We view primary-process homeostasis, secondary-level learning and memory, and direct perception as neural systems that co-evolved and provided the necessary space for the decoupling of affect from its here-and-now functions to encapsulate complex social and imaginative representational capabilities.
The property of decoupleability refers to offline processing of information. It indicates thinking of a concept when it is not part of the creature’s present action, target of action, or perceptual context. Decoupling is the process that cleaves present-tense perceptual indicative percepts from instrumental proto-beliefs (see our discussion of body grammar in the last post). Affect as conative motivational drive is amenable to being decoupleable because it predates—and remains functional—through all evolutionarily later cognitive abilities; that is, its primacy ensures that it has a use within any mental context. And, unlike other mental functions, affect can filter through any mental operation, infusing pertinent elements with salience; affect dyes our thoughts with value and meaning. Accordingly, we have described several roles played by affect including, as a mode of presentation, as an intentional arrow, and as motivation for locking onto appropriate affordances.
Perception and action depend upon the imperative forms of informational transfer between creature and environment, which we described as affordances. Salience within the perceptual world occurs via affective goads that dynamically covary with homeostatic needs and lead to action patterns, such as further information-seeking behaviors. The function of affect in perception is as a mode of experience, and specifically, as a subjective motivational force. Affect functions as an approach/avoid value in affordance space, whether it be social space or the spatial navigation landmarks described below.
A core concept in the cognitive sciences, representation, has the functional role of acting as a decoupleable surrogate for specifiable extra-neural states of affairs. While full-blown decoupleability is certainly a paradigmatic quality of representational mind, our thrust here is that affect is susceptible to a more simple decoupling process that could function via mnemonic processes and direct perception of landmarks in a minimally representational mind. We believe the varied role of affect has long been overlooked in this regard and claim that, in both spatial navigation and dream states, affect is a decoupleable mental quality that, in the evolution of mind, may have served as a bridge to tertiary-level representational processes. As an attractor in spatial navigation and as a somatic marker in dreams and mnemonic processes, affect could be said to specify inner neural states of affairs in a feedback relation to extra-neural states of affairs, thus establishing the decoupling of primary-level affective states from their origins in the here-and-now of homeostatic functioning.
We claim the initial instances of functional decoupleability arose within a perception-action core, with affective valence-coding as goads in spatial navigation, and within the dream state wherein affective valence tags are unconsciously manipulated for the purpose of organizing social space. In the involuntary imagination of dreams, the landmarks in cognitive maps, and the burgeoning imaginative faculty, affective tags may be decoupled from their initial objects, creating a fluid space for the abstraction of value and meaning. It is only after these formats are in place that symbol systems and the subsequent compositional explosion of the linguistic mind becomes possible.
Affect and Spatial Navigation
Mediating between mind, behavior, and world, cognitive maps are mental structures that offer a substantive example of how a structural relation between observer and environment can afford movement, simulation, and opportunities for decision-making. Such spatial cognition necessitates fine-tuning of motor movement, executive planning, and even inhibition-related delayed gratification. The most primitive body structure in the brain is the somatosensory and proprioceptive body image; next are spatial egocentric and allocentric maps. This layering of maps of the bodily self suggests a functioning isomorphism between a set of homeostatic and movement processes in the brain and behaviorally important aspects of the world, such that elements of the map can be considered as symbols. As the inner model develops from an egocentric to an allocentric map, it eventually allows the capability for simulation (i.e., simulating different possible routes) and structured mnemonic storage of information. Allocentric maps differ from evolutionarily later forms of symbolic representation, like language, in several ways; most notably, they are iconic, lack predication, are analogue, and are not compositional. These allocentric maps are good candidates for nonconceptual content, specifically as landmark affordance simulation spaces.
We contend that the value or meaning of the landmark in a cognitive map is an affective valence tag that summarizes mnemonic information regarding the landmark as a motivational vector. The affective tag may simply average, via conditioning, past affective states at the landmark—for example, “Food was in this spot” equals positive valence—to indicate “approach” or “avoid” as a simplified attractor goad on a cognitive map. In more complex maps, or layers of maps, a landmark can indicate relational geometrical information pertaining to destination, relative to goal states. Goal states may be generated by primary-level homeostatic needs and then associated with landmarks by secondary-level learning and mnemonic processes like operant conditioning. A recent ethological example of such a process is the so-called episodic memory in scrub-jay food caching, which demonstrates how elements of a spatial map can have values determined by the length of time a food article remains edible. In these studies, it was determined that scrub-jays seem to keep track of which articles of food are stored in which areas. Our interpretation of the data is that spatial cognition and valence of map elements (e.g., the affective urgency of retrieval) is responsible for the bird’s behavior.
Elaborating on our earlier discussion of the mind as an embodied-embedded-extended process, we suggest that affect in spatial navigation plays an affordance- like role; it suggests basic behaviors—viz. approach/avoid—toward landmarks.
Decoupling in dreams
In dreams, when atonia, REM, and changes in neurotransmitter levels occur, the dreamer is largely absent from the perceptual here- and-now and the mind enters a (re-)calibration zone of emotional training. At this level of consciousness, the dreamer indeed still experiences affective states; we know this because motivational mechanisms and dopaminergic activity are essential for the generation of dreams.
In being decoupled from agency and here-and-now perceptual-motor tasks, synthetic mental processes are able to organize/reorganize affect relative to mnemonic content. Imagination in the dream state, where the mind unconsciously and involuntarily mediates between perception, memory, and judgment, is a good candidate for predecessor-to-conscious forms of explicit decoupling, simulation, and voluntary imagination.
The adaptive function of dream decoupling in our story is as a form of mnemonic-affect schema consolidation; this could occur for spatial landmarks, as well as in social animals, for the purposes of generation and refinement of appropriate social behaviors. The hippocampus is known for its role in mediating cognitive maps, receiving affective components of incoming sensory stimuli, creating declarative memories, computing spatial orientation, and providing information about the context in which conditioning has taken place. Evidence from behavioral neuroscience concerning the role of the hippocampus dovetails well with our hypothesis that both spatial landmark goads and dream memory reconsolidation occur when affect is decoupleable.
It is in the dream state that, we hypothesize, a type of memory reconsolidation occurs, wherein malleable emotional memories like dominance hierarchies can be modified and recalibrated. Experiments suggest that fear conditioning, as well as procedural memories and some memory distortion in episodic memory, can be modified in the reconsolidation process—in or outside of the dream state. We suggest a dynamic reconsolidation process where affective tags (or landmarks) can be shifted and learned in an implicit, unconscious format at the level of secondary-level affective processes, in both spatial navigation and in the involuntary imagination of the dream state.
In the next post, a parallel story will be told about the development of concepts, which themselves may have developed from cross-modal affect-perception units.
 Cf. Griffiths, Paul E., and Andrea Scarantino. 2005. Emotions in the wild: The situated perspective on emotion. In Cambridge Handbook of Situated Cognition. Cambridge: Cambridge University Press.
 Schafer, M. & Schiller, D. 2018. Navigating Social Space. Neuron 100, 476-489.
 An egocentric map is a map based around the body axis of one’s self, and allocentric maps encode information about objects relative to each other.
See SELF in Panksepp, J. 1998. Affective Neuroscience. USA: Oxford University Press.
Asma, Stephen, and Thomas Greif. 2012. Affective neuroscience and the philosophy of the self. Journal of Consciousness Studies 19 (3/4): 6–48.
Damasio, A. 2010. Self comes to mind: Constructing the conscious brain. New York: Random House.
 Rescorla, Michael. 2009. Cognitive maps and the language of thought. British Journal for the Philosophy of Science 60 (2): 377–407.
Millikan, R. 2004. On reading signs: Some differences between us and the others. In D. Kimbrough Oller, Ulrike Griebel, Gerd B. Muller, Gunter P. Wagner, and Werner Callebaut, eds., Evolution of communication systems: A comparative approach, 15–29. Cambridge, MA: MIT Press.
Sloman, Aaron. 1978. The computer revolution in philosophy: Philosophy, science, and models of the mind. Hassocks, UK: Harvester Press.
 See Clayton, N. S., and A. Dickinson. 1998. Episodic-like memory during cache recovery by scrub jays. Nature 395 (6699): 272–274. Though the scrub jay is not a mammal, we think analogous processes arose in the mammalian clade. Also see on insect navigation, Gallistel, C. R. 1990. Learning, development, and conceptual change. The organization of learning. Cambridge, MA: MIT Press.
 O’Keefe, J. & Nadel, L. 1978. The Hippocampus as a Cognitive Map. USA: Clarendon Press. Furthermore, according to Solms, Mark. 2000. Dreaming and REM sleep are controlled by different brain mechanisms. Behavioral and Brain Sciences 23 (6): 843–850 (discussion 904), normal spatial cognition has been found to be essential for dreaming.
 Schafer & Schiller, ibid.
 Cf. Nader, L., and O. Hardt. 2009. Update on memory systems and processes. Neuropsychopharmacology 36 (1): 251–273.
 For fear conditioning, see Kindt, Merel, Marieke Soeter, and Bram Vervliet. 2009. Beyond extinction: Erasing human fear responses and preventing the return of fear. Nature Neuroscience 12: 256–258.
For procedural memory, see Walker, M. P., T. Brakefield, J. A. Hobson, and R. Stickgold. 2003. Dissociable stages of human memory consolidation and reconsolidation. Nature 425 (6958): 616–620.
For episodic memory, see Hardt, Oliver, Einar Örn Einarsson, and Karim Nader. 2010. A bridge over troubled water: Reconsolidation as a link between cognitive and neuroscientific memory research traditions. Annual Review of Psychology 61: 141–167. Hupbach, Almut, Rebecca Gomez, and Lynn Nadel. 2013. Episodic memory reconsolidation: An update. In Cristina M. Alberini, ed., Memory reconsolidation, 233– 247. San Diego, CA: Academic Press.