Social Brain Hypothesis

Traditionally, evolutionary processes have been attributed to natural selection, favoring the

“survival of the fittest” [96]. Accordingly, the evolution of brains was ascribed to an improvement of

the capacity to process ecologically relevant information. This seems plausible if one considers the

fact that some particular non-human primates and apes acquired complex problem-solving

capacities (also called extractive foraging [97]), but fails to explain why all primates – including the

folivores – have larger brains than other mammals [17]. This is particularly important when keeping

in mind that the adult human brain only makes up about 2% of body weight but consumes about

20% or total energy intake [98], leading potentially to a very disproportionate balance between

costs and benefits. Consequently, an alternative brain evolution hypothesis was brought forward by Robin Dunbar (1998), named the social brain hypothesis. The latter argues that “… the competition for social skills led to the evolution of cognitive mechanisms for outsmarting others, and fuelled the expansion of the human brain and perhaps the elaboration of certain neural systems …” [95].

Of central importance in the social brain hypothesis is the fact that only the neocortex ratio has been found to positively correlate with different indices of social complexity – such as social group size for the anthropoid primate species [99] or the mating system in primates [100]. No such relations could be established for other brain areas, like the hippocampus [101], primary visual cortex, or sub-cortical areas involved in emotion processing like the amygdala [102]. There is even the notion that “emotional” brain areas have progressively reduced in their relative size in favor of more “executive” brain centers, reflecting a shift away from emotional to deliberate control of behavior [103]. Taken all together, it is unlikely that the human brain has increased in size during evolution due to purely visual or emotional mechanisms, or because of an increase in memory capacity per se [17].

Along these lines, an appealing alternative for the driving force of human brain development is seen in the ability to manipulate information about social relationships [17]. This notion is corroborated by several lines of evidence, including – amongst others – 1) a positive correlation between the use of tactical deception rates and neocortex size in primates – a behaviour that is also indicative of mentalizing and theory of mind skills [104]; 2) a negative relation between neocortex size and male rank correlation with mating success in polygamous primates – meaning that the bigger the neocortex of lower ranking males, the more mating success they have, presumably because they can deploy more sophisticated social skills to circumvent dominant males [105]; 3) a positive correlation between adult neo-cortex size and the length of the juvenile period in primates – implying that it is not the hardware embryological development, but the

“software programming” during social learning that accounts for large brains in primates [106]; and 4) a positive correlation between grooming clique size and relative neo-cortex size in primates, including humans (see Figure M5) – being interpreted as “… a direct cognitive limitation on the number of individuals with which an animal can simultaneously maintain a relationship of sufficient depth that they can be relied on to provide mutual support when one of them is under attack …”

[17].

Figure M5: Mean grooming clique size plotted against mean neo-cortex ratio for individual primate genra (reproduced from [17]). The square is Homo sapiens. Prosimians include primates that are not monkeys or apes (i.e. lemurs), whereas anthropoids comprise the group of great apes (like gorilla, chimpanzee, and orang-utan).

The account of Saxe (2006) on the two uniquely human social cognition skills (the capacity for theory of mind and the representation of triadic relations) converges with the social brain hypothesis [17] that also stresses theory of mind as being the primary driving force of evolutionary neocortex size increase, and suggest that specific social cognition skills might mark the most significant difference between humans and non-human primates and other mammals, and that they evolved due to increasing computational demands as a function of social complexity.

For the present thesis, the social brain hypothesis does not represent a crucial account

regarding experiment design or interpretation of obtained results. Nonetheless, it provides valuable

information on the development of some higher-order cognitive processes involved in social

perception, such as theory of mind and mentalizing. Because the latter will be mentioned

repeatedly in the remaining part of the thesis like, i.e., in the results and discussion chapters, it is

useful to know which biological functions they are thought to subserve and how these functions

emerged during human (brain) evolution.

CHAPTER III

Functional Neuro-Anatomy

III.1. The Social Brain

After having explored the fundamental mechanisms and core processes of social perception and emotion processing in general in Chapter I, and discussed the emergence of specific social skills as the driving evolutionary force of human neo-cortex development in Chapter II, the question now is where these social computations are localized within the human brain. As already mentioned several times, social perception includes a multitude of processes, such as cognitive, emotional and motivational, as well as personal and contextual mechanisms. Hence, it is not surprising that almost the whole brain is involved in perceiving, interpreting, and modulating social emotional information.

Figure M6: Familiar Face Recognition Model (original in [33]). It includes brain areas coding for the invariant and dynamic visual features of faces, the retrieval of associated person knowledge, as well as emotional components regarding social behavior and social decision making (see Chapters I.3.2 and III.2.).

When it comes to delineating a functional anatomical model of specific brain areas

implicated in social perception and emotion recognition, two major models are traditionally

considered, representing accounts of either face or emotion expression recognition. As already

pointed out in Chapter I.3.1, one of the most elaborated models of face recognition was described by Gobbini & Haxby [33] (see Figure M6 above), while the most comprehensive model for processing of emotional facial expressions was put forward by Adolphs [34] (see Figure M7 below).

Figure M7: Emotional Facial Expression Processing Model (original in [25]). This model is mostly concerned with early processing steps separating structural versus conceptual representations, including the initial perception of an emotion at the top, and the identification of an emotion at the bottom.

Even though these models provide a sound basis for describing the neural substrates of social perception, separating “core” and “extended” systems that represent the differential encoding of emotion versus identity and featural information about social stimuli (see Chapter I.3.1.), they seem somewhat restricted to basic perceptual stimulus properties, particularly in the case Adolphs’

(2002) model. Moreover, both accounts do not take consider the distinct pre-frontal cortical and

cingulate contributions to social perception. Because the present doctoral thesis was not only concerned with the recognition of emotional facial expressions, but also investigated more complex social perception processes such as the influence of context and individual differences on responses to social partners as well as the regulation of social emotions (see Chapter IV and following), another approach might be more appropriate to describe the brain networks important for social neuroscience.

An alternative approach (see Figure M8 below) distinguishes between different neural systems for social perception and separates between automatic (reflexive, x-system) versus controlled (reflective, c-system) mechanisms [6, 18] (see Figure M8 below and Chapter I.2.2.). The most important conceptual distinctions between these two systems are summarized in Table I.

Table I: Basic features associated with x- and c-systems (original in [6]).

Even though such a dual-process model separating automatic from controlled mechanisms does

not necessarily represent the complexity of social behavior in an ideal manner, it still provides a

very useful bridge between already well established behavioral work and new evidence from

neuro-imaging. In addition, the dual-process model can readily be transcribed into the context of

the present thesis, where emotion perception could be attributed to processes mainly involving the

x-system, whereas emotion regulation could be linked with preferential processing within the

c-system. As a matter of course, such a distinction will not be applicable without any exceptions, as

will become evident by reading the next two sections on the neural substrates of automatic and

controlled social perception, respectively.

Figure M8: Reflexive (x-system; left) and reflective (c-system; right) brain systems (reproduced from [6], original in [24]). For more details, please refer to text in Chapter III.