Touching to connect, explore, and explain: how the human brain makes social touch meaningful

ABSTRACT Human touch has an enormous power to engender and mediate meaning in the human mind, from the emotional to the pragmatic, and from the linguistic to the symbolic. Can a functional-neuroanatomical perspective on social touch contribute to a general understanding of the biological workings of such meaning-making? I argue here that it can, and that the ways the brain accomplishes this are manifold. I identify and explore three main neural subsystems which operate in concert to generate the emotional and semantic complexion of social touch. These subsystems underlie how humans: 1) touch to connect with others; 2) explore the physical and social worlds; and 3) explain the significance of a touch within our own knowledge and experience, especially with regard to the way we interpret the world through language and culture. I therefore propose that what makes social touch meaningful has much to do with the functional and evolutionary roots of these brain subsystems. Although they can be distinguished and analyzed, in the “wild” human brain these subsystems are functionally intertwined, and their processes are integrated to generate a unified subjective experience of social touch. This view also acknowledges the intertwined nature of the embodied individual within society, thus carrying potential implications for theoretical analysis in such terms.


Introduction
In Arnold Bennett's novel Hilda Lessways, the character Edwin Clayhanger hurries away from a brief encounter with Hilda Lessways " . . . with a delicate photograph of the palm of her hand printed in minute sensations on the palm of his" (Bennett 1911, 177). What makes Hilda's touch more than just a mechanical indentation of Edwin's skin? For that matter, what makes any touch from another person meaningful? Humans use touch to mediate emotional connection with other people. We also use it to explore and manipulate our surroundings, including the social surroundings. Not least, a human touch is colored and transformed by memory, expectation, and the cultural and semantic frames that allow us to interpret and explain our daily world. This paper therefore explores social touch in terms of its role in these three important aspects of touch embodiment: social attachment (touching to connect), exploration, learning, and play (touching to explore), and contextualization (touching to explain).
Although the main vantage point of this paper is the human brain, it explicitly acknowledges that a person is not bounded by their skull or even by their skin. Rather, the human brain is a collaboration between evolution, experience, and society. Just as populations of neurons and regions of the cerebral cortex are connected with one another in dynamic, mutually-influential interactions, so individual humans are connected with one another in webs of relationships. These webs are, in turn, causally inextricable from the workings of individual brains. With this in mind, I attempt to address the question of how the brain sciences might contribute to a general understanding of social touch, by describing relevant aspects of neural anatomy and function. What follows is necessarily speculative, but is grounded as far as possible in the current evidence.
The human brain carries millions of years of history with it. Its ways of meaning-making are multifarious and rich, but they are also integrated. In fact, they are so seamlessly integrated as to seem more or less unified on a subjective level (Fulkerson 2014). However, on the level of the underlying mechanisms, myriad neural subsystems have emerged over evolutionary history in response to specific yet distinct pressures. Though they have come to coexist mostly harmoniously within an adult human nervous system, these neural subsystems are nevertheless intertwined; they are wrought into a whole from different sources, like a tapestry made from a variety of materials close to hand. Part of the beauty of the evolving and malleable brain is that emergent features can become bases for new aptitudes, new propinquities, and new adaptations. The neural underpinnings of human affective touch thus reflect how multiple, evolutionarily-intertwined systems can enable -or even create -essential functional aptitudes.

Touching to connect
The roots of embodied social attachment lie in regulatory systems that calibrate the body's physiological functions using touch information, whether it be tactile, thermal, or "proprioceptive" (the "self-sensing" perception of the position in space of one's body parts). But what might these kinds of inner physiological processes have to do with interactions and relationships with other people? It is because, in a real sense, the regulatory systems of individual bodies are connected to others by invisible threads that keep us together and safe. These invisible threads form part of the mutuallydependent interactions between people. Not only that, but they have been engineered by evolution to create feelings that tug on our hearts.
At times this is quite literally the case, since many of these regulatory systems that give rise to subjective feelings are involved in modulating the activity of the heart, lungs, and viscera. They include the autonomic branches of the nervous system as well as those which send cascades of messenger molecules into the body's causeways to influence organ function. As the author Anne Enright has noted, "I think we can safely say that when we are moved, it is some liquid that starts moving: blood, or milk, or salt water" (Enright 2013). These physiological regulatory systems have been summed up under the label "homeostatic," or sometimes "allostatic" systems (Craig 2002;McEwen 2000).
These systems have traditionally been studied in terms of causes and effects within an individual body. The study of stress mechanisms is an example, centering on the so-called hypothalamic-pituitary-adrenal (HPA) axis and the sympathetic-adrenal-medullary (SAM) axis. The HPA axis aids in mobilizing glucose (a sugar) in muscle and organ cells to produce rapidly-available energy (McEwen 2000), while the SAM can facilitate fast oxygen delivery to muscles and other tissues via in-the-moment modulation of heartbeat and breathing (McEwen 2000;Gunnar and Hostinar 2015). However, the individual body is not a closed loop with respect to the social environment. Other people in the social environment can influence individual-level regulatory systems such as the functioning of stress responses, as can factors such as socioeconomic status (e.g. Johnson et al. 2021). It is thus worth exploring what bearing "external" systems of social relationships might have on "internal" systems of bodily regulation and feeling (Mayo and Heilig 2019).
The kernel of an answer may lie in a very fundamental fact about humans: we are born helpless and vulnerable. The mammalian trait of giving birth to immature young is called "altriciality." Most mammals are altricial, as are many bird species (Scheiber et al. 2017). The likely advantage to this arrangement is that it can give experience and learning a longer window to shape our brains and behavior compared to, for example, a tiny lizard that hits the ground running directly after emerging from the egg. Evolution by natural selection at the gene level cannot directly produce quick adaptive responses at the sometimes millisecond-scale of environmental change faced by the individual. Genetic change is simply too slow. Instead, we evolved nervous systems which are plastic and relatively quick-changing, and which can be tailored adaptively to an individual's local conditions within a lifetime. The disadvantage is that this can take a long time to set up, at least in the beginning of a life. The required time for reaching maturity can be longer or shorter depending on the species, but humans lie well toward the longer end of the spectrum. However, altriciality itself does not seem to be related to the complexity of a species' social environment (Scheiber et al. 2017).
Despite its advantages, altriciality presents a stubborn practical issue. The issue is simply that after birth two or more bodies, which need to remain together for the successful maturation of the offspring, become physically separate entities. This is a risky arrangement: at any moment any of these entities can wander off, or fall down a hole, or get snatched away by a predator. I call this the "proximity problem" (Morrison 2016b). As many parents are keenly aware, someone needs to care for and nurture young altricial animals until they are sufficiently mature. That job can fall to one or both parents, and in some species even more distantly related kin can also lighten the burden. In any case, parents and offspring need some way of sticking together that does not rely on a physical lifeline, like an umbilical cord or a marsupial pouch. Accordingly, strong evolutionary selection pressures have fostered behaviors that keep parent and offspring together during this critical period. From the inside, these behaviors are reinforced by associated feeling states -spinning out the invisible threads.
Touch acts as a fundamental mediator for a range of behavior and feeling states engendered by the brain in social attachment. It is not the only one. For example, vocalizations and odor are two other major mediators of social attachment in mammalian species. But in humans and other primates, touch has a place at or near center stage when it comes to solving the "proximity problem" by forging and maintaining social connection. The brain achieves this via mechanisms which either reduce the likelihood of separation, or increase the likelihood of reunion after separation (Morrison 2016b). It does so via neural processes which give rise to changes in motivational and emotional states which, in turn, alter an individual's behavior with respect to another individual.
What might such neural processes look like? Evidence from mammals and birds indicates that a kind of lock-and-key neuron signaling mechanism, the "mu-opioid" system, is associated with a feeling state which has been labeled "separation distress" (Nelson and Panksepp 1998;Panksepp 1998;Panksepp et al. 1978;Panksepp, Nelson, and Bekkedal 1997), especially with reference to immature animals. Mu-opioid receptors (the lock) are expressed in brain cells and can change cell signaling when occupied by opioid molecules (the key). Being within touch-proximity to kith or kin keeps an animal in plentiful supply of these opioid molecules, such as endogenous endorphins. Yet becoming separated and out of touch-proximity brings empty, unoccupied receptors. On a subjective level, this can contribute to a feeling of bereftness -abandonment, loneliness, social isolation -a species of feelings which we work hard to avoid by seeking social interaction.
The hormone oxytocin, on the other hand, acts as a key to the receptor-locks within systems that mediate stress responses and the reinforcement of learned behavior, with the power of dampening the former and enhancing the latter (Feldman et al. 1994;Kurosawa et al. 1995;Tang et al. 2020;Wang et al. 2017;Xiao, Priest, and Kozorovitskiy 2018). Oxytocin-related signaling within the brain and body can manifest in behaviors that increase the likelihood of sticking with the familiar rather than roaming to new pastures; it can also bolster the existing patterns of social relationships, as well as gradually adapting itself to new ones (Quintana and Guastella 2020;Handlin et al. 2023).
The late neuroscientist Jaak Panksepp proposed that core functional properties of the mu-opioid system, as well as the oxytocin system, have been co-opted into the social realm and further elaborated by natural selection (Nelson and Panksepp 1998). Under this proposition, opioid systems attain a functional role in brokering those woeful separationdistress feelings, prompting "seeking" behavior (Panksepp et al. 1978) which increases the likelihood of reunion. Likewise, oxytocin's effects on smooth muscle action during reproduction, parturition (childbirth), and lactation (nursing) may have come more and more under the sway of behavior-dependent brain signaling over evolutionary time (Quintana and Guastella 2020). For example, oxytocin's role in modulating cardiac responses and other stress-regulatory processes is likely under influences from the cortex, particularly medial prefrontal cortex (Handlin et al. 2023), which is heavily involved in emotional and cognitive processing and implicated in social interactions (Kietzman and Gourley 2023). The OT receptor is expressed here and elsewhere throughout the brain (Quintana et al. 2019). When it comes to social touch, human oxytocin levels change in a contextdependent manner depending on who is doing the touching and on the prior history of social interaction with a given individual (Handlin et al. 2023). Apart from hormonal modulation, other human neuroimaging studies have shown a broad involvement of reward-related structures during affectively-relevant touch stimulation (e.g. Perini, Morrison, and Olausson 2015;Handlin et al. 2023).
Hormones and receptors may fluctuate, but more stable neural hardware has also been implicated in emotional touch. A particular type of nerve fiber in the skin has a history going back to our shared ancestors with reptiles and fish (Morrison 2022;Semmes 1969). This type of nerve fiber is called a "C afferent" ("C" is the notation for the category of nerve fiber, and "afferent" means that the signaling goes from the body's periphery toward the brain). C afferents follow a certain pathway up to the brain via the spinal cord, and this pathway is present in all mammals studied so far. This older afferent system is very slow-conducting -it lacks the fatty electrical insulation of other kinds of afferents -and it tends to send out side-branches to communicate with other parts of the spinal cord and nervous system on its rather leisurely way up to the brain. C afferents and their various subtypes are associated with homeostatic domains such as temperature, itch, and pain (Craig 2003). A subtype of C afferents ("C-tactile" or "CT" afferents) that respond to caresslike touch and a range of other stimuli have been specifically implicated in affective touch in humans (Löken et al. 2009;Morrison 2012;Morrison et al. 2011;Morrison, Loken, and Olausson 2010).
This class of afferent nerve also includes those which respond to changes in temperature, hinting at an intriguing link between touch and temperature at this level. Mice born without an important cell receptor channel for C afferents (Nav1.7) showed blunted sensitivity to gradual skin cooling and tended to spend their time in cooler parts of the experimental environment than mice with intact Nav1.7 channels (Middleton et al. 2021). The putative equivalent of CT afferents in mice (referred to as "C low-threshold mechanoreceptors or C-LTMRs) also respond to decreases in skin temperature (Abraira and Ginty 2013). In humans, CT afferents respond preferentially to caress like touch stimuli at a "creature temperature" of 32° Celsius (Ackerley et al. 2014).
From the Middle Ages and onward Western culture has strongly associated warmth with comfort (Claessen 2012;Crowley 2005), perhaps especially in chillier European climates. Beneath these associations tick biological temperature regulation mechanisms, which respond to changes in outside temperature by signaling which ensures the maintenance of the exquisitely narrow temperature range in which the body can operate (Morrison, Nakamura, and Madden 2008;Nakamura and Morrison 2008). Getting too cold can require releasing heat from the body's existing resources in the form of burning energy, either as immediately available caloric energy (as in shivering) or that stored as fat. If such physiological fixes fail, the nervous system prompts the body to use its skeletomuscular apparatus to move from a colder to a warmer place (Farrell and Alberts 2007), for example to seek the warmth of a fireside.
When the comfort of warmth is derived not from a sunbeam or a hearthside but from a fellow creature, this is referred to as social thermoregulation (Alberts 2007;Farrell and Alberts 2007;Gilbert et al. 2007Gilbert et al. , 2010Gilbert et al. , 2012. Social thermoregulation provides a particularly clear example of how a regulation mechanism "designed" to solve a specific problem can ramify into other domains and extend from infancy further into the life span. Many newborn mammals lack the ability to efficiently regulate their body heat in the face of cold. Luckily, however, parents and siblings can serve as cheap and convenient space heaters. Being close to others thus grants a metabolic benefit by siphoning the heat energy of other bodies. Other individuals thus become part of the thermoregulatory equipment of the newborn, in a way that co-occurs with touch. Indeed, the researcher Alberts described the rat dam as a "moveable feast of experiences" for the pups huddling beneath and around her, a feast which "remains in place while they root, squirm, shift, probe, and push. It undulates with each breath. Throughout such contact, conductive heat exchange is maintained" (Alberts 2007).
Crucially, when sensory and homeostatic systems guide the individual in following the local social warmth gradient (moving toward the living heat source when too cold, and away when too warm), it involves not just warmth but touch. Getting warmed up after having been cold feels good. By extension, being close to another living being and feeling their touch can also feel very good. Warmth and touch may be coextensive not just because others' bodies are warm to the touch, but also because warmth and touch can be coded by the same nerve "hardware," as suggested by the CT and C-LTMR evidence mentioned above. Social touch thus becomes open to becoming a metabolically-lessexpensive shorthand for "safe" and "warm," from the point of view of the nervous system's energy economies. This is because it takes more energy to warm oneself up on the basis of thermal signaling than for touch to engender a sense of warmth on the basis of touch signaling. Social touch may be telling the body that everything is fine, and we experience the corresponding neural signaling as contentment and pleasure (Morrison 2016b). In general, social interaction may buffer stress responses (Chun et al. 2022), thus lowering the likelihood of needing an emergency metabolic loan from the HPA or SAM. After all, the body would much rather spend its energy budget on long-term investments such as growth and healing -but also on luxuries such as exploration and play.

Touching to explore
Think of a baby on the verge of being coordinated enough to investigate an interesting colorful object hanging from a mobile: she stretches out these very useful devices mounted on the ends of her arms and aims them at the object. She can have no idea that she is only at the beginning of a lifetime of manual action and manipulation. Our exploratory use of touch contributes to our understanding of the world and to our ability to extract meaning from it (Tuan 2005). In this, the hands are central; more often than not they are the leading edge of our interactions with objects and people. In this regard, the hands likely do not respect distinctions between objects and people, but can be enlisted in the exploration of each.
The roots of such embodied exploration, learning, and manipulation lie in sensorimotor circuits centering on motor and sensory regions of the cortex. These systems underlie the range of aptitudes that grant humans remarkably fine hand control, tactile acuity, and, arguably, certain quintessential features of human cognition such as the way we perceive the world in terms of action affordances (Gibson 1962). In a real sense, our brains have "oozed into the muscles of [our] fingers," as those of the carpenter onboard the Peaquod in Moby Dick (Melville 1851, ch 107). This high level of sensorimotor hand control has earned our species large dividends. Dexterity brings certain survival benefits, all the way from increased calorie extraction from tricky-to-handle foods (think of eating a prickly pear fruit with your hands), to tool use, to literally digital computational thinking.
How do the underlying sensorimotor circuits grant meaning to touch? It begins with relatively direct input from fast-conducting and generously-insulated afferents (Aβ afferents) projecting from the body's periphery to the brain via the spinal cord. These reach primary and secondary somatosensory cortices (SI and SII), which are heavily involved in sensorimotor control of forepaws/hands, lips, and (where applicable) whiskers. This system has high spatial and temporal acuity, and it registers tactile events on organized body map in the brain's cortex in easily-quantifiable terms such as spatial coordinates and intensity metrics (Case et al. 2016(Case et al. , 2017Sailer et al. 2016).
In contrast, SII regions, along with nearby posterior insular cortex, have been associated with aspects of affective touch (Grandi 2016;Morrison 2016a;Perini, Morrison, and Olausson 2015). "Affective touch" lacks a precise operational definition, but it generally refers to cutaneous tactile stimulation that is associated with a reported positive hedonic subjective rating (e.g. pleasantness), regardless of stimulation site or stimulus type (Morrison 2016a). Since the mid-twentieth century, affective touch has been conceived as a vital component of social affiliation and bonding, with investigations extending into the realms of social behavior and health, and caregiving (Burnside 1973;Seaman 1982). Sometimes it is used (somewhat circularly) to refer to CT-mediated touch perception.
Despite various region-level dissociations and specializations, the brain's processing of touch ultimately occurs on the level of networks (Masson and Op de Beeck 2018). By working together in a network, the brain can increase flexibility and speed of sensing and responding, reduce uncertainty and noise surrounding the touch event, discover any correlations between the touch and other events, and transform social touch sensations into future behavior. In these sensorimotor brain circuits and networks, somatosensation is tightly bound to action.
Goal-directed intentionality is what distinguishes an action from a movement. Movements become orchestrated into actions in the service of an intention to do something, whether that is grasping a coffee mug or talking to one's friend. The relationships between intentions, goals, and meaning are tightly linked in networks that span the parietal lobes, which include many somatosensory-specific regions, alongside regions of the frontal lobes which contain many neural populations key for turning goals and intentions into actions (e.g. de Haan and Dijkerman 2020; Dijkerman and de Haan 2007). Another hallmark of these sensorimotor networks is that they are highly predictive. In recent years, support has been gathering for a perspective in which the brain's sensory perception is not geared simply for describing and elaborating stimulus features, but rather for adapting behavior to the circumstances surrounding the stimulus (e.g. Clark 2013). For example, cortical sensorimotor cortices may be organized not primarily as a spatial body map, but instead according to complex behavior repertoires -whole actions -relevant to the acting organism, such as grasping a piece of food and bringing it to the mouth (Graziano 2016).
Once in place, the useful features of fine sensorimotor dexterity -tactile sensitivity, exploration, and intentional action -can be co-opted to serve social roles. For example, they provide a basis for predictions or "simulations" of one's own actions that can help in planning actions and anticipating their consequences "offline," without necessarily having to perform them first (Clark 2013). This trick of "predictive coding" can be extended into the social realm to infer meaning from others' actions and even others' pain (e.g. Morrison 2005). When it comes to social touch, its mutuality implies a dynamic process, starting with an active, intentional initiation of touch which must be intricately coordinated with any receptivity to being touched, and not least of any reciprocation. This process unfurls over time, as well as over countless re-enactments in close relationships, all involving a dance of intention, perception, and action, which is socially and culturally infused (Kirsch et al. 2018;McIntyre et al. 2021;Masson and Op de Beeck 2018;Suvilehto, Cekaite, and Morrison forthcoming). However, it should be noted that in laboratory experimental contexts, the cooperative compliance of volunteer research participants is probably often mistaken for true receptivity to an experimenter's touch, with true receptivity being the spontaneous willingness to engage in, and respond to, social touch under naturalistic conditions -for example hugging your child when she comes home from school. Such receptivity can elude capture under artificial laboratory conditions.
A less well-investigated aspect of touch is its role in play. Play is part and parcel of exploration. It is how we find limits and test possibilities. Sensorimotor systems are engaged when we "fiddle around" with objects or even apply more hard-won skills such as improvisational dexterity on a musical instrument. But play also looms large in the social realm. We use our hands for playing with each other too, across various contexts from the intimate to the ritualistic.
Mammals use play-biting or other derivative displays in which teeth are bared, and in which participants send and receive signals (like laughter) which distinguish play from real aggression (Graziano 2018). During such interactions, defensive responses are relaxed and a receptivity to touch is probably increased. In a touch-play-attack such as a tickle fight with a child, laughing and other cues (perhaps even the tickle sensation itself) convey signals that no harm is meant or done. Could play-touch like tickling be analogous to smiling and laughing? There is very little research on this idea, though the observation that rats produce special high-pitched trills during touch-mediated social play has entered the public imagination (Panksepp and Burgdorf 2003). Touch-play skates a fine line between the poles of play and defense, fun and discomfort, advance and retreat, vulnerability and receptivity. Its meaning emerges as a product of constant, reciprocal, hand-to-skin negotiations between the interactants -and it is one in which context must play a crucial role.

Touching to explain
The central nervous system's processing of any sensory event comes down to a single question "asked" by the brain: What does this mean for me? "Meaning" here can invoke the possible implications of a situation for behavior and action, but it can also invoke the relational properties of a human touch within complex and interacting networks of brain, body, and social relationships: in other words, its context. As humans, we have a strong tendency to explain the world and its phenomena: we try to make sense of things and circumstances and strive to render them intelligible. We generate interpretations of the world through abstract structures such as language and mathematics, which can take shape as narrative and theory, for example. The nature of the concepts "meaning" and "explanation" have been treated extensively (and quite variously) by philosophers, so a detailed discussion is beyond the scope of this review. However, a general trend in these philosophical ideas is to invoke a relation between entities, for example between the word "cake" and any of the various manifestations of sweet, spongy baked goods one might encounter in the world. Language-and culture-mediated systems of meaning can thus play a large role in how humans interpret and perform social touch.
The temporal lobe is central in these respects. One of the most wonderful specializations of the human brain is that it enables us not only to speak, but to create poetry and to be moved by hearing it; to sing and to be moved by song. At the root of this kind of meaning-making lie neural subsystems, headquartered in the temporal lobe, that associate and integrate sensory information with memory, contextual features, semantic meaning (with reference to language and similar systems such as logic), and semiotic interpretation (with reference to signs and symbols, for example gestures or icons). Temporal lobe cortex is wrapped around a delicate twist of tissue. This twist is the seahorse-shaped hippocampus, the vital seat of memory. Temporal cortex has dense connections with hippocampus and neighboring entorhinal cortex, which are critical in forming, preserving, and altering associations (Erickson, Rauschecker, and Turkeltaub 2017). The temporal lobe is also highly connected with further-flung brain regions associated with somatosensory and spatial function in the parietal cortex, such as the ones discussed in the previous section (Igelstrom and Graziano 2017;Igelstrom, Webb, and Graziano 2015;Yeo et al. 2011). It also contains auditory subregions necessary for processing sound signals from the ear. At a coarse level, the temporal lobe performs basic functions in integrating sensory and other relevant contextual information to derive meaning (Braunsdorf et al. 2021;Carter and Huettel 2013).
"Meaning" is notoriously difficult to define, but in the case of the temporal lobe it involves the transformation and embedding of sensory information from multiple sources within associative networks of abstract features. Language provides a good example. The word "rich" can be pronounced or read: in the former case it is a temporal pattern of sound waves, in the latter a static visual pattern of letter shapes. But in neither case is it experienced as such. Instead, when we hear or see the word "rich" we understand its meaning, as well as its contextual uses and potential connotations. It might even evoke imagery or memories, such as a picture of Gatsby in his opulent library full of uncut books (Fitzgerald 1925), or a recollection of a rich slice of cake tasted on some long-ago evening at a beachside café. Such integrative and essentially semantic hallmarks of language are consistent with the current evidence for various specializations of the temporal lobes in humans.
Over evolutionary history, the temporal cortex has expanded in humans compared to other great apes (Braunsdorf et al. 2021;Bryant et al. 2019). In humans, certain functions of the temporal lobe show pronounced hemispheric lateralization (a concentration of certain functions within the left or right counterpart brain structure), most notably for language (Braga et al. 2020), but also for recognition of faces and graphical symbols in category-specific temporal visual cortex (Burns, Arnold, and Bukach 2019;Kanwisher and Yovel 2006;Vinckier et al. 2007). It also is associated with emotional aspects of speech prosody and music (Mauchand and Zhang 2022;Norman-Haignere et al. 2022), and temporal lobe seizures can have emotional and autonomic effects, which are sometimes experienced as sublime or revelatory (Armstrong 2005). This makes sense given the emotional nature of memory and its importance in contextualization of experience, perhaps particularly when it comes to complex, socially-relevant information.
Indeed, evidence has been growing for a role of temporal lobe subregions in social and affective touch. Yet any such role has been under-researched, despite the region's recurring yet inconsistent appearance across human brain imaging studies of affective touch (e.g. Brauer et al. 2016;Davidovic et al. 2016;Singh et al. 2014;Gordon et al. 2013;Voos, Pelphrey, and Kaiser 2013). Superior temporal sulcus (STS) and the nearby temporoparietal junction (TPJ) are implicated in multisensory integration, as well as in aspects of semantic processing and mental state inference (Allison, Puce, and McCarthy 2000;Igelstrom and Graziano 2017;Perrett et al. 1985). From a functional connectivity perspective, this region also participates in a network that is strongly connected with insula and sensorimotor cortex (Igelstrom, Webb, and Graziano 2015) and has probable indirect connections with parts of the brain that play a strong role in reward and learning (Depue and Morrone-Strupinsky 2005;Xiao, Priest, and Kozorovitskiy 2018;Gothard and Fuglevand 2022;Mosher et al. 2016). Temporal lobe engagement during social touch has also recently been shown to depend on the levels of oxytocin in the blood (Handlin et al. 2023). Further, pharmacological manipulations using the drug 3,4-methylenedioxymethamphetamine (MDMA or "ecstasy") indicate that this "feelgood" amphetamine enhances the subjective pleasantness of touch and increases attention to happy faces (de Wit and Bershad 2020) and and blood oxytocin levels (Kirkpatrick et al. 2014).
A given touch from another human being, like the one Hilda gave to Edwin, becomes an event among these existing webs of semantic and social relationships, and is experienced as meaningful by virtue of how it is accommodated within them. In these relationships, it is conceptually useful to identify core contextual features of touch behavior in social interactions (Suvilehto, Cekaite, and Morrison forthcoming): why touch has come to be functionally important (this centers around evolution and mechanisms); who is involved in a given interaction and what their relationships are (this encompasses individual-and relationship-level features such as gender and status); and how touch is used in a given situation (this mainly occupies the sphere of cultural practice). It is important to consider the contribution of each of these aspects to the question of how contextual factors participate in the brain's meaning-making during social touch (see also Ellingsen et al. 2015;Sailer and Leknes 2022).
The why of social touch has to do with the evolutionary dimensions of its affective nature, such as the ones discussed in an earlier section. Another example regarding possible reasons why we have inherited social touch from our primate ancestors is the grooming of others ("allogrooming,"). Socially close individuals in many mammalian species groom each other, which has demonstrable hygienic benefits in removing dirt and parasites (Nelson and Geher 2007). But in many primate species, including humans, allogrooming has expanded beyond hygiene to become linked to satisfaction and pleasure (Keverne, Martensz, and Tuite 1989) as well as to social status and hierarchy and, intriguingly, social uses of language such as gossip (Dunbar 2010). In humans, touch can also communicate intentions and emotional states (Hertenstein et al. 2006(Hertenstein et al. , 2009Thompson and Hampton 2011;Kirsch et al. 2018;McIntyre et al. 2021;Hauser et al. 2019). Despite so frequently being studied as an isolated channel of communication, touch is often embedded in the stream of talk during social interactions (Goodwin and Cekaite 2018;Bergnehr and Cekaite 2018;Cekaite 2020;Jones and Yarbrough 1985;Wingenbach et al. 2019).
The who of social touch is pervasive and includes individual-level features such as age, gender, social status, social role, and personality, as well as what kind of relationship pertains between the interacting individuals (Suvilehto et al. 2015(Suvilehto et al. , 2019Thompson and Hampton 2011). For example, in healthcare settings, touch may be instrumental and "task-based" or affective and comforting (Burnside 1973;Seaman 1982) and is affected by the gender and social roles of professional and patient (Kelly et al. 2018;Routasalo 1999). Such issues also lie at the interfaces between medicine, identity, and technology (De Falco et al. 2022). In addition to mechanisms of attachment and social relationships discussed here, the social-level dynamics of reciprocity, hierarchy, and coalition are also likely critical (Prescott and Robillard 2021;Bugental 2000), though under-researched. All can to some extent be mediated by touch. Unfortunately, there is a dearth of research addressing the likely large role of individual differences in social touch experience and behavior (but see Luong et al. 2017;Sailer, Hausmann, and Croy 2020), despite the intuition that " . . . some of those cunningly fashioned instruments called human souls have only a very limited range of music, and will not vibrate in the least under a touch that fills others with tremulous rapture or quivering agony" (Eliot 1959, ch IX).
The how of social touch becomes especially relevant where the goals and intentions of the interactors feed the meaning and interpretation of a touch in a social context (Sailer and Leknes 2022), and of course where the behavior is highly inflected or even determined by cultural practice (Suvilehto, Cekaite, and Morrison forthcoming). Self-report of affective touch experience has revealed cultural variation across 45 countries (Sorokowska et al. 2021). Touch taboos are an important domain, especially when it comes to the location on the body where touch occurs (Jourard 1966;Suvilehto et al. 2015Suvilehto et al. , 2019 and how it is interpreted by the interactants (Nguyen, Heslin, and Nguyen 1975).
In such acts of interpretation, the meaning-making surrounding touch again spills out into extended contexts and domains. For example, humans can infer aspects of emotional states and relationships merely by observing others' interactions (Morrison, Loken, and Olausson 2010;Nguyen, Heslin, and Nguyen 1975;Beckes, IJzerman, and Tops 2015;Linden 2014). Indeed, as Georg Christoph Lichtenberg remarked in the 18th century, "It is possible to stroke someone's cheek in such a way that a third party feels as if he's had his ears boxed" (Waste Books L, 62). In every sphere from attachment to action, meaning-making and touch combine like chemical elements to energize and illuminate the cosmos within.

Summary and conclusions
The neural underpinnings of social touch are heterogeneous, reflecting the dynamic activity of intertwined subsystems and processes. This review has identified three such subsystems that make major functional contributions to how humans derive forms of meaning from social touch, and considers the conceptual bases and neuroanatomical evidence for each. First, neural mechanisms supporting social attachment underlie emotional responses and bonding, allowing individuals to connect via touch. Second, touch is central in touch exploration and in many forms of play. Third, the integration of sensory processing with memory, context, language, and culture can imbue touch with meanings that foster interpretation of tactile contact during social interactions.
Touching to connect enlists neural mechanisms that regulate the body's physiological functions via processes which can alter motivational and emotional states and, ultimately, behavior. For example, social separation or isolation decreases the level of endorphins available in the brain, which can give rise to unpleasant feelings of loneliness or abandonment which we seek to remedy through active behavior, such as seeking out others. When we touch to explore by using our hands to interface with the physical and social worlds, we are relying on sensorimotor circuits which underlie a range of aptitudes that grant humans remarkably agile hand control and fine tactile acuity during the performance of actions. Since these sensorimotor circuits can also predict action consequences, they may contribute to an immediate understanding of the actions, intentions, and feelings of other people through inference of goals and intentions. Finally, we can touch to explain via neural pathways which integrate sensory perception with memory, feeling, and relevant contextual and associative information. This variety of meaning-making lends itself easily to the ways we interpret the world in terms of language, art, and frames of cultural practice.
These subsystems and processes form a gyre that widens out from the individual brain to encompass the body, friends, family, society, and the furthest reaches of our cultural habitats. The neuroscience of social attachment, our talents for sensorimotor coordination of hand actions, and our human knack for emotional and semantic contextualizationeach of these human aptitudes shed their own useful light on the nature of social touch. The resulting view of social touch as a manifold yet functionally integrated and subjectively unified phenomenon can provide a conceptual basis for further theoretical exploration of social touch across disciplines, including the biological and social sciences and the humanities.