- Email: [email protected]
‘Seeing’ sound
News & Comment
movement, eye position, the reach to a target and attentional orienting. Monkeys were given a task where the goal was to reach for a food pellet, which was either in a stationary food well or in a well attached to a moving robot. The monkey knew the location of the food at the start of each trial. Crucially, it was the robot that signalled to the monkey when to reach for the food. So, regardless of the location of the target, the monkey had to attend to the robot for the first part of a trial. As expected, many dorsal PMC cells responded to eye position, saccade direction and the direction of the planned reach. However, about 20% of the cells were defined as ‘attention-tuned’, meaning that they responded to what the monkey was attending to, independently of motor or oculomotor factors. The PMC therefore also has a role to play in attention. HJB
TRENDS in Cognitive Sciences Vol.5 No.5 May 2001
is more uniform. The 144-electrode system, which, according to Bach-y-Rita requires 50 hours of training for proficiency, has already been used to pilot subjects in simple navigation tasks. The next step will be to create a wireless link to an inconspicuous camera, to be mounted on spectacles. Another application – to help divers see in murky waters with a sonar rather than a video camera – has already been tested by the US Navy. MW
It’s all in the timing
A taste for space Can we see only with the retina? Paul Bach-y-Rita has previously shown that people, including the blind, can also see when a 2-D array of tactile stimulators transduces signals from a camera – as long as the camera is actively controlled by the observer (as the eye is). Recently, Bach-y-Rita and Paul Kaczmarek at the University of Wisconsin have improved the system by trading tactile for mild electrical stimulation, and placing the stimulators inside the mouth on the tongue. Because of the sensitivity of the tongue and the presence of saliva, which is a good conductor, smaller currents can be used to deliver stimulation that
auditory–visual interaction might occur [Groh, J.M. et al (2001), Neuron 29, 509-518]. Researchers trained monkeys to listen to sounds played from a semicircle of speakers, while looking left, right or straight ahead. They then recorded from cells in the inferior colliculus (IC), an early part of the auditory pathway. As expected, IC cells responded differently to sounds from different locations. Crucially however, a third of IC cells also altered their response to a particular sound if the direction of the monkey’s gaze changed. This is the first demonstration that such a ‘low-level’ area is involved in integrating hearing and vision. HJB
Memories stuck in a rut
Bird-watching at a dog show? If an expert bird-watcher can instantly tell apart two similar species that are indistinguishable by an untrained observer, how do the brains of the expert and nonexpert differ? James Tanaka and Tim Curran have uncovered part of the answer in an elegant study that used event-related potentials [Psychol. Sci. (2001) 12, 43–47]. Experts on birds and dogs, respectively, classified pictures of both types of animals. When bird experts saw pictures of birds, they had a stronger early negative component, called the N170, than when they saw pictures of dogs, for which they had no expertise. The opposite was true of dog experts. This double dissociation proves that the effect is due to expertise, not to the visual information presented. Interestingly, the N170 potential has also been tied to face recognition – a domain in which we are all experts. MW
189
Timing is crucial to many aspects of human performance – from catching a cricket ball to delivering a punch line. A recent study sheds new light on the identity of the brain’s timekeeper [Rao, S.M. et al. (2001) Nature Neurosci. 4, 317–323]. Subjects were asked to compare the time intervals between two pairs of sounds. Using event-related fMRI, the researchers were able to separate out activation occurring at different stages of the task. Activation after the first pair of sounds reflected encoding of timing information, whereas later activation associated with comparison of the time intervals reflected more general processes of decision making and response preparation. Early activation was detected in the basal ganglia and right parietal cortex. This concurs with the observation that people with Parkinson’s disease, which affects dopamine transmission in the basal ganglia, are impaired with regard to time perception. By contrast, activation of the cerebellum, which is often hypothesized to be the brain’s timekeeper, was detected only later in the task. HJB
‘Seeing’ sound
Humans have recall that is far from perfect, but memory can also act like a stuck record, repeatedly recalling the same mistaken information. Work by Catherine Fritz and colleagues, reported in the British Journal of Psychology, suggests that we tend to recall what we previously recalled, and fail to correct the memory. This exploration of the ‘failure-of-furtherlearning’ effect follows up past experiments by Harry Kay, in which subjects recalled a passage that was read to them. Between each re-presentation of the text the subjects (college students) received other, unrelated information. They then attempted to recall the passage accurately. Catherine Fritz’s new experiments reinforce and extend Kay’s surprising result – that re-presentation of the same text does not significantly improve recall or understanding. The initial recall tends to persist, errors included, as if once the memory is fixed, it can’t be unlearned. Fritz and colleagues also show that even when the identity and style of the presenter is varied from one presentation to the next, no improvement is seen; and that providing the listener with a list of previous errors results in even poorer recall. By contrast, subjects were good at remembering the meaning of the passages after just one presentation. Students may be surprised to discover that revision seems to be a waste of time! But these experiments involve a very specific task, verbatim recall, and presentation of varied material was not tested. So perhaps it is premature to predict the end of revision. DPB
When we watch a ventriloquist perform, our perception that the voice is coming from the dummy relies on the fact that what we hear is influenced by what we see. A new study has identified an area of the brain where this
http://tics.trends.com 1364-6613/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved.
In Brief articles written by Heidi Johansen-Berg Dominic Palmer-Brown and Mark Wexler
movement, eye position, the reach to a target and attentional orienting. Monkeys were given a task where the goal was to reach for a food pellet, which was either in a stationary food well or in a well attached to a moving robot. The monkey knew the location of the food at the start of each trial. Crucially, it was the robot that signalled to the monkey when to reach for the food. So, regardless of the location of the target, the monkey had to attend to the robot for the first part of a trial. As expected, many dorsal PMC cells responded to eye position, saccade direction and the direction of the planned reach. However, about 20% of the cells were defined as ‘attention-tuned’, meaning that they responded to what the monkey was attending to, independently of motor or oculomotor factors. The PMC therefore also has a role to play in attention. HJB
TRENDS in Cognitive Sciences Vol.5 No.5 May 2001
is more uniform. The 144-electrode system, which, according to Bach-y-Rita requires 50 hours of training for proficiency, has already been used to pilot subjects in simple navigation tasks. The next step will be to create a wireless link to an inconspicuous camera, to be mounted on spectacles. Another application – to help divers see in murky waters with a sonar rather than a video camera – has already been tested by the US Navy. MW
It’s all in the timing
A taste for space Can we see only with the retina? Paul Bach-y-Rita has previously shown that people, including the blind, can also see when a 2-D array of tactile stimulators transduces signals from a camera – as long as the camera is actively controlled by the observer (as the eye is). Recently, Bach-y-Rita and Paul Kaczmarek at the University of Wisconsin have improved the system by trading tactile for mild electrical stimulation, and placing the stimulators inside the mouth on the tongue. Because of the sensitivity of the tongue and the presence of saliva, which is a good conductor, smaller currents can be used to deliver stimulation that
auditory–visual interaction might occur [Groh, J.M. et al (2001), Neuron 29, 509-518]. Researchers trained monkeys to listen to sounds played from a semicircle of speakers, while looking left, right or straight ahead. They then recorded from cells in the inferior colliculus (IC), an early part of the auditory pathway. As expected, IC cells responded differently to sounds from different locations. Crucially however, a third of IC cells also altered their response to a particular sound if the direction of the monkey’s gaze changed. This is the first demonstration that such a ‘low-level’ area is involved in integrating hearing and vision. HJB
Memories stuck in a rut
Bird-watching at a dog show? If an expert bird-watcher can instantly tell apart two similar species that are indistinguishable by an untrained observer, how do the brains of the expert and nonexpert differ? James Tanaka and Tim Curran have uncovered part of the answer in an elegant study that used event-related potentials [Psychol. Sci. (2001) 12, 43–47]. Experts on birds and dogs, respectively, classified pictures of both types of animals. When bird experts saw pictures of birds, they had a stronger early negative component, called the N170, than when they saw pictures of dogs, for which they had no expertise. The opposite was true of dog experts. This double dissociation proves that the effect is due to expertise, not to the visual information presented. Interestingly, the N170 potential has also been tied to face recognition – a domain in which we are all experts. MW
189
Timing is crucial to many aspects of human performance – from catching a cricket ball to delivering a punch line. A recent study sheds new light on the identity of the brain’s timekeeper [Rao, S.M. et al. (2001) Nature Neurosci. 4, 317–323]. Subjects were asked to compare the time intervals between two pairs of sounds. Using event-related fMRI, the researchers were able to separate out activation occurring at different stages of the task. Activation after the first pair of sounds reflected encoding of timing information, whereas later activation associated with comparison of the time intervals reflected more general processes of decision making and response preparation. Early activation was detected in the basal ganglia and right parietal cortex. This concurs with the observation that people with Parkinson’s disease, which affects dopamine transmission in the basal ganglia, are impaired with regard to time perception. By contrast, activation of the cerebellum, which is often hypothesized to be the brain’s timekeeper, was detected only later in the task. HJB
‘Seeing’ sound
Humans have recall that is far from perfect, but memory can also act like a stuck record, repeatedly recalling the same mistaken information. Work by Catherine Fritz and colleagues, reported in the British Journal of Psychology, suggests that we tend to recall what we previously recalled, and fail to correct the memory. This exploration of the ‘failure-of-furtherlearning’ effect follows up past experiments by Harry Kay, in which subjects recalled a passage that was read to them. Between each re-presentation of the text the subjects (college students) received other, unrelated information. They then attempted to recall the passage accurately. Catherine Fritz’s new experiments reinforce and extend Kay’s surprising result – that re-presentation of the same text does not significantly improve recall or understanding. The initial recall tends to persist, errors included, as if once the memory is fixed, it can’t be unlearned. Fritz and colleagues also show that even when the identity and style of the presenter is varied from one presentation to the next, no improvement is seen; and that providing the listener with a list of previous errors results in even poorer recall. By contrast, subjects were good at remembering the meaning of the passages after just one presentation. Students may be surprised to discover that revision seems to be a waste of time! But these experiments involve a very specific task, verbatim recall, and presentation of varied material was not tested. So perhaps it is premature to predict the end of revision. DPB
When we watch a ventriloquist perform, our perception that the voice is coming from the dummy relies on the fact that what we hear is influenced by what we see. A new study has identified an area of the brain where this
http://tics.trends.com 1364-6613/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved.
In Brief articles written by Heidi Johansen-Berg Dominic Palmer-Brown and Mark Wexler