It is the view in cognition that different domains of knowledge are organized in the brain and processed by distinct neural mechanisms. In the present paper, the authors suggest an extension to the currently existing domain-specific hypothesis, based on connectivity between different functional regions of the brain. They claim that domain expertise in a region is a result of innate connectivity between a group of regions that also process information about the same domain.
Patients with brain-injuries who show category specific semantic impairments also show conceptual level impairments[0]. The categories of category-specific semantic deficits are animate objects, inanimate biological objects and artefacts[1]. The studies suggests that domain-specific constraints dictate organization of conceptual knowledge in the brain.
A lot of research on object knowledge has focused on the ventral visual pathway which has its roots in the early visual areas. It plays a primary role in object recognition and classification. The ventral region exhibits spatial biases according to semantic distinctions, i.e., to say that it differentially processes objects belonging to different categories, eg: faces, tools, body parts, words, fruits etc. Such bias raises questions as to what factors encourage such differential processing in the ventral region. The thrust of the majority of research on the ventral region has been on stimulus driven response. Coming to categorical distinctions, this approach doesnt seem valid in light of the fact that response depends on more than just the primary input, often drawing upon prior knowledge. A classic example here would be the Stroop Test.
Herein, the authors lay their core idea: that what the ventral stream does after object recognition depends entirely on how it is connected to the rest of the brain. The paper studies visual object recognition and taxonomy, but similar ideas should also apply to object recognition via other faculties. The authors argue that such connectivity in the brain is innate in nature, and the driving factor behind semantic specialization in the ventral stream, drawing upon evidence to support their theory.
A domain specific neural network in the brain has two primary characteristics. First is distributed processing, i.e., it consists of distinct regions in the brain each processing a different type of information about the domain[0]. Interesting to note, is the fact that such domains are evolutionary important domains so as to permit a dedicated processing network, good examples being the domain of faces and the domain of tools. Secondly, the set of computations over such a domain must be sufficiently erratic to motivate the formation of such a pathway.
For example, social information is called for when a face is identified. Handling expertise is called upon on recognising a tool or similarly shaped objects. As a personal example, I seem to think of grasping sticks when I notice a peacock's neck. The authors propose that categorical specialization in the ventral stream is expressed as connectivity patterns between the ventral stream and the rest of the brain[0]. For example, specialization of the fusiform face area in the ventral pathway to identify faces is driven by its networking with areas that store socially relevant information. Specificity for tools and manipulable objects in the medial fusiform gyrus is driven, in part, by connectivity between that region and regions of parietal cortex that subserve object manipulation [23-26].
There has been recent evidence refuting the need for visual experience for classification of some types of data in the ventral stream. The authors themselves are credited with the finding that the same medial-to-lateral bias that is present in sighted people, is also present in congenitally blind subjects.
But, if visual experience does not drive organization in the ventral stream, what does? Connectivity! Only through such connections in the brain, can the same areas in a congenitally blind person be activated by the same input that elicits a similar response in a sighted person.
Note that this also points out the fact that connectivity is innate in nature.
Figure 1. Congenitally blind and sighted participants were presented with auditorily spoken words of living things (animals) and nonliving things (tools, non-manipulable
objects) and were asked to make size judgments about the referents of the words. The sighted participants were also shown pictures corresponding to the same stimuli in a
separate scan. For sighted participants viewing pictures, the known finding was replicated that nonliving things such as tools and large non-manipulable objects lead to
differential neural responses in medial aspects of the ventral occipital-temporal cortex. This pattern of differential BOLD responses for nonliving things in medial aspects of
the ventral occipital-temporal cortex was also observed in congenitally blind participants and sighted participants performing the size judgment task over auditory stimuli.
These data indicate that the medial-to-lateral bias in the distribution of category-specific responses does not depend on visual experience. For details of the study, see [44].
The distinguishing feature of an innate property is its similarity across individuals of a species and sometimes even cross-species similarities. Genetic variables capturing the property are a strong indication towards an innate structure. We look at the following cases:
A couple of reports highlight a greater level of functional similarity between monozygotic twins than dizygotic twins. These studies found that in face-related tasks, the responses and performances are much more similar in monozygotic twins than in dizygotic ones(Polk and colleagues[50], Wilmer and colleagues[51]). Interestingly enough, such similarities are not found in verbal and visual memory tasks, indicating genetic selectivity in behaviours.
Another piece of evidence supporting genetic contribution to facial recognition comes from study of prosopagnosia patients, wherein patients are selectively impaired to recognise faces. A recent study by Thomas and colleagues[55] has found, that in such patients, major white matter tracts(responsible for connectivity) linking the posterior occipital temporal cortex with other regions of the brain are missing, thus, highlighting the importance of a network level analysis in object organisation.
Another piece of evidence supporting genetic contribution to facial recognition comes from study of prosopagnosia patients, wherein patients are selectively impaired to recognise faces. A recent study by Thomas and colleagues[55] has found, that in such patients, major white matter tracts(responsible for connectivity) linking the posterior occipital temporal cortex with other regions of the brain are missing, thus, highlighting the importance of a network level analysis in object organisation.
In functional imaging studies in monkies, and more recently in macaques[57] and chimpanzees[58], it was found that atleast for facial recognition, similar clusters of face-preferring voxels in the brain can be found in the temporal cortex of monkeys, as they do exist in humans[0]. It may be argued, that such patterns might exist only due to visual similarity, as is known to happed in the IT cortex. But, continued attempts to explain away this observations in terms of visual similarity have failed, often revealing tight taxonomic structures.