relative visual motion may serve as a cue for the visual system to rule out a “scotoma” interpretation of the target, thus eradicating MIB Experiment 2: Filling in During MIB Predicting that when the target disappears, any surrounding texture should fill in that region of the visual field Finding → like visual scotomas, MIB involves not only a perceptual disappearance, but also an active filling-in process Experiment 3: MIB without Motion? The perceptual-scotoma account does not require motion, but requires only sustained global visual changes → predicting that MIB should occur even when such changes are implemented via cyclic changes in global luminance Finding → motion may not be required for “motion” induced blindness as long as a sustained global change forces the visual system to question the interpretation of the target as a distal stimulus Unconscious inferences could help give rise to MIB Week 3 Day 1: Attention and Memory Inattentional Blindness while Walking/Talking on a Cell Phone (Unicycling Clown) Cell phone use can cause inattentional blindness even during a simple activity that doesn’t need many cognitive resources Fougnie and Marois argued that inattentional blindness is more likely when the divided attention task involves processing by working memory rather than simply maintaining information Conversations involve demanding cognitive manipulations → often induce inattentional blindness Students argue that years of cell phone conversations and of driving have made these tasks automatic so that attentional capacity is not unduly overloaded Study examines whether behaviors get impacted by divided attention Study 1: observing people walk with a cell phone, music player, in pairs(?) etc Measuring time it took to cross the square, number of direction changes, weaving, trips/stumbles, collisions or near-collisions, if individual acknowledged other by people People in driving simulators (unrelated to present study) found cell phone conversations more distracting than conversations with someone in the car Study 2: interviewing random people w/ a clown nearby (???) Asking if 1) people saw anything unusual. If not, they were then asked directly about the unicycling clown People with music players noticed it, those actively looking at phones didn’t Tasks even as simple as walking can be disrupted by cell phone conversations A gender- and sexual orientation-dependent spatial attentional effect of invisible images Invisible erotic information can either attract or repel observers’ spatial attention depending on gender and sexual orientation Interrocular suppression to render stimuli invisible If the suppressed images exert a location-specific effect on the attentional system, they could potentially act as attentional cues Posner cueing paradigm tests the effect of spatially directed attention Gabor patch Results show that spatial distribution of observers’ attention can be modulated by the presence of certain types of visual images even when they’re interocularly suppressed and invisible Studies have shown that the amygdala can be activated by invisible emotional stimuli Emotion-laden stimuli are effective in modulating selective attention The Abyss: Music and Amnesia (New Yorker, Oliver Sacks) Clive Wearing gets brain infection, left with a memory span of only seconds Both retrograde and anterograde amnesia To imagine the future was no more possible for Clive than to remember the past Clive gets moved to a small country residence for the brain-injured. “Clive was calmer and sometimes jolly, a bit more content, but often with angry outbursts still, unpredictable, withdrawn, spending most of his time in his room alone” Jokiness can go with a weakening of the usual social frontal-lobe inhibitions His “memories” may have just reflect his knowledge about events. Expressions of semantic memory rather than episodic memory To catch sight of his wife was always a relief Perfect preservation of Clive’s musical abilities Procedural, unconscious memory is unimpaired in amnesia His self is seemingly untouched by amnesia, even if his autobiographical self is virtually lost Listening to music is not a passive process but intensely active, involving a stream of inferences, hypotheses, expectations, and anticipations “Clive’s at-homeness in music and in his love for me are where he transcends amnesia and finds continuum Week 3 Day 3 Carey: Origin of Concepts (Number) Core Cognition: Number Plurality is a quantificational commitment Piaget: numerical representations are built from logical capacities. Believed that children could not represent numbers until early school years Disproven Object-file system of representations “Number sense” system, paradigm example of core cognition Core System 1: Analog Magnitude Representations of Number Physical magnitude thats roughly proportional to the number of individuals in the set being enumerated 1) animals represent number Rats trained to press bar a given number of times before a feeder was loaded → mean of each distribution of number of presses is just a little higher than the correct number for that condition Weber’s Law – discriminability is a function of ratio Rats, crows, pigeons, a parrot, monkeys, apes, dolphins have experimentally demonstrated analog magnitude number representations 2) one system of number representation is analog Evidence for Human Infants’ Analog Magnitude Reprsentation of Number Preverbal infants can represent cardinal values of sets of individuals Continuity Through the Life Span Human adults, like infants and animals, create analog magnitude representations of numbers of quite different kinds of individuals Magnitude and distance effects typical of Weber’s law Characterizing the Input Analyzers that Create Analog Magnitude Representations of Number Earliest proposal was the accumulator model of Russell Church and Warren Meck (1984) Gallistel and Gelmen (1992): the accumulator model is formally identical to the explicit counting procedures defined over culturally constructed lists of integer words (one two three four five…) Suggestion that animals perform a computation that depends on two timing mechanisms Temporal interval between onsets of successive tones Overall duration of the tone sequence Iconic Format Number representations are conceptual, content goes beyond spatio-temporal and sensory vocabulary Representation of Sets Not until the end of “preoperational thought” (~5 yrs old) that children were capable of set-based quantification A Second Core System with Numerical Content: Parallel Individuation of Small Sets Number is only implicitly coded, there are no symbols for number at all, not even analog magnitude ones Parallel individuation system (????) underlies performance on many small number experiments The Set-Size Signature of Individual-File Representations Rehsus macaques spontaneously represent number in small sets of objects, can compare two sets with respect to which one has more Numerical Computations Carried out over Parallel Individuation Infants compute 1-1 correspondence between representations of small sets held in short-term memory, determining numerical equivalence and numerical order Infants also sensitive to number in a simple habituation task, as long as the individual sets are distinct from one another Lack of Sensitivity to Number in Properly Controlled Studies of “Number” Representation of Small Sets Number is confounded with continuous variables such as total surface area of the stimuli, total contour length of the stimuli, and so on Parallel Individuation of Individuals Other than Objects Young infants show sensitivity to the distinction between two- and three-syllable words, a puppet jumping, etc Infants compare models on the basis of summed continuous variables in preference to 1-1 correspondence between numbers of events Representation of Sets, Redux Infants able to index at least two sets in parallel A Second Core Cognition System with Numerical Content Purpose of parallel individuation is to create working-memory models of small sets of individuals in order to represent spatial, causal, and intentional relations among them Plays crucial role in the creation of explicit verbal number list representation of positive integers Conclusions: Relations Between Analog Magnitude Number Representations and Paralell Individuation Sensitivity of analog magnitude representations is limited by the ratio of the sets being discriminated Capacity of parallel individuation is limited (in infants) to three individuals Object-file and event-file representations consist of symbols for each individual, properties of those individuals Analog magnitude representations are computable for small sets by infants as well as adults, but in many situations infants instead focus on the individuals The Representation Underlying Infants’ Choice of More: Object Files vs. Analog Magnitudes (Feigenson, Carey, Hauser) Support for the idea that infants rely on object-file representations, comparing mental models via total volume or surface area rather than one-to-one correspondence between object files Quality discrimination is subject to Weber’s law: ratio between quantities determines whether they are successfully discriminated Main question: do infants rely on object-file representations or on analog-magnitude representations in any given quantity task? What computations can be performed over such representations? Experiment 1 Explore whether infants spontaneously represent more/less relationships 1 vs 2, 2 vs 3, 3 vs 4, 2 vs 4, 3 vs 6 Establishing relationships between two hidden quantities to choose the opaque container with more crackers Infants showed clear preferences with small numbers (1v2, 2v3), but not with larger ones (3v4, 2v4, 3v6) Experiment 2 2a: exploring whether infants failed with large numbers for motivational reasons (ex. 3 crackers may be satisfying, so they wouldn’t care to get 6) → they successfully chose the transparently greater quantities of cracker 2b: complexity control, testing possibility that infants failed in large number comparisons because the events were too complex/long to hold interest 2c: duration control, exploring alternative interpretations for the failures in large-number conditions. Addition of hand wave Experiment 3 Determining whether comparisons were made by number of items or total amount of cracker 3a: greater-area condition, total surface area of 1 cracker was twice the sum of surface areas of two crackers 3b: equal area condition. One cracker area = two crackers area Infants appeared to choose on the basis of total quantity, not number of items (surface area!) General Discussion Proposal that object files subserved performance in the present experiments, and likely in other experiments concerning small sets of individual objects Week 4 Day 1: Learning Language Critical period for first language acquisition (Friedmann & Rusou 2015) Examining language outcomes for children who missed critical period for acquiring a first language (isolation, hearing impairment, thiamine deficiency) Critical period ends during first year of life, missing it results in severe syntactic impairments Some suggest that there are critical periods with different time frames for different language abilities This study focus on domain-specificy for syntax acquisition First theory from Penfield and Roberts, later developed by Lenneberg Natural acquisition from mere exposure occurs during a critical period (begins at 2, ends at puberty) Begins after a certain maturation of the brain, ends after certain loss of cerebral plasticity Question → why does the critical period end? What changes in the brain cause this? Lenneberg → loss of plasticity is bc of “lateralization of language functions”. Usually the left hemisphere will become more dominant for language Some think balance between excitatory and inhibitory activity defines critical period What is the functional role of the offset? Theory: it’s required to help keep the structure that’s already been learned, allow for consolidation and stabilization First language acquisition requires a neurologically prepared mind and sufficient language input Chomsky Presenting studies that indicate language input is crucial for normal acquisition of a first language, specifically for syntax Isolated and feral children: insufficient language input These children later failed to acquire language ven after exposure to language and sometimes formal teaching Ex: Genie was kept in isolation from 20 months to ~13 years old Her vocabulary consistently grew, but she failed to develop normal syntax. No question words, demonstratives, particles Single words can be acquired later in life, whereas syntax can no longer be acquired normally Children with hearing impairment: insufficient language input Raised without sign language Don’t receive sufficient input until they’re provided with hearing devices. Often show syntactic impairments Children showed difficulties understanding and producing relative clauses (ex. “This is the girl that the grandma drew”) and Wh-questions like “which girl did grandma draw?” (theme of the action follows the agent) Children who lost their hearing after their first year of life and those whose loss was only monoaural showed normal syntactic abilities Children who did not receive language input during the first year were impaired Some studies suggest that syntactic learning during the first year is derived from other linguistic abilities (prosody, phonology, word frequency) Thiamine deficiency during the first year Vitamin necessary for brain development Severe syntactic difficulties when they were five and nine years old Difficulty with syntactic movement such as Wh-questions, object relative clauses, and topicalized sentences Differences between other types of language abilities → there are different critical periods for different language domains These children were able to acquire new lexical items, but could not recover from syntactic impairment Pharmacological hope Using excitatory-inhibitory balance theory to try manipulating critical periods in pharmacological ways There was a study on development of phoneme discrimination that found children whose mothers used SSRIs during pregnancy showed accelerated phonemic perception (WOAH) Will we all speak emoji? “Calling emoji language is like calling a whale a fish…emoji and language don’t do the same thing” Emoji as a supplement to language, not replacing it entirely A new view of language acquisition (Kuhl 2000) Infants perceptually map critical aspects of ambient language in the first year of life before they can speak Historical Theoretical Positions BF Skinner: language was an “operant” that developed in children as a function of external reinforcement and shaping Chomsky: posited a “language faculty” that included innately specified constraints on the possible forms human language could take. Specification of a universal grammar and universal phonetics. Language was on the primary examples of a module: domain-specific, informationally encapsulated, and innate Emerging view suggests that infants engage in a new kind of learning in which language input is mapped in detail by the infant brain. Six principles reflected this view are covered in this paper Initial Perception Parses Speech Correctly and Is Universal, but Not Domain Specific or Species Specific How do infants parse the auditory world into critical units of language? Universal, phonetic parsing abilities across languages Partitioning seen for speech is not limited to humans or to speech Detecting “phonetic boundary” parsing of phonetic units from birth is from general auditory processing mechanism, not necessarily evolved for speech Acoustic differences strongly influenced selection of phonetic units used in language. Runs COUNTER to two principles: i) view that phonetic units were prespecified in infants ii) view that language evolved in humans without continuity with lower species Development is Not Based on Selection Selectionist view: linguistic experience produced either maintenance or loss. Detectors stimulated by ambient language were maintained, whereas those not stimulated by language input atrophied Listeners do NOT completely lose ability to discriminate non-native contrasts To refute selectionist position, studies must demonstrate that infants listening to ambient language are engaged in some other kind of learning process, not fundamentally subtractive in nature Infants’ Learning Strategies “Map” Language Input Demonstrations that by simply listening to language, infants acquire sophisticated information about its properties Pattern detection in language input Ability to detect and use distributional and probabilistic information contained in ambient language to identify higher-order units Infant perception is literally warped by experience to hance language perception. No speaker of any language perceives acoustic reality; in each case, perception is layered in the service of language Infants can detefct patterns in phonetic units, global pro