adaptive behavior turing instabilities in biology, culture, … under disinhibited, non-ordinary...

Original Paper

Adaptive Behavior21(3) 199–214� The Author(s) 2013Reprints and permissions:sagepub.co.uk/journalsPermissions.navDOI: 10.1177/1059712313483145adb.sagepub.com

Turing instabilities in biology, culture,and consciousness? On the enactiveorigins of symbolic material culture

Tom Froese1,2,3, Alexander Woodward1 and Takashi Ikegami1

AbstractIt has been argued that the worldwide prevalence of certain types of geometric visual patterns found in prehistoric artcan be best explained by the common experience of these patterns as geometric hallucinations during altered states ofconsciousness induced by shamanic ritual practices. And in turn the worldwide prevalence of these types of hallucina-tions has been explained by appealing to humanity’s shared neurobiological embodiment. Moreover, it has been proposedthat neural network activity can exhibit similar types of spatiotemporal patterns, especially those caused by Turinginstabilities under disinhibited, non-ordinary conditions. Altered states of consciousness thus provide a suitable pivotpoint from which to investigate the complex relationships between symbolic material culture, first-person experience,and neurobiology. We critique prominent theories of these relationships. Drawing inspiration from neurophenomenol-ogy, we sketch the beginnings of an alternative, enactive approach centered on the concepts of sense-making, value, andsensorimotor decoupling.

KeywordsEnaction, sense-making, representation, hallucination, Turing patterns, human cognition

1 Introduction

Forms of artistic expression, rituals, and language areuniversal features of all known human cultures. Whileanimal behavior is mostly governed by immediate envi-ronmental demands and the meaning of animal com-munication is generally fixed by biological evolution,human symbolic practices and their meanings are thehistorical outcome of a seemingly open-ended socialprocess of cultural evolution. Explaining the origin ofsuch symbolic practices in terms of our biologicalembodiment and the social circumstances of our prehis-toric past is one of the major outstanding challenges ofscience. Indeed, the origin of symbolic representationconcerns the origin of the human condition as such,hence the idea of modern humans as the ‘‘symbolicspecies’’ (Deacon, 1997) or ‘‘Homo symbolicus’’(Henshilwood & d’Errico, 2011b). This is an issue thatreaches across the great divide between the natural andsocial sciences, and has implications beyond the remitof science itself.

The enactive approach to cognitive science is in agood position to advance the debate about the originsof the earliest symbolic practices. Whereas traditionalcognitive science presupposes the existence of an inter-nal ‘‘symbol system’’ as its basic starting point (Newell

& Simon, 1976), the enactive approach instead startswith the biologically embodied mind, and must there-fore address the ‘‘cognitive gap’’ between adaptivebehavior and abstract human cognition (De Jaegher &Froese, 2009; Froese & Di Paolo, 2009). Thus, ratherthan simply assuming the notion of representation asits most basic conceptual foundation for explaining theinternal mechanisms of human cognition, it tries tounderstand the biological, social, and historical originsof the phenomenon of symbolic representation as anaspect of our cultural environment. This approachhelps to avoid a lot of the unnecessary linguistic andtheoretical confusion, which has plagued cognitive sci-ence for many years (Harvey, 2008). And it also helps

1Ikegami Laboratory, Department of General Systems Studies, Graduate

School of Arts and Sciences, University of Tokyo, Tokyo, Japan2Instituto de Investigaciones en Matematicas Aplicadas y en Sistemas

(IIMAS), Universidad Nacional Autonoma de Mexico (UNAM), Mexico3Centro de Ciencias de la Complejidad (C3), Universidad Nacional

Autonoma de Mexico (UNAM), Mexico

Corresponding author:

Departamento de Ciencias de la Computacion, Instituto de

Investigaciones en Matematicas Aplicadas y en Sistemas (IIMAS),

Universidad Nacional Autonoma de Mexico (UNAM), Apdo. 20-726,

01000 Mexico D.F., Mexico.

Email: [email protected]

to bring to light some unresolved foundational issues.The question therefore becomes how primary adaptiveprocesses of sense-making of the here and now couldhave been transformed into secondary forms of sym-bolic sense-making of the absent and the imaginary(Froese, 2012).

Enactive accounts of this profound qualitative tran-sition are still in their infancy, but there is broad agree-ment that there is no one single biological or socialexplanatory factor. We are dealing with a historicalprocess emerging from the interaction between biologi-cal processes, social practices, and the cultural back-ground (Froese, in press; Hutchins, 2010; McGann,2007; Stewart, 2010). The development of an accountthat theoretically bridges the cognitive gap thereforepromises to simultaneously provide an interdisciplinarybridge between the natural and social sciences.Furthermore, because the interaction between culturalprocesses and biological embodiment is mediated at thepersonal level through lived experience, this scientificintegration is not limited to objective processes, butalso includes phenomenology of the first-person per-spective. Cognitive science thereby becomes an‘‘anthropologically informed cultural neurophenome-nology’’ (Laughlin & Throop, 2009). The aim of thispaper is to contribute to this enactive approach to thevarious interplays between biology, culture, and con-sciousness, in particular by critically examining the roleof altered states of consciousness in the generation ofgeometric prehistoric art.

1.1 Overview of the paper

We first take a closer look at the one of the most promi-nent theories of the origin of prehistoric art and high-light its main shortcomings. The prevalence of certaingeometric patterns in the symbolic material culture ofmany prehistoric cultures, starting shortly after theemergence of our biological species and continuing insome indigenous cultures until today, is explained interms of the characteristic contents of biologicallydetermined hallucinatory experiences. However, weargue that the correlation between the first artisticmotifs and typical hallucinatory experiences is not suffi-cient to serve as a full explanation. In particular, thereis a lack of consideration of the value associated withaltered states of consciousness, both in terms of phe-nomenology and function. What is it about these non-ordinary visual patterns that made them more attractivefor artistic expression than most others of an almostinfinite set of possible patterns, both physical and ima-ginary?1 Given that humans appear to be in principlecapable of arbitrarily associating any kind of stimuluswith any kind of meaning, as epitomized by language asan open-ended symbol system, there is a need to explainthe shared selective biases that are in evidence acrossprehistoric cultures. In other words, we need to account

for the cross-culturally shared value of these specifickinds of geometric patterns.

We then evaluate current models of the neural basisof geometric hallucinations, because basic neural pro-cesses are prime candidates for explaining this cross-cultural selective bias. These models typically proposeto view the visual system as a potentially excitablemedium whose autonomous dynamics are unleashed bydisinhibiting altered states of consciousness. Essentially,it is suggested that the disinhibited visual system gener-ates spatiotemporal patterns of neural activity due toTuring instabilities. Researchers also generally claimthat the geometric hallucinations experienced by thesubject are mental representations of these neural pat-terns. However, while these neural models are capableof reproducing some of the geometric patterns that arefound in prehistoric art and non-ordinary visual experi-ences, their range remains severely limited. In addition,the models tend to trivialize the relationship betweenthe structure of subpersonal neural processing and thecontent of personal visual experience by assuming thatvisual experience is a representation of inverse opticsapplied to neural activity in region V1 of the visual ner-vous system.

We then propose an enactive approach to resolvingthese issues. By drawing inspiration from the methodof neurophenomenology, we argue that the role of self-sustaining neural activity, which is closely associatedwith neural Turing instabilities, has been underappre-ciated. According to an enactive approach, self-sustaining neural dynamics can generate their ownintrinsic value in relation to their conditions of self-maintenance, and they can also serve as a neuralmechanism by which to decouple autonomous brainactivity from the influence of environmentally mediatedsensorimotor dynamics. Both of these aspects can helpto explain the aesthetic selective biases of the first art-ists, in particular their interest in inner experience asexemplified by abstract hallucinations and imaginaryphenomena, which are not directly related to thedemands of their physical environment. We speculatethat the self-sustaining dynamics may account for whythese geometric hallucinations were experienced asmore significant than other phenomena, and that at thesame time their underlying neural dynamics may haveserved to mediate and facilitate a form of imaginarysense-making that is not bound to immediatesurroundings.

2 On the origins of the symbolic mind

Starting with Darwin’s (1871) evolutionary approach tothe origins of human language in terms of natural andsexual selection, the date for the beginning of symbolicthought and creative imagination has been continuallypushed back into prehistory. It was long assumed that

200 Adaptive Behavior 21(3)

symbolic material culture originated during the last IceAge in Europe. This standard view was supported bythe available archaeological evidence, for instance bythe famous Paleolithic cave paintings that were createdby the first immigrating Homo sapiens starting fromaround 40,000 years ago. However, more recently thearchaeological consensus has begun to be overturnedby new evidence of an even older symbolic material cul-ture (Henshilwood & d’Errico, 2011a). These findingswere uncovered in Africa, and indicate the presence ofsymbolic practices over 100,000 years ago, includingthe manufacture of paint and body ornamentation. Inaddition, and contrary to the idea that artistic expres-sion began with primitive pictorial representations ofthe external environment, there was a tradition ofengraving pieces of red ochre with a variety of abstractgeometric patterns (Henshilwood, d’Errico, & Watts,2009). An example is shown in Figure 1.

This discovery connects well with the ethnographicalobservation that geometric visual patterns have beenplaying an important role in prehistoric cultures fromaround the world (Froese, in press). Interestingly, thesepatterns represent only a small subset of the potentiallyinfinite set of possible abstract visual patterns thatcould be created, and it seems that certain kinds of pat-tern are particularly frequent, such as dots, circles,cross-hatchings, parallel wavy lines, and especially vari-ous kinds of spirals (Lewis-Williams & Dowson, 1988).

An important clue to the origin of this selective biaswas found by psychologists investigating the experien-tial effects of various kinds of alterations of conscious-ness. They discovered that the patterns of many visualhallucinations could be categorized into a small num-ber of so-called ‘‘form constants’’ (Kluver, 1967). In the1920s, Kluver had systematically studied the effects ofmescaline (the main psychoactive compound of thepeyote cactus, Lophophora williamsii) on the experienceof its users (including on himself). He observed that thenon-ordinary visual experiences were often character-ized by similar kinds of abstract geometric patterns,which he classified into four categories of form con-stants: (1) gratings, lattices,2 fretworks, filigrees, hon-eycombs, and checkerboards; (2) cobwebs; (3) tunnelsand funnels, alleys, cones, vessels; and (4) spirals.Kluver’s form constants have since been found tooccur in a variety of other kinds of altered states(Bressloff, Cowan, Golubitsky, Thomas, & Wiener,2001). Intriguingly, these form constants turned outto resemble many of the abstract motifs that are oftenassociated with prehistoric art from around theworld, including Paleolithic cave art in Europe. Theseinsights led to the provocative hypothesis that muchof the content of prehistoric art was inspired by pat-terns and visions seen during altered states of con-sciousness (e.g., Clottes & Lewis-Williams, 1998;Lewis-Williams, 2002; Lewis-Williams & Dowson,1988; Winkelman, 2002).

Of course, it still remains to be explained why theseparticular motifs were highly regarded by the artistsand how these people became artists capable of sym-bolic expression in the first place. According to Lewis-Williams (2002), the symbolic capacity was already pro-vided by the evolution of an appropriate mutation inthe brain, and the value of the particular content wasderived from the hierarchical power structures enforcedby shamans who had exclusive access to the visionaryexperiences. However, this hypothesis has several short-comings. While complex social stratification and strifewas clearly in evidence during the Neolithic period inEurope (Lewis-Williams & Pearce, 2005), this kind ofsocial class conflict may not be generalizable to earlierperiods or other parts of the world.

For instance, the Jomon culture of prehistoricJapan, one of the oldest pottery traditions in the worlddating back to around 16,500 BC, was the first culturein the Japanese archipelago to produce pottery deco-rated with abstract geometric patterns. This symbolicmaterial culture flourished for well over 10,000 years,producing large quantities of pieces with recognizableform constants (e.g., Figure 2). However, compared tolater symbolic material cultures in this region, theJomon period is remarkable for its long-term stability,and because of the absence of strong archaeologicalevidence for a complex social hierarchy, social strife,and intergroup warfare (Habu, 2004).

The hypothesis that the first symbolic practices werean outcome of a prehistoric class struggle is lackingsolid archaeological evidence. More importantly, evenif we could find an early prehistoric society that fitsLewis-Williams’ hypothesis, the existence of a rigidsocial hierarchy and restricted accessibility of alteredstates of consciousness by an elite class does not explainwhy such altered states were highly valued by thesepeople in the first place. If the ritualized alteration ofconsciousness and its symbolic expressions were to be atool of empowerment restricted to the elite, then thegeneral population must have already valued such

Figure 1. Illustration of a piece of red ochre with abstractgeometric incisions that were made on one side around73,000 years ago. This piece was found during excavations inBlombos Cave, South Africa. This and similar pieces from thesame location are currently the oldest known human expressionof abstract geometric patterns. Reproduced with kindpermission from Elsevier (Henshilwood et al., 2009).

Froese et al. 201

non-ordinary experiences; otherwise their restrictioncould hardly have served as an instrument of control. Inaddition, this hypothesis leaves any functional connec-tion of altered states of consciousness to the origin ofsymbolic practices, including their underlying neuraleffects, unexplored. It seems plausible that, at least undersome circumstances, a temporary alteration of normalsense-making can lead to heightened levels of imagina-tion and creativity (Dobkin de Rios & Janiger, 2003).

3 Neural mechanisms

Turing patterns consist of various geometric forms,including spots, traveling waves, grids, and spirals,which emerge out of the distributed activity of non-linear dynamical systems with local excitatory andsparse inhibitory connectivity. Some illustrative exam-ples derived from a simple Gray–Scott reaction–diffusion system are shown in Figure 3.

Simple spatiotemporal pattern formation was firstdescribed by Turing (1952) for the case of chemicalreaction–diffusion systems. Turing proposed that thesekinds of systems could help to explain morphogenesis,i.e., pattern formation during biological development.Since then, Turing patterns have been found to be ubi-quitous in a variety of biological systems (Goodwin,

2001). Regarding visually expressive phenomena, it hasbeen shown that the skin patterns of some animals canbe generated by underlying reaction–diffusion systems(Kondo & Miura, 2010). Given the similarity betweenTuring patterns and some of the motifs frequentlyfound in prehistoric art and experienced by subjects inaltered states of consciousness, it makes sense to inves-tigate whether the biological mechanisms underlyingthe production of these visual phenomena is amenableto an analysis in terms of Turing instabilities. Indeed,this proposal is in line with Kluver’s observation that‘‘the hallucination is [.] not a static process but adynamic process, the instability of which reflects aninstability in its conditions of origin’’ (cited in Bressloffet al., 2001).

3.1 The neural Turing mechanism

One of the originally proposed mechanisms for geo-metric hallucinations is that of a neural Turing mechan-ism, embodied in the Wilson–Cowan equations (H. R.Wilson & Cowan, 1973) and first mathematically ana-lyzed in terms of visual hallucinations by Ermentroutand Cowan (1979). This is the neural context for thefamous Turing Instability, which describes a reaction–diffusion process in which an ordered macroscopicstructure can emerge from local interactions under thenon-equilibrium state (Turing, 1952). Turing’s idea canbe readily transposed to the functioning of the nervoussystem. For example, action potential propagationalong a neuron’s axon can be directly described byreaction–diffusion equations, and reaction–diffusionequations are analogical to the Wilson–Cowan neuralnetwork equations (H. R. Wilson, 1999, pp. 267–268).We can think of the reaction component as the interac-tions between neuronal cells, and of the diffusion com-ponent as the spread of neural activity through localsynaptic connections. Similarly, the local structure ofneural interconnectivity dictates the type of emergentphenomena that can be produced. Neural networkmodels of geometric hallucinations have graduallyincorporated these empirical insights from neural anat-omy and physiology, including the spatial arrangementof different neuronal cell types.

The first model of geometric hallucinations to followthis idea, by Ermentrout and Cowan (1979), developedan isotropically connected two-layer neural network ofexcitatory and inhibitory neurons to represent the pri-mary visual cortex (V1). Increasing an excitability para-meter destabilized the rest state, bifurcating the system,and led to macroscopic spatial patterns emerging due tothe lateral interactions of negative feedback. Such ageneral description of a locally connected networkshows up the reaction–diffusion type properties of suchsystems, but they require specificity in parameters inorder to generate patterns. The main limitation of thisearly work is that they do not consider the neural

Figure 2. A deep pot covered with various spiral patterns; anearly artistic example of the spiral type of Kluver’s formconstants. This particular style of vessel, known as Moroiso, wasmade around 4000 BC by people of the Jomon culture(Kobayashi, 2004, p. 31), a relatively peaceful and non-stratifiedsociety of prehistoric Japan (Habu, 2004). (Photo taken by TomFroese at the Archaeological Museum of Kokugakuin University,Tokyo, Japan).


architecture of V1: the arrangement of neurons inhypercolumns and their properties as low-level featuredetectors. These are addressed in Bressloff et al.’s(2002) work, to be discussed later in this section.

3.2 The location of geometric hallucinations withinthe brain

Where exactly in the brain do such geometric hallucina-tions emanate from? Also, is it correct to say that thereis a single region that constitutes the generation of avisual pattern, knowing that the brain is a highly inter-connected dynamical system? Considering that geo-metric hallucinations are unaffected by the physicalmovement of the eyes means that such phenomenamust emanate from the endogenous activity of thebrain. Secondly, that we ‘‘see’’ such phenomena sup-ports the assumption that they must be dependent onthe brain’s visual sub-system.

All models of geometric hallucinations propose thatsuch patterns emerge from V1, the ‘‘first’’ region of thevisual cortex to receive visual input. This assumption issupported by fMRI scans that have shown that V1 issometimes activated during mental imagery, i.e., whenone imagines something visual in either an awake ordreaming state—even when they are not in an

externally coupled state (Miyashita, 1995), althoughthere is some variability in these results (Kosslyn &Thompson, 2003). Other results have pointed to theprimary visual cortex interacting with the hippocampusduring offline episodic memory consolidation; high-order replay of episodic memories occurs with quitespecific patterns of visual activity in V1 (Ji & Wilson,2007).

The visual cortex is located in the posterior regionof the brain’s occipital lobe. V1 takes information fromthe Lateral Geniculate Nucleus (LGN), a region of thethalamus, which in turn receives information from ret-inal ganglion cells that each receive information fromaround 5–100 photoreceptors. Anatomically, V1 is anarray of hypercolumns involved in the low-level pro-cessing of visual input, with feature detectors tuned tovisual aspects such as color, edges, contours, motion,etc. (S. W. Wilson, 1983). Between regions, there is aprogressive mapping to 2D cortical surfaces and all ofthe models to be discussed operate in such a dimension.After V1, higher-order processing of visual informationoccurs in roughly 30 identified regions within the visualcortex (Miyashita, 1995).

But it is presumptive to speak of these brain pro-cesses happening in some sort of projective sequence.For example, in cat brains it has been identified that

Figure 3. Examples of dynamic patterns exhibited by the Gray–Scott reaction–diffusion system under various parameters. Patternsare chosen as exemplars of various phenomena; see Pearson (1993) for a more systematic classification. (a) A spiraling pattern; (b) achaotic pattern of travelling waves; (c) a line pattern, whereby lines grow at the ends and then bend to fill space; (d) a labyrinthpattern; (e) a hole pattern; (f) a pattern of unstable spots, whose fluctuating population is maintained by a balance betweenreproduction and disintegration; (g) a stable spot pattern, whereby spots reproduce to fill empty space. Reproduced with kindpermission from Nathaniel Virgo (Virgo, 2011).

Froese et al. 203

retinal input accounts for about 7% of synaptic con-nections to relay cells in the LGN, therefore 93% ofinput is non-retinal, coming from other regions in thebrain (Van Horn, Erisxir, & Sherman, 2000). It is certainthat similar physiology exists within the human visualsystem and this shows that a great deal of dynamicfeedback processes must be occurring. Accordingly,models that do not account for this massive feedbackhave questionable biological plausibility.

3.3 The retinotopic map

In general, in current models of geometric hallucina-tions, an inverse retinotopic map is applied to neuralactivity in V1 in order to virtually project back intowhat would have been the causally responsible retinalspace under normal conditions. The resulting image isintended to enable us to see what the subject wouldvisually experience. The retinotopic map describes howretinal input projects onto the visual cortex. This phy-siological mapping has been studied extensively and, insimplified terms, is a conformal projection from retinalpolar coordinates into cortical rectangular coordinates.More space in V1 is assigned to the foveal region thanto the periphery and this is why our spatial detail ishighest at this focal point in our vision.

The reasoning principle underlying the applicationof an inverse mapping to the neural network models isthat under normal awake conditions, when a subject isobserving the external world, an undistorted view ofthe world enters the retina and then, for physiologicalreasons, gets distortedly mapped to V1. However, wenever consciously experience this distorted view of V1.Therefore, it is argued, there must be some subsequentinverting process such that, in some manner, phenom-ena generated in V1 are actually experienced in terms

of the original retinal frame of reference. This reason-ing has important implications for explaining geometricvisual hallucinations. For example, taking a uniformlytextured ‘‘image’’ (in V1) and going from rectangular(cortical) to polar (retinal) coordinates can generate atunnel like warping of the image. This is an effect com-mon in many geometric hallucinations and perhapsalso related to the tunnel seen in near death experiences(Blackmore & Troscianko, 1989).

However, the tunnel like warping effect of theinverse retinotopic mapping is not always fitting. Forexample, to accommodate lattice type phenomena thatdo not have a clear center of origin or convergence,Ermentrout and Cowan (1979) state that these mustoccur at the periphery of the visual field. On the otherhand, a cobweb with a clearly defined center is said toappear nearer the fovea (see Figure 4 for illustrativeexamples of these distinct kinds of form constants).

But this seems a rather odd prediction. If this modelis indeed biologically accurate then a large number ofgeometric patterns that do not exhibit a tunnel-likeappearance should be constrained to only appearing atthe periphery of our visual experience. As far as weknow, the extant psychological literature does not makeany mention of a strict division between such focal andperipheral types of hallucinatory visual patterns. Thus,although this is a prediction that still deserves to be fur-ther investigated experimentally, it is more likely anindication that the relationship between the retinotopicmapping and geometric hallucinations is less direct thanoriginally assumed by Ermentrout and Cowan.

Another shortcoming of using inverse retinotopicmapping to derive an image of the subject’s experienceis that it lacks consideration of the temporal dimensionof consciousness (Varela, 1999). Empirical researchsuggests that abnormal activity in the visual system

Figure 4. Illustration of typical form constants. (a) Cobweb; (b) lattice consisting of parallelograms; and (c) lattice consisting out ofhexagons. In contrast to the cobweb form, the lattice forms do not converge on a center point. Reproduced with kind permissionfrom Springer (Ermentrout and Cowan, 1979).


only becomes consciously experienced in the form ofgeometric hallucinations after more extensive recurrentprocessing (e.g., P. C. J. Taylor, Walsh, & Eimer, 2010),thus again implying a more indirect mapping betweenactivity in V1 and conscious visual experience. Moregenerally, we can also take issue with the phenomenol-ogy of mental imagery that is implied by the idea ofderiving the subject’s experience from a neural ‘‘image’’.Under closer phenomenological examination, the expe-rience of mental imagery is qualitatively different fromthe experience of seeing a 2D picture (Thompson, 2007).

3.4 The ubiquity of spiral and wave dynamics in thebrain

As stated earlier, most models assume that hallucina-tory phenomena are produced in V1 and then pursue adescription of their neural underpinnings. But spiraland wave (circular wavefront) dynamics seem to be aubiquitous feature of brains, not just in V1, and also ofbiologically realistic neural models. Huang et al. (2010)state that spiral waves which emanate from a centralpoint, as an emergent property, can organize and modu-late cortical population activity on the mesoscopic scaleand may contribute to both normal cortical processingand to pathological patterns of activity, e.g., epilepsy.Such wave activity is found to be robust when consid-ered within local excitatory interactions such as within acortical sheet (a locally connected 2D network). But thebrain has modulatory long-range connections that canchange the properties of local sub-networks that couldserve as regulatory protective mechanisms.

For example, Rulkov, Timofeev, and Bazhenov(2004) developed a set of computationally efficient yethighly realistic map-based neuron models to explorelarge-scale cortical networks and their dynamics. Inparticular, the spiking activity in different types of cor-tical pyramidal (PY) (fast spiking, regular spiking,intrinsically bursting) and interneurons (INs) was devel-oped. Experiments showed that wave properties weredictated by coupling parameters and that self-sustaineddynamics were a function dependent on synaptic inter-action between neurons, and not spontaneously firingcells. These findings support the idea of neural architec-ture being an important factor for the emergence ofgeometric patterns that are potentially experienced ashallucinations. That spontaneous waves of activitycould persist within the network indefinitely is alsointeresting with respect to the ideas of self-sustainabilityduring externally decoupled mental states, maintainingnetwork wide activity to prevent neuronal atrophy fromdisuse, or the brain’s default network—where self-sustained dynamics allows the brain to be in a primed,high-dimensional state.

Fohlmeister, Gerstner, Ritz, and van Hemmen(1995) investigated the type of spontaneous excitations

that can emerge from a realistic non-linear model of acortical sheet of noisy spiking neurons. Their modelincluded a drug parameter and they found that under anormal state of activity there was incoherent low-frequency firing. As they increased the excitation of thenetwork, beginning from random initial conditions,they categorized the following emergent properties: (1)stripes, (2) spirals—found to be very stable, (3) rings,and (4) collective bursts. Collectively these were termedfirst stage imagery, corresponding to the idea of formconstants. Second stage imagery, which goes beyondgeometric hallucinations to include complex imagery,without any doubt involves several other areas of thebrain, including memory systems (de Araujo et al.,2012), and was not modeled. Interestingly, once spiralsexist within the modeled neural system they cannot beextinguished, even if the drug parameter is adjustedinto the ranges to support categories 3 and 4. We notethat this kind of hysteresis can be useful during normalbrain functioning. Summarily, they consider form con-stants to be elementary excitations in the primary visualcortex and these spatiotemporal patterns involve thesame kind of mechanism as in experimentally observeddrug induced epilepsy (we will return to epilepsybelow).

Milton, Chu, and Cowan (1993) developed similarwork for generating spirals from a fixed excitation cen-ter that mimic cardiac arrhythmia. This also demon-strates how changes in neural parameters generate quitevaried macroscopic phenomena despite having a consis-tent local connectivity. In general, however, these mod-els do not generate all four categories of form constant.And despite some authors claiming that the ‘‘perfor-mance of a large network does not depend on details ofthe model once the neural essentials have been incorpo-rated’’ (Fohlmeister et al., 1995), other models indicatethat more of the characteristic form constants could beproduced if the exact structure of V1 was considered.

3.5 Modeling geometric hallucinations as a functionof V1 neural architecture

Bressloff et al. (2001, 2002) developed a model thataimed to connect geometric visual hallucinations withthe structure of V1. It is well known that V1 consists ofa number of neural circuits that are involved in low-level ‘‘image processing’’, responding preferentially tofeatures such as oriented edges and local contours, etc.Their work proposes that the specific spatial arrange-ment and local coupling of these feature detectors,along with the neural Turing mechanism combine togenerate form constants. According to their model,under normal conditions the visual system is in a stablestate, but the introduction of drugs destabilizes V1 andspontaneously generates cortical activity that reflectsthe underlying architecture of V1. The possible form

Froese et al. 205

constants of this model are specific linear combinationsof the eigenfunctions of the neural architecture, calledPlanforms, which possess both spatial and temporalproperties. Since V1 must avoid falling into these statesduring normal visual processing, if the influences of theretinal input are to have any effect in driving the sys-tem, this places strong constraints on the evolution ofthe visual cortex architecture, for example in terms ofthe sparsity of long-range inhibitory connectivity(Butler et al., 2012).

Bressloff et al.’s work represents perhaps the mostcomplete description of a neural mechanism for gener-ating geometric hallucinations, however, it is not capa-ble of generating all of Kluver’s form constants anddoes not seem to account for the ubiquitous spiral wavedynamics of the brain, which are realized by the neuraldynamics in a model such as Folhmeister et al.’s. Toaddress this, it would be interesting to incorporate amore biologically plausible, spiking neural model.Secondly, the authors only considered oriented edgesand local contours in their architecture; what genera-tive power would the modeling of other feature maps,such as spatial frequency, binocular disparity, motion,and color provide?

Furthermore, none of the models seems to accountfor the entirety of geometric hallucinations that arecommonly reported in the archeological and anthropo-logical literature, for example as described by Lewis-Williams and colleagues (Lewis-Williams & Dowson,1988). Future work along the lines of Bressloff et al.’smodel may someday show that the intrinsic propertiesof V1 are enough to describe all geometric hallucina-tory phenomena, but given the large variety of pat-terns, this is unlikely. It is interesting to consider whatother brain mechanisms could contribute to their for-mation, such as the large number of feedback pathwaysthat exist within the visual system.3

3.6 The role of neural feedback in visual processing

Complex and varied experimental results on the func-tion of feedback pathways have been found, but it isgenerally considered that higher-level cortical regionsare involved in large-scale perceptual integration whichcombine with the fine-scale resolution provided by theearlier visual regions (Sillito, Cudeiro, & Jones, 2006).Feedback is involved in modulating perception; such asfigure ground separation, suppressing areas of lowinformation to generate a sparse codification of thevisual signal, coordination and tuning of visual featuredetectors, and attention (Williams et al., 2008). Back-projecting pathways are therefore thought to facilitate,inhibit, and synchronize neural activity (Przybyszewski,1998). Under the influence of drugs, or in an alteredstate of mind induced by other means, these pathwayscould all be involved in moving the visual system

toward one of the stable Planforms as proposed inBressloff’s model.

Secondly and more generally, we know that V1 andother regions of the visual pathway show neural activityduring imagination (Ganis, Thompson, & Kosslyn,2005), which shows that they can be integrated intohigher-level processing in the absence of external input.Could a feedback loop between these brain areas gener-ate geometric hallucinations? This is much like the ideaof a video-feedback system, where a video camera ispositioned to record the same video monitor to which itis connected (Crutchfield, 1984, 1988). Despite a videofeedback system being physically different, a conceptualanalogy can be formed by making firstly a connectionbetween the neural Turing mechanism and modelingthe video-feedback as a reaction–diffusion system(Crutchfield, 1988), and secondly with the known abun-dance of feedback pathways within the brain. The largenumber of geometric patterns that can be produced bysuch physical video feedback systems, such as logarith-mic spirals and phyllotaxis, lends some credence to thisnovel hypothesis.

However, what could constitute a neural visual feed-back mechanism for geometric hallucinations is stillunknown. It seems plausible that somehow the abun-dance of feedback loops could play this role, especiallywhen considering that at some point, during an alteredstate of mind, geometric hallucinations may be com-bined, modulated by, and reinterpreted with hallucina-tions from memory and iconic imagery that must begenerated elsewhere within the cerebral cortex (deAraujo et al., 2012).

Another pathway that could be involved is throughthe superior colliculus (SC)—a major region of the ver-tebrate midbrain that forms an important part of thevisual system, and which is considered to play a criticalrole in the development of blind sight (Takaura,Yoshida, & Isa, 2011). The mutual interaction betweenthe primary visual cortex and this subcortical neuralstructure, which is also involved in visual processing,could open new possibilities for explaining geometrichallucinations in the absence of external visual stimuli.In addition, the SC’s role in multisensory integration(Hall & Moschovakis, 2004) might help to explain howvisual hallucinations become integrated with hallucina-tions in other modalities, including synesthesia-likeeffects.

3.7 Beyond structural isomorphism

This overview of the putative neural mechanismsunderlying geometric hallucinations points to the keycomponents being the architecture in V1 supporting aneural Turing mechanism along with the retinotopicmapping, the ubiquity of spiral and wave dynamicswithin the brain, and widespread and long-rangingfeedback connections. But despite these insights, the


full range of reported geometric hallucinations has yetto be accommodated by a single mathematical neuralnetwork model. The incorporation of additional phy-siological detail may increase the range of geometricphenomena that can be modeled, especially long-rangefeedback connectivity. We must consider the neuralprojections into and out of V1, such as its direct con-nection back to the LGN, to the other 30 or so visualprocessing regions, and the rest of the brain.

Visual hallucinations can be induced in a wide num-ber of ways, but all models make use of a single drugparameter that affects the entire system. Yet how doesthe experienced pattern differ depending on the routetaken to such a state of mind? Dietrich (2003) proposesthat most altered states of mind (drug induced or other-wise) involve a transient decrease in prefrontal cortexactivity, and that the difference between types of expe-rience arises from how the induction method affects thevarious prefrontal neural circuits. Connecting this func-tion to models of geometric hallucinations and to thetypes of patterns specific to certain altered states couldmake for some interesting predictions. We can alsowonder at what type of widespread effects the prefron-tal cortex can have on the whole visual cortex and therest of the brain, not just in V1.

It also remains to be seen whether Turing instabil-ities in the rest of the brain, including the peripheralnervous system, play a role in generating geometric hal-lucinations. Indeed, because of the generality of thisdynamical phenomenon, its potential role may not belimited to electrical interactions between neurons,either. We know that Turing patterns are crucial for anumber of other biological processes, including in thechemical domain (Goodwin, 2001). Therefore, morerealistic models of geometric hallucinations might wantto include details of the physical and chemical pro-cesses of the central and peripheral nervous system,e.g., reaction and diffusion of various neurotransmit-ters, vitreous liquids inside the eye, oily residues on thesurface of the cornea, etc., which could give rise toTuring patterns under disinhibited conditions, and thusinfluence the outcome of visual processing. As a simpleproof of concept, it has been shown that the dynamicsof a Gray–Scott reaction–diffusion system (see patternsin Figure 3) can be used in order to control a mobilerobot, giving rise to shape discrimination and memory(Dale & Husbands, 2010).

But as we incorporate these additional features intothe models, it may also turn out that the hypothesis ofa strict structural isomorphism between patterns ofneural activity in V1 and geometric hallucinations willbecome less tenable. As a case in point, the function ofthe inverse retinotopic map can be called into questionwhen considering that it implies that many of thereported geometric hallucinations have to appear extra-foveally. Accordingly, rather than looking for a directone-to-one structural mapping between patterns of

neural activity and patterns of hallucinatory experi-ence, it may be more productive to consider other func-tionally relevant properties of such neural patternsthan their precise spatiotemporal arrangement alone.

For example, all modeled neural patterns showproperties of robustness and self-sustainability. In themodels of Ermentrout and Cowan (1979) and Bressloffet al. (2001), patterns appear as the stable configura-tions of the system after bifurcating from a low activitystate. The neural model of Rulkov et al. (2004) displaysself-sustained wave dynamics that can spontaneouslyemerge from low-level random neural activity. Asalready mentioned, in the model of Fohlmeister et al.(1995), once spiral waves emerge they cannot bedestroyed, even when increasing a drug parameter intothe range where other geometric phenomena usuallyappear. This case of apparent irreversibility is probablythe strongest example of robustness and self-sustainability in the surveyed models, and it is sugges-tive as a biological explanation for the fact that spiralsare one of the most prevalent motives in prehistoricart.

How could this property of robust self-maintenancerelate to the hallucinatory experiences? A striking fea-ture that differentiates geometric hallucinations fromother visual experiences is that they are generatedintrinsically when the subject has been decoupled fromits environment. The patterns form irrespective of ourlifetime learned memories. Indeed, they could be con-sidered internally directed perceptual experiences, sinceif the proposed models hold true, they are directlyformed from the actual biological structure of thevisual system. We are said to have a strange subjectiveexperience of looking into oneself, where the patternswe see directly expose the underlying operation of ourbrains. At the same time, the robustness of the underly-ing neural patterns also reinforces this decoupling fromthe environment. In the next section, we argue thatunder some conditions such a neural mechanism ofdecoupling could be adaptive. Humans are adept at notonly understanding patterns from external sensoryinput, but importantly, we have a powerful faculty foractive generation—be it for purposes such as predic-tion, imagination or playing.

4 Steps toward a new approach to theorigins of symbolic material culture

One promising approach for systematically evaluatingthe relationship between brain functioning and humanexperience is known as neurophenomenology, a methodthat was proposed by Varela (1996) as part of the enac-tive approach, and which has subsequently been elabo-rated by others (e.g., Petitmengin, Navarro, & Le VanQuyen, 2007; Thompson, Lutz, & Cosmelli, 2005). Thebasic idea of neurophenomenology is to relate the

Froese et al. 207

empirical data of neuroscience and verbal descriptionsof first-person experience in a mutually informing andenriching manner. In this way both third-person andfirst-person aspects of a phenomenon are taken intoaccount, but without privileging one domain over theother. While the aim is to find a formal model at whichthe dynamics of the two domains can be integrated, thisdoes not entail a necessity of structural isomorphism.Instead, there is a mutual braiding of structural con-straints, which links the activity of the two domainsinto one unfolding process (Varela, 1999).

In addition, because perception is conceived of as arelational process of sense-making that is extended overthe entire brain–body–environment as a whole, theexternal–internal distinction becomes relativized (Lenay& Steiner, 2010). That is, the perceptual experience ofan environment as being external is itself an outcome ofsense-making, and therefore becomes potentially rever-sible and dissolvable as the conditions of sense-makingare modified under unusual conditions. In the case ofgeometric hallucinations, it could therefore be that weare in fact dealing with an abnormally oriented percep-tual process, where the object of perception is an aspectof the nervous system itself. This is surely a strangehypothesis, but the consideration of perceptual media-tion opens up a new perspective on the problem of themismatch between the impoverished neural patternsand the rich geometric experiences. A similar mismatchhas been observed to hold between the impoverishedretinal input and our rich visual experience of the world(Noe, 2002). Accordingly, on this alternative view, themain shortcoming of the current neural network modelsis not an insufficiently detailed structural isomorphism,but rather a failure to take the enriching effects ofsense-making into account.4

4.1 Neurophenomenology of value

How might neurophenomenology address the corre-spondence between spatial patterns in the cultural,experiential and neurobiological domains? The firstthing to note is that the artistic patterns under consider-ation form an extremely limited set compared to thetotal set of all possible patterns that could have beencreated. This universal selection bias is something thatmust be explained, but which has so far remained mys-terious. It is not sufficient to note the correlation withvisual experiences during altered states of conscious-ness. Even if we accept that internally oriented visualperception is possible during altered states of conscious-ness, this does not in itself explain why those perceptualexperiences are more frequently expressed artistically.Why should these particular patterns be preferred overother kinds of ordinary and non-ordinary visual experi-ences? This selective bias must therefore be accountedfor in terms of the value these patterns have had for theartists who made them.

This consideration places a significant constraint onour explanation from the side of cultural practices,because the artists must have valued their altered visualexperiences. Some researchers have tried to address thisconstraint by appealing to social practices. For exam-ple, Lewis-Williams (2002) argued that the value of theexperiences derives from their value as a tool to enforcesocial hierarchies by restricting access to elite individu-als via prohibition. But this only shifts the originalproblem without resolving it, because we must thenexplain why people would value these experiences asworthy of restriction to elites.

One promising alternative is to assume that whenthese visual patterns are seen during altered states ofconsciousness they are directly experienced as highlycharged with significance. In other words, the patternsare directly perceived as somehow meaningful andthereby offer themselves as salient motifs for use inrituals. This possibility seems to accord with verbalreports about the range of emotional effects that can beelicited by a variety of non-ordinary experiences (e.g.,Dobkin de Rios & Janiger, 2003; Huxley, 1956;Shanon, 2002; Strassman, 2001). This appeal to anintrinsic value places constraints on the underlyingneural dynamics. We therefore need to discuss howvalue could be realized in neurobiology, and then wedetermine whether this neurological mechanism is com-patible with the neurological mechanism underlying theexperience of the most common forms of geometricphenomena.

4.2 Sketches of an enactive neurobiological accountof value

In the history of cognitive science, the question of valuehas been a profoundly puzzling one. The computa-tional theory of mind holds that cognition is rule-basedmanipulation of abstract symbols, as epitomized by thedigital computer program. However, this logical syntaxleaves the concrete meaning of the symbols unac-counted for; they are simply empty placeholders. Thisproblem of meaning has resulted in major stumblingblocks in theory of mind, as well as in the practice ofAI and robotics, such as the symbol grounding prob-lem and the frame problem. The enactive critique ofthese theoretical and practical problems is centered onthe grounding of value in the self-maintenance ofautonomous forms of biological organization (for areview, see Froese & Ziemke, 2009).

Most importantly, the enactive approach tries toprovide a viable alternative by defining value in opera-tional terms. A good starting point is the work by DiPaolo, Rohde and De Jaegher (2010, p. 48): ‘‘We pro-pose to define value as the extent to which a situationaffects the viability of a self-sustaining and precariousnetwork of processes that generates an identity.’’ There


are a number of entangled concepts implicated in thisdefinition, which cannot be unpacked in detail here.We briefly summarize the main ideas. To start with,the network’s identity is not a reified entity; it can bedefined as operational closure, i.e., the mutual depen-dence of the processes of the network on each other fortheir existence. This co-dependence provides them withan internal relationship that integrates them with eachother and that does not depend on a unifying distinc-tion by an external observer. And this co-dependencealso makes the network’s identity precarious, since itsexistence is not externally guaranteed but must be self-sustained by its own ongoing activity. In other words,the network is only viable within certain boundary con-ditions. The value of an event is therefore definedin relation to its impact on these conditions, i.e.,whether it contributes to the self-sustaining of the iden-tity or not.

The paradigmatic example of this kind of value-generating system is the minimal living system (Weber& Varela, 2002). But the same kind of organization canalso be found in the nervous system, where we have thetemporary emergence of cell assemblies (Varela,Lachaux, Rodriguez, & Martinerie, 2001). The firingpatterns of some neurons (or of collections of neurons)can become entrained with each other, thereby consti-tuting a precarious, self-sustaining network of pro-cesses, which in this case is a neurocognitive identity(Varela, 1997). This kind of neural pattern is thereforea suitable candidate for a process of value-generation.

Now if we consider the models of the neural mechan-isms underlying some of the common geometric halluci-nations, we also find the same kind of self-sustainingcell assembly. Moreover, it is likely that the assembly’srobustness is enhanced considerably, because activity inthe visual brain regions during altered states is generallyassumed to be disinhibited. In addition, under theseconditions the sensory input from external events nolonger has the power to modulate the cell assembly’sboundary conditions, thus enhancing its viability. Inthe most extreme cases, this kind of autonomous, self-sustaining network can manifest itself in a complete lossof sensorimotor interaction. For example, it has longbeen known that seeing flickering light at certain fre-quencies can induce neural entrainment and geometri-cal hallucinations (for recent work in this area, see, e.g.,Billock & Tsou, 2007), but when the brain as a wholebecomes too entrained with those frequencies it cancause an epileptic fit. According to the enactive accountof value, we would expect these kinds of geometricexperiences to be experienced as significant, and thenon-ordinary experiences encountered before or duringepileptic fits as intensely meaningful.5

Verbal reports of people with epilepsy confirm thishypothesis. Sometimes the seizures leave people uncon-scious, but when they do have experiences during theprodromal phase and at the onset of the epileptic fit

they are complex in many ways (Le Van Quyen, 2010;Petitmengin et al., 2007). What is particularly remark-able is the reported vividness or intensity of theseexperiences. The content of the experiences depends onthe person and the precise brain region of the epilepticdischarge, but commonalities can be identified.Consider, for example the case of emotional experience:

Special emotional auras consist of an attack of anguishand terror that suddenly takes over the consciousness withsuch intensity that the subject has the impression she or heis losing control of the situation, which will have a terribleend, perhaps madness or even death. In other emotionalauras, the experience consists of a sudden state of joy withno apparent cause, and it takes over the consciousnesspassively for a few short moments, filling it with awe andstrangeness. (Le Van Quyen, 2010, pp. 247–248)

Clearly, neurophenomenology is currently notadvanced enough to explain the particular content ofthese experiences, but it does successfully explain whythe experiences are characterized by such an intenselyfelt significance. Generalizing this insight to the experi-ences of geometric hallucinations, we can predict thatthese, too, are felt with an increased intensity of signifi-cance, as long as self-sustaining processes of neuralactivity generate them. In other words, intrinsic value isbuilt into the experience of some altered states of con-sciousness from the start. The ambivalence of the non-ordinary experiences reported by Le Van Quyen, i.e.,whether they are experienced as good or bad, is notspecific to altered states of consciousness induced byepilepsy, but is characteristic of altered states more gen-erally (Huxley, 1956). Perhaps this inherent polarity ofconsciousness alteration, which Huxley famously con-trasted as scenarios of ‘‘heaven’’ and ‘‘hell’’, has to dowith the fact that the organism is a whole meshwork ofself-sustaining processes (Varela, 1991), and the intrin-sic value of a specific self-sustaining neural process hasto be balanced against the needs of the other self-sustaining processes of the organism as a whole.

4.3 Ideas for an enactive neurobiological accountof imagination

The problem of explaining the origins of the humanimagination is related to the transformation of basicsense-making of the here and now to the enaction ofthe absent and imaginary (Froese, 2012). We speculatethat one manner of realizing this transformation is tomediate the organism’s sensorimotor interactions insuch a way that the autonomous dynamics of the ner-vous system temporarily become decoupled from theenvironment. Abstract forms of cognition are likely tobe facilitated during such highly introverted states ofconsciousness, which would also help to explain theirprehistoric origin.

Froese et al. 209

As we have already indicated, a basic dynamicneural mechanism for realizing a temporary mode ofenvironmental decoupling can be derived from the exis-tence of self-sustaining processes of neural activity.These neural processes might not always be easily iden-tifiable in terms of spatial patterns, especially when weare considering parts of the nervous system that do notexhibit such a regular spatial embedding as the retinaor V1.6 Nevertheless, it is helpful to think of theseneural processes in terms of an analogy with chemicalreaction–diffusion systems, which are formally closelyrelated to the activation–inhibition models of the visualsystem (H. R. Wilson, 1999, pp. 267–268). Chemicalreaction–diffusion systems are a well-researched area,both formally and empirically, and therefore provide auseful source of inspiration. For example, it has beendemonstrated that chemical Turing patterns, and espe-cially spiral patterns, are highly robust against the influ-ence of external perturbations such as attacks byparasitic chemical species (Boerlijst & Hogeweg, 1991).Accordingly, it seems plausible that the disinhibition ofthe visual cortex during altered states, perhaps via top-down projections from prefrontal cortex to the pathwayfrom retina to V1 regions, could lead to Turing patternsof neural activity in V1 that are robust against perturba-tions from visual sensory input, thereby effectively shut-ting down external visual influence (Butler et al., 2012).

The analogy with chemical reaction–diffusion sys-tems also suggests that some of these neural Turing pat-terns may even exhibit their own self-individuatedspatiotemporal identity, as well as the ability to engagein adaptive interactions with other emergent processesand structures. These kinds of emergent behaviors havebeen demonstrated to occur in chemical reaction–diffusion systems, including self-moving droplets withtail structures (Froese, Virgo, & Ikegami, in press). It istherefore tempting to speculate that some neural Turingpatterns can adapt to, and perhaps even make sense of,the brain’s own activity and structures during alteredstates of consciousness. Could this help to explain thecommonly reported transition from purely geometrichallucinations to iconic hallucinations that appear toexhibit their own agency (Lewis-Williams, 2002)?

Future work in systems neuroscience needs to verifyif any of these possibilities are borne out by the empiri-cal evidence. Nevertheless, an enactive approach iscompatible with current models of the neural mechan-isms underlying geometric hallucinations, and alreadyhas more explanatory power than merely assuming adirect structural isomorphism between visual experi-ence and brain activity: (1) it allows for the possibilitythat the patterns of non-ordinary visual experience aremore varied and richly detailed than the underlyingneural patterns; (2) it can help to explain why these pat-terns were taken to be highly significant; and (3) it sug-gests that there may have been a functional linkbetween the ritualistic alteration of consciousness and

the origin of abstract cognition and symbolic practicesin terms of temporary environmental decoupling ofneural processing.

5 Conclusion

We have criticized some of the current theories thathave been proposed to explain the origins of symbolicpractices, especially as exemplified by Lewis-Williams’account. We have argued that Lewis-Williams’ insis-tence on social conflict and class struggle as the pri-mary driving force behind changes in the first symbolicpractices is not compelling for a variety of reasons.Nevertheless, his claim that the cultivation of alteredstates of consciousness was involved in the origins ofthe first symbolic material cultures may still turn out tobe correct, although for alternative reasons which hefails to consider.

On the basis of a review of current mathematicalmodels of neural mechanisms underlying geometric hal-lucinations, we proposed a neurophenomenologicalapproach that emphasizes the enactive account ofsense-making and value. In particular, by consideringsome kinds of altered state of consciousness as largelyinternally mediated forms of perceptual sense-making,we can better account for the richness of non-ordinaryexperiences. Furthermore, the value generated by self-sustaining cell assemblies may help to explain theselective bias of the first artists for the kinds of geometricpatterns that are typically experienced during thesealtered states of consciousness. The enactive approachalso supports the idea that such altered states could havesignificantly influenced the operation of the nervous sys-tem, especially by temporarily decoupling the autono-mous activity of the brain from the usual environmentalinfluences. This switch from immediate sensorimotorsense-making, which is normally directed toward theexternal here and now, to a more internally mediated,decoupled sense-making of mental and bodily structurescould thereby have facilitated the creation and diversifi-cation of abstract cognition and symbolic practices.

Notes

1. Some researchers prefer the hypothesis that these prehis-toric patterns are highly abstract representations of nor-mal perceptual experience rather than direct expressionsof abnormal hallucinatory experience. For example,spiral patterns could be derived from ‘‘swirling waterflows, swirling winds, winding stems of vines and windingsnakes’’ (Takaki & Ueda, 2007, p. 133). However, thishypothesis presupposes a process of geometric abstrac-tion. Moreover, other common kinds of prehistoric pat-terns are not so readily found in nature. Mostimportantly, even if all of the patterns could beabstracted from environmental phenomena, the problemof explaining the cross-cultural significance of these pat-terns would remain the same.


2. Note that the engraved pattern shown in Figure 1 can beinterpreted as a lattice pattern made of triangles (i.e., oneof Kluver’s form constants).

3. In this respect it is suggestive that even the hallucinatoryexperience of small spots of light (phosphenes), inducedby applying transcranial magnetic stimulation (TMS)over the human visual cortex, apparently cannot bereduced to a local neural mechanism, but arises only aftermore widespread recurrent processing (P. C. J. Taylor etal., 2010). A distribution across multiple neural substratesalso helps to explain how such simple geometric halluci-nations can be transformed into full-blown iconic halluci-nations, for example under certain kinds of pathologicalconditions (J.-P. Taylor et al., 2011).

4. This shift in explanatory approach helps us to resolve oneof the major outstanding puzzles of the current neu-roscience of geometric hallucinations, namely the appar-

ent mismatch between the resolution of neural Turingpatterns and visually experienced patterns. However, it isnot a solution to the explanatory gap of the mind-bodyproblem as such. We thank Julien Hubert for helping usto clarify this important difference.

5. Strictly speaking, the enactive approach only predicts thatthere is value for the self-sustaining neural process itself.More remains to be said about how this value alsobecomes significant from the perspective of the self-sustaining organism as a whole. But considering thatneural processes are constitutive of the whole organism(Varela, 1991), it seems reasonable to assume that theirvalues are also constitutive of the values of the wholeorganism. Nevertheless, future work on an enactive the-ory of the personal self is needed in order to better con-nect these distinct levels of description. Thanks toNathaniel Virgo for highlighting this issue.

6. Thanks to Chris Buckley for pressing this point duringseveral discussions.

Acknowledgments

We thank Chris Buckley, daboo, Julien Hubert, GuillaumeDumas, Etienne Roesch, Nathaniel Virgo and one anon-ymous reviewer for their thoughtful comments on an earlierdraft of this paper. Many ideas of this paper took shape dur-ing several seminars of the Ikegami Laboratory, and theyhave benefited from generous audience feedback.

Funding

Tom Froese and Alexander Woodward were each financiallysupported by a Grant-in-Aid of the Japanese Society for thePromotion of Science (JSPS) during the initial phase of thiswork.

References

Billock, V. A., & Tsou, B. H. (2007). Neural interactions

between flicker-induced self-organized visual hallucina-

tions and physical stimuli. Proceedings of the National

Academy of Sciences of the USA, 104(20), 8490–8495.Blackmore, S. J., & Troscianko, T. S. (1989). The physiology

of the tunnel. Journal of Near-Death Studies, 8(1), 15–28.Boerlijst, M. C., & Hogeweg, P. (1991). Spiral wave structure

in pre-biotic evolution: Hypercycles stable against para-

sites. Physica D: Nonlinear Phenomena, 48, 17–28.

Bressloff, P. C., Cowan, J. D., Golubitsky, M., Thomas, P. J.,& Wiener, M. C. (2001). Geometric visual hallucinations,

Euclidean symmetry and the functional architecture of stri-ate cortex. Philosophical Transactions of the Royal Society

B: Biological Sciences, 356, 299–330.Bressloff, P. C., Cowan, J. D., Golubitsky, M., Thomas, P. J.,

& Wiener, M. C. (2002). What geometric visual hallucina-

tions tell us about the visual cortex. Neural Computation,

14, 473–491.Butler, T. C., Benayoun, M., Wallace, E., von Drongelen, W.,

Goldenfeld, N., & Cowan, J. D. (2012). Evolutionary con-

straints on visual cortex architecture from the dynamics ofhallucinations. Proceedings of the National Academy of

Sciences of the USA, 109(2), 606–609.Clottes, J., & Lewis-Williams, D. (1998). The shamans of pre-

history: Trance and magic in the painted caves. New York,

NY: Harry N. Abrams.Crutchfield, J. P. (1984). Space-time dynamics in video feed-

back. Physica D: Nonlinear Phenomena, 10(1–2), 229–245.Crutchfield, J. P. (1988). Spatio-temporal complexity in non-

linear image processing. IEEE Transactions on Circuits and

Systems, 35(7), 770–780.Dale, K., & Husbands, P. (2010). The evolution of reaction–

diffusion controllers for minimally cognitive agents. Artifi-

cial Life, 16, 1–19.Darwin, C. (1871). The descent of man, and selection in relation

to sex. London, UK: John Murray.de Araujo, D. B., Ribeiro, S., Cecchi, G. A., Carvalho, F.

M., Sanchez, T. A., Pinto, J. P., de Martinis, B. S., Crippa,J. A., Hallak, J. E. C., & Santos, A. C. (2012). Seeing with

the eyes shut: Neural basis of enhanced imageryfollowing Ayahuasca ingestion. Human Brain Mapping,

33, 2550–2560.De Jaegher, H., & Froese, T. (2009). On the role of social

interaction in individual agency. Adaptive Behavior, 17(5),444–460.

Deacon, T. W. (1997). The symbolic species: The co-evolution

of language and the brain. New York, NY: W. W. Norton& Company, Inc.

Di Paolo, E. A., Rohde, M., & De Jaegher, H. (2010). Hori-

zons for the enactive mind: Values, social interaction, andplay. In J. Stewart, O. Gapenne & E. A. Di Paolo (Eds.),

Enaction: Toward a new paradigm for cognitive science (pp.33–87). Cambridge, MA: The MIT Press.

Dietrich, A. (2003). Functional neuroanatomy of alteredstates of consciousness: The transient hypofrontality

hypothesis. Consciousness and Cognition, 12, 231–156.Dobkin de Rios, M., & Janiger, O. (2003). LSD, Spirituality,

and the creative process. Rochester, VT: Park Street Press.Ermentrout, G. B., & Cowan, J. D. (1979). A mathematical

theory of visual hallucination patterns. Biological Cyber-netics, 34, 137–150.

Fohlmeister, C., Gerstner, W., Ritz, R., & van Hemmen, J.

L. (1995). Spontaneous excitations in the visual cortex:Stripes, spirals, rings, and collective bursts. Neural Compu-

tation, 7, 905–914.Froese, T. (2012). From adaptive behavior to human

cognition: A review of Enaction. Adaptive Behavior, 20(3),209–221.

Froese, T. (in press). Altered states and the prehistoric rituali-

zation of the modern human mind. In C. Adams, A.Waldstein, B. Sessa, D. Luke & D. King (Eds.), Breaking

Froese et al. 211

convention: Essays on psychedelic consciousness. London,UK: Strange Attractor Press.

Froese, T., & Di Paolo, E. A. (2009). Sociality and the life-

mind continuity thesis. Phenomenology and the Cognitive

Sciences, 8(4), 439–463.Froese, T., Virgo, N., & Ikegami, T. (in press). Motility at the

origin of life: Its characterization and a model. Artificial Life.Froese, T., & Ziemke, T. (2009). Enactive artificial intelli-

gence: Investigating the systemic organization of life and

mind. Artificial Intelligence, 173(3–4), 366–500.Ganis, G., Thompson, W. L., & Kosslyn, S. M. (2005). Brain

areas underlying visual mental imagery and visual percep-

tion: an fMRI study. Cognitive Brain Research, 20, 226–241.Goodwin, B. (2001). How the leopard changed its spots: The

evolution of complexity. Princeton, NJ: Princeton Univer-

sity Press.Habu, J. (2004). Ancient Jomon of Japan. Cambridge, UK:

Cambridge University Press.Hall, W. C., & Moschovakis, A. (Eds.). (2004). The superior

colliculus: New approaches for studying sensorimotor inte-

gration. New York, NY: CRC Press.Harvey, I. (2008). Misrepresentations. In S. Bullock, J. Noble,

R. A. Watson & M. A. Bedau (Eds.), Artificial Life XI:

Proceedings of the Eleventh International Conference on the

Simulation and Synthesis of Living Systems (pp. 227–233).

Cambridge, MA: The MIT Press.Henshilwood, C. S., & d’Errico, F. (2011a). Middle Stone

Age engravings and their significance to the debate on the

emergence of symbolic material culture. In C. S. Henshil-wood & F. d’Errico (Eds.), Homo symbolicus: The dawn of

language, imagination and spirituality (pp. 75–96). Amster-dam, Netherlands: John Benjamins Publishing Company.

Henshilwood, C. S., & d’Errico, F. (Eds.). (2011b). Homo

symbolicus: The dawn of language, imagination and spiri-

tuality. Amsterdam, Netherlands: John Benjamins Pub-lishing Company.

Henshilwood, C. S., d’Errico, F., & Watts, I. (2009). Engraved

ochres from the Middle Stone Age levels at Blombos Cave,South Africa. Journal of Human Evolution, 57, 27–47.

Huang, X., Xu, W., Liang, J., Takagaki, K., Gao, X., & Wu,

J.-y. (2010). Spiral wave dynamics in neocortex. Neuron,

68, 978–990.Hutchins, E. (2010). Enaction, imagination, and insight. In J.

Stewart, O. Gapenne & E. A. Di Paolo (Eds.), Enaction:Toward a new paradigm for cognitive science (pp. 425–450).

Cambridge, MA: The MIT Press.Huxley, A. (1956). The doors of perception and heaven and hell.

New York, NY: Harper & Brothers.Ji, D., & Wilson, M. A. (2007). Coordinated memory replay

in the visual cortex and hippocampus during sleep. Nature

Neuroscience, 10(1), 100–107.Kluver, H. (1967).Mescal and the mechanisms of hallucination.

Chicago, IL: The University of Chicago Press.

Kobayashi, T. (2004). Jomon reflections: Forager life and cul-

ture in the prehistoric Japanese archipelago. Oxford, UK:Oxbow Books.

Kondo, S., & Miura, T. (2010). Reaction–diffusion model as

a framework for understanding biological pattern forma-tion. Science, 329, 1616–1620.

Kosslyn, S. M., & Thompson, W. L. (2003). When is early

visual cortex activated during visual mental imager? Psy-

chological Bulletin, 129(5), 723–746.

Laughlin, C. D., & Throop, C. J. (2009). Husserlian medita-tions and anthropological reflections: Toward a cultural

neurophenomenology of experience and reality. Anthropol-ogy of Consciousness, 20(2), 130–170.

Le Van Quyen, M. (2010). Neurodynamics and phenomenol-

ogy in mutual enlightenment: The example of the epilepticaura. In J. Stewart, O. Gapenne & E. A. Di Paolo (Eds.),

Enaction: Toward a new paradigm for cognitive science (pp.245–266). Cambridge, MA: The MIT Press.

Lenay, C., & Steiner, P. (2010). Beyond the internalism/

externalism debate: The constitution of the space of per-ception. Consciousness and Cognition, 19, 938–952.

Lewis-Williams, D. (2002). The mind in the cave: Conscious-

ness and the origins of art. London, UK: Thames &Hudson.

Lewis-Williams, D., & Dowson, T. A. (1988). The signs of alltimes: Entoptic phenomena in Upper Paleolithic art. Cur-

rent Anthropology, 29(2), 201–245.Lewis-Williams, D., & Pearce, D. (2005). Inside the Neolithic

mind: Consciousness, cosmos and the realm of the gods.

London, UK: Thames & Hudson.McGann, M. (2007). Enactive theorists do it on purpose:

Toward an enactive account of goals and goal-directedness.

Phenomenology and the Cognitive Sciences, 6(4), 463–483.Milton, J. G., Chu, P. H., & Cowan, J. D. (1993). Spiral

waves in integrate-and-fire neural networks. In S. J. Han-

son, J. D. Cowan & C. L. Giles (Eds.), Advances in neural

information processing systems 5 (pp. 1001–1006). San

Mateo, CA: Morgan Kaufmann Publishers.Miyashita, Y. (1995). How the brain creates imagery: Projection

to primary visual cortex. Science, 268(5218), 1719–1720.Newell, A., & Simon, H. A. (1976). Computer science as

empirical enquiry: Symbols and search. Communications of

the Association for Computing Machinery, 19(3), 113–126.Noe, A. (2002). Is the visual world a grand illusion? Journal of

Consciousness Studies, 9(5–6), 1–12.Pearson, J. E. (1993). Complex patterns in a simple system.

Science, 261(5118), 189–192.Petitmengin, C., Navarro, V., & Le Van Quyen, M. (2007).

Anticipating seizure: Pre-reflective experience at the center

of neuro-phenomenology. Consciousness and Cognition,

16(3), 746–764.Przybyszewski, A. W. (1998). Vision: Does top-down process-

ing help us to see? Current Biology, 8, R135–E139.Rulkov, N. F., Timofeev, I., & Bazhenov, M. (2004). Oscilla-

tions in large-scale cortical networks: Map-based model.

Journal of Computational Neuroscience, 17, 203–223.Shanon, B. (2002). The antipodes of the mind: Charting the

phenomenology of the Ayahuasca experience. Oxford, UK:

Oxford University Press.Sillito, A. M., Cudeiro, J., & Jones, H. E. (2006). Always

returning: feedback and sensory processing in visual cortexand thalamus. Trends in Neurosciences, 29(6), 307–316.

Stewart, J. (2010). Foundational issues in enaction as a para-digm for cognitive science: From the origin of life to con-sciousness and writing. In J. Stewart, O. Gapenne & E. A.

Di Paolo (Eds.), Enaction: Toward a new paradigm for cog-

nitive science (pp. 1–31). Cambridge, MA: The MIT Press.Strassman, R. J. (2001). DMT: The spirit molecule. Rochester,

VT: Park Street Press.Takaki, R., & Ueda, N. (2007). Analysis of spiral curves in

traditional cultures. Forma, 22, 133–139.


Takaura, K., Yoshida, M., & Isa, T. (2011). Neural substrateof spatial memory in the superior colliculus after damageto the primary visual cortex. The Journal of Neuroscience,

31(11), 4233–4241.Taylor, J.-P., Firbank, M., Barnett, N., Pearce, S., Living-

stone, A., Mosimann, U., Eyre, J., McKeith, I. G., &O’Brien, J. T. (2011). Visual hallucinations in dementiawith Lewy bodies: Transcranial magnetic stimulationstudy. The British Journal of Psychiatry, 199, 492–500.

Taylor, P. C. J., Walsh, V., & Eimer, M. (2010). The neuralsignature of phosphene perception.Human Brain Mapping,

31, 1408–1417.Thompson, E. (2007). Look again: Phenomenology and men-

tal imagery. Phenomenology and the Cognitive Sciences, 6,137–170.

Thompson, E., Lutz, A., & Cosmelli, D. (2005). Neuropheno-menology: An introduction for neurophilosophers. In A.

Brook & K. Akins (Eds.), Cognition and the brain: The phi-losophy and neuroscience movement (pp. 40–97). New York,NY: Cambridge University Press.

Turing, A. M. (1952). The chemical basis of morphogenesis.Philosophical Transactions of the Royal Society B: Biologi-

cal Sciences, 237(641), 37–42.Van Horn, S. C., Erisxir, A., & Sherman, S. M. (2000). Relative

distribution of synapses in the A-laminae of the lateral gen-iculate nucleus of the cat. The Journal of Comparative Neu-

rology, 416, 509–520.Varela, F. J. (1991). Organism: A meshwork of selfless selves.

In A. I. Tauber (Ed.), Organism and the origins of self (pp.79–107). Dordrecht, Netherlands: Kluwer AcademicPublishers.

Varela, F. J. (1996). Neurophenomenology: A methodologi-cal remedy for the hard problem. Journal of ConsciousnessStudies, 3(4), 330–349.

Varela, F. J. (1997). Patterns of life: Intertwining identity and

cognition. Brain and Cognition, 34(1), 72–87.Varela, F. J. (1999). The specious present: A neurophenome-

nology of time consciousness. In J. Petitot, F. J. Varela, B.

Pachoud & J.-M. Roy (Eds.), Naturalizing phenomenology:

Issues in contemporary phenomenology and cognitive science

(pp. 266–317). Stanford, CA: Stanford University Press.Varela, F. J., Lachaux, J.-P., Rodriguez, E., & Martinerie, J.

(2001). The Brainweb: Phase synchronization and large-

scale integration. Nature Reviews Neuroscience, 2, 229–239.Virgo, N. (2011). Thermodynamics and the structure of living

systems. D.Phil. Dissertation, University of Sussex,

Brighton, UK.Weber, A., & Varela, F. J. (2002). Life after Kant: Natural

purposes and the autopoietic foundations of biological

individuality. Phenomenology and the Cognitive Sciences,

1, 97–125.Williams, M. A., Baker, C. I., Op de Beeck, H. P., Shim, W.

M., Dang, S., Triantafyllou, C., & Kanwisher, N. (2008).

Feedback of visual object information to foveal retinoto-

pic cortex. Nature Neuroscience, 11(12), 1439–1445.Wilson, H. R. (1999). Spikes, decisions and actions: The dyna-

mical foundations of neuroscience. New York, NY: Oxford

University Press.

Wilson, H. R., & Cowan, J. D. (1973). A mathematical theory

of the functional dynamics of cortical and thalamic ner-

vous tissue. Kybernetik, 13, 55–80.Wilson, S. W. (1983). On the retino-cortical mapping. Interna-

tional Journal of Man-Machine Studies, 18, 361–389.Winkelman, M. (2002). Shamanism and cognitive evolution.

Cambridge Archaeological Journal, 12(1), 71–101.

About the Authors

Tom Froese received an MEng in computer science and cybernetics from the University ofReading, UK (2004). He then obtained a DPhil. in cognitive science from the University ofSussex, UK (2010). He was a postdoctoral research fellow at the neurodynamics and con-sciousness laboratory of the Sackler centre for consciousness science, University of Sussex,UK (2010). Froese then became a JSPS postdoctoral fellow at the Ikegami laboratory of thedepartment of general systems studies, University of Tokyo, Japan (2010–2012). Currently,he is a postdoctoral researcher at the self-organizing systems laboratory of the Instituto deInvestigaciones en Matematicas Aplicadas y en Sistemas, Universidad Nacional Autonomade Mexico, Mexico (2012–2013). His research is focused on developing enactive approachesto understanding the biology, phenomenology, and dynamics of life, mind, and sociality.

Alexander Woodward received his PhD in computer science from the University ofAuckland, New Zealand in 2009. He is currently a postdoctoral fellow at the Ikegami labora-tory of the University of Tokyo (since 2010). His research into computer vision has led himto ask questions about the nature of visual perception in living systems. Current researchfocuses on computer vision, neural networks and reservoir computing, subjective time in thebrain, scientific computing on the GPU, and a maximalist approach to artificial life.

Froese et al. 213

Takashi Ikegami received his doctorate in physics from the University of Tokyo. Hisresearch interest is to build and study artificial life systems ranging from chemical dropletsand evolutionary robotics to Web dynamics. Some of these results have been published inLife emerges in motion (Seido, 2007) and also The Sandwich theory of life (Kodansha, 2012).Takashi Ikegami gave the keynote address at the 20th anniversary of the Artificial Life con-ference in Winchester, UK. He is also a member of the editorial boards of Artificial Life,Adaptive Behavior and BioSystems.


adaptive behavior turing instabilities in biology, culture, … under disinhibited, non-ordinary...

Documents