Psychology of Speech Perception

Perception Chapter 3

• Perception is a constructive process that interprets sensory information to create internal representations of the world.
• These representations are not exact copies of the real world.
• Perceptions are based on estimates from sensory signals, experience, and expectations.
• Perception is similar to problem-solving.
• Sensory information is limited and can be noisy and distorted.
• The inverse projection problem refers to the challenge of determining the object responsible for an image on the retina.
• The retinal image is ambiguous, as many different object configurations can produce the same image.
• Parts of objects can be hidden, and images can be blurred or distorted.
• Objects appear differently from different viewpoints, impacting the retinal image (viewpoint invariance).

Sensing the World

• Senses are adaptations developed to respond to selective pressures in different environments.
• Different organisms have evolved different sensory capabilities to fill different niches.
• Sensory examples include a frog feeding on flying insects, a moth detecting female pheromones, and humans hearing the range of human voice.

Sensory Systems

• To perceive the world, physical energy must be detected and converted into neural signals via transduction.
• Sensory organs contain transducers that convert physical stimuli (sights, sounds, smells) to neural impulses.
• Examples of sensory organs include eyes, ears, and skin.

Sensation and Perception

• Sensation: The encoding of stimulus energy into neural impulses, and extracting basic information about the stimulus.
• Perception: Selecting, organizing, and interpreting sensory signals to form the conscious experience of objects and their relationships.

Overview: From Stimulus to Perception

• Energy carries information about the world, which can be limited, noisy, and distorted.
• Sensory organs capture and modify energy.
• Receptors transduce energy into a neural response.
• Sensory nerves transmit the encoded response to the brain.
• The thalamus processes and relays the neural response to specialized areas of the cortex for processing.
• Brain areas process this information in parallel to create the perception of the world.

Influences on Perception

• Bottom-up processing: Perception based on incoming stimuli and sensory signals being registered by receptors.
• Top-down processing: Perception based on stored information in the brain (prior knowledge, experience, expectations).

Approaches to Understanding Perception

• Direct perception theories: Bottom-up processing, where sensory stimuli drive perception by assembling parts into a coherent whole.
• Constructive perception theories: Top-down processing, where perceivers actively construct perception by using prior expectations and knowledge to organize stimuli.

Top-Down Processing

• Top-down processing influences perception by using previous knowledge and context to make inferences.
• This is combined with bottom up processing to create perceptions. This is very rapid and automatic.
• The visual system simultaneously considers sensory information from an object and the surrounding environment with feedback based on prior knowledge and expectations.

Gestalt Laws of Perceptual Organisation

• Gestalt's laws describe rules of perceptual organization, explaining how sensory elements are grouped to form objects.
  • Law of good continuation: Lines tend to be perceived as following the smoothest path. Objects overlapped by other objects are perceived as continuing beyond overlap.
  • Law of good figure or Pragnanz: Patterns tend to be seen as simple as possible.
  • Law of similarity: Similar objects are grouped together.
  • Law of proximity: Objects that are close together are grouped together.
  • Law of common fate: Objects moving in same direction or related are grouped .
• Gestalt principles predict common environmental configurations.
• Gestalt laws often work well but are heuristics, not perfectly accurate.

Heuristics and Algorithms

• Heuristic: A practical method to solve a problem quickly, based on experience and a "rule of thumb", often approximate but fast.
• Algorithm: A step-by-step procedure guaranteed to solve a problem, often slow but accurate if the right information is available.

Environmental Regularities

• The environment shows many frequent and predictable stimuli.
• The brain uses this information to improve perception.
• Perception is most sensitive to stimuli that are present often in the environment.
  • Oblique effect: Verticals and horizontals are perceived more readily than other orientations due to frequent environmental experiences.
  • Light-from-above heuristic: Light usually comes from above, which helps perceive depth and distance via shadows.

Semantic Regularities

• Semantic regularities: Characteristics associated with the function and location of objects in different environments.
• These experiences allow predictions to be made about objects and scenes.
• The brain is influenced by individuals' prior experiences of objects and locations.
• Supporting research demonstrates that context plays a significant role in perception, confirming predictions influenced by familiarity (Palmer 1975).

Bayesian Inference

• The brain utilizes Bayesian inference to combine prior knowledge (prior probabilities) with incoming sensory information (likelihoods) to estimate the probability of various possible perceptions.
• Prior probabilities: Estimate the general probability of something occurring (prior experiences).
• Likelihoods: Estimate probability of an event given current evidence.

Neuronal Codes and the Environment

• Our perceptual abilities are determined by neurons tuned to frequently occurring stimuli.
• Natural selection has shaped these abilities through experience.
• Experience can change how neurons respond (experience-dependent plasticity), adapting to environmental regularities.
• This adaptability is particularly apparent in adapting to stimuli in the environment (e.g. Greeble training demonstrates how experiences and training change neural responses to categories of stimuli).

Perception and Action

• Separate brain systems are responsible for perception and controlling motor actions.
• The what pathway (ventral pathway) aids conscious object identification; while the where pathway (dorsal pathway) controls coordinated actions with objects.

What & Where System: Dissociation Logic

• Dissociation logic is used to understand a complex system, by understanding what mechanisms contribute to specific functions by observing malfunctions.
  • Single Dissociation: Damage to one area affects one function but not another, suggesting separate mechanisms.
  • Double Dissociation: Damage in two opposite ways affect two different functions, supporting separate mechanisms for the functions. Examples of this with patients or animals.
• Brain ablation is often used for these kinds of studies.

Neuropsychological Dissociations

• Cases of people with neurological problems can be studied to understand what areas of the brain are responsible for specific functions and how malfunctions reveal the separate systems.