Two-streams hypothesis

History

Several researchers had proposed similar ideas before the hypothesis was formalized. The authors credited inspiration from work on blindsight by Weiskrantz and previous neuroscientific vision research. Schneider proposed two visual systems for localisation and identification in 1969, and Ingle described two independent visual systems in frogs in 1973. Trevarthen had offered an account of two separate mechanisms of vision in monkeys in 1968.

In 1982, Ungerleider and Mishkin distinguished dorsal and ventral streams from lesion studies of monkeys, proposing the original where versus what distinction. This framework was later superseded by that of Milner and Goodale. Experimental work on visual agnosic patient D.F., first reported by Goodale and colleagues in 1991, has been a hugely influential source of information, though the model has faced criticism for perceived over-reliance on findings from a single case.

Two visual systems

Goodale and Milner amassed anatomical, neuropsychological, electrophysiological, and behavioural evidence for their model. According to their data, the ventral 'perceptual' stream computes a detailed map of the world from visual input for cognitive operations, while the dorsal 'action' stream transforms visual information into egocentric coordinates for skilled motor planning. The model posits that visual perception encodes spatial properties relative to other objects, using relative metrics and scene-based frames of reference, whereas visual action planning uses absolute metrics via egocentric frames of reference.

Grasping movements directed at objects in size-contrast-ambiguous scenes have been shown to escape the effects of illusions, as different frames of reference and metrics are involved. Norman proposed a similar dual-process model describing eight main differences between the two systems. The dorsal stream projects from the primary visual cortex to the posterior parietal cortex and was initially termed the 'where' pathway. Later research on Patient D.F. revealed that the dorsal stream processes visual information needed to construct representations of objects one wishes to manipulate, leading to its updated nickname as the 'how' pathway.

The dorsal stream is interconnected with the parallel ventral stream, which runs downward from V1 into the temporal lobe, and is involved in spatial awareness and guidance of actions, containing a detailed map of the visual field and detecting movements.

Two auditory systems

The two streams hypothesis of audition was largely proposed by Hickok and Poeppel in 2004, building on the visual two streams hypothesis. It suggests that cortical structure and processing merge perceptual circuitry with motor coordination to match sound onto meaning and onto articulatory-based representations. Hickok and Poeppel also theorize that both dorsal and ventral streams may have a level of bidirectionality, allowing them to function in various ways together or apart to facilitate speech perception and production.

A ventral auditory pathway emerges from the primary auditory cortex, where phonemes are processed posteriorly to syllables and environmental sounds. This information then joins the visual ventral stream at the middle temporal gyrus and temporal pole, converting auditory objects into audio-visual concepts. The ventral system is colloquially referred to as the 'what' pathway, focusing on lexical interface to map sound onto meaning. The function of the auditory dorsal pathway is to map auditory sensory representations onto articulatory motor representations.

Hickok and Poeppel claim this pathway is necessary because learning to speak is essentially a motor learning task, requiring a neural mechanism that codes and maintains instances of speech sounds to guide the tuning of speech gestures.

Criticisms

Goodale and Milner's innovation shifted the perspective from input distinctions to the functional relevance of vision for perception or action. Contemporary perspectives, informed by empirical work over the past two decades, offer a more complex account than a simple separation of function into two streams. Recent experimental work has challenged findings, suggesting that the apparent dissociation between effects of illusions on perception and action may be due to differences in attention, task demands, and other confounds.

However, other empirical findings provide strong support that skilled actions like grasping are not affected by pictorial illusions. Recent neuropsychological research has questioned the validity of the dissociation between visual agnosia and optic ataxia, with Hesse and colleagues demonstrating dorsal stream impairments in patient DF, and Himmelbach and colleagues applying more rigorous statistical analysis showing the dissociation was not as strong as first thought.

A 2009 review concluded that while the spirit of the model has been vindicated, the independence of the two streams has been overemphasised. Goodale and Milner themselves proposed the analogy of tele-assistance, viewing the dorsal stream as semi-autonomous under guidance of executive functions informed by ventral stream processing. The emerging perspective is that a two-systems framework was a necessary advance, but the reality likely involves considerable interaction between vision-for-action and vision-for-perception.