Thursday, October 10, 2013

Is Jung's Archetypal Memory Real? If So, Can We Identify it in the Brain?

For anyone interested in Jungian studies, this is an interesting new entry into the neuroscience literature, where Jung and his ideas have essentially been ignored. Of the three best known off-shoots from the original Freudian group (Freud and psychoanalysis, Jung and Analytical Psychology, and Alfred Adler and Individual Psychology), only the psychoanalytic model has continued to evolve in any serious way. With the emergence of Self Psychology in the 1970s and 1980s, then the growth of both relational psychoanalysis and intersubjective systems theory (both of which includes elements of attachment theory and neuroscience [interpersonal neurobiology, or affective neuroscience]), the modern psychoanalytic school is thriving.

Not so true for Jung and the post-Jungians. There has been very little research into Jungian ideas and constructs - and those that have been developed, for example his typology model that has become widely known as the Myers-Briggs model, have been dismissed. Ideas of the unconscious (though now more accepted by some neuroscientists), or archetypes, or synchronicity have all been laughed at and dismissed as worthy of investigation.

Slowly, that is changing.

The study presented below (introduction only) attempts to ascertain if there is any validity to the idea of collective archetypes or a collective unconscious.

Full Citation:
Sotirova-Kohli, M; Opwis, K; Roesler, C; Smith, SM; Rosen, DH; Vaid, J; Djonov, V. (2013, Oct 9). "Symbol/Meaning Paired-Associate Recall: An “Archetypal Memory” Advantage?" Behavioral Sciences; 3(4): 541-561. doi:10.3390/bs3040541

Symbol/Meaning Paired-Associate Recall: An “Archetypal Memory” Advantage?

Milena Sotirova-Kohli [1],, Klaus Opwis [1], Christian Roesler [1], Steven M. Smith [2],
David H. Rosen [3], Jyotsna Vaid [2], and Valentin Djonov [4]
1. Department of Psychology, University of Basel, Missionstrasse60/62, Basel 4055, Switzerland;
E-Mails: (K.O.); (C.R.)
2. Department of Psychology, Texas A&M University, College Station, TX 77843, USA;
E-Mails: (S.M.S.); (J.V.)
3. School of Medicine, Oregon Health and Science University, Portland, OR 97239, USA;
4. Institute of Anatomy, University of Bern, Balzerstrasse 2, Bern 3000, Switzerland


The theory of the archetypes and the hypothesis of the collective unconscious are two of the central characteristics of analytical psychology. These provoke, however, varying reactions among academic psychologists. Empirical studies which test these hypotheses are rare. Rosen, Smith, Huston and Gonzales proposed a cognitive psychological experimental paradigm to investigate the nature of archetypes and the collective unconscious as archetypal (evolutionary) memory. In this article we report the results of a cross-cultural replication of Rosen et al. conducted in the German-speaking part of Switzerland. In short, this experiment corroborated previous findings by Rosen et al., based on English speakers, and demonstrated a recall advantage for archetypal symbol meaning pairs vs. other symbol/meaning pairings. The fact that the same pattern of results was observed across two different cultures and languages makes it less likely that they are attributable to a specific cultural or linguistic context.

1. Introduction

The notions of archetypes and the collective unconscious, which are central to analytical psychology, have generally remained outside the domain of inquiry of mainstream academic psychology. Nevertheless, there are emerging efforts to integrate ideas from analytical psychology and those drawn from cognitive psychology, neuroscience and even physics, e.g., [1–9], etc. To date, these efforts have largely aimed at a theoretical or conceptual integration. Attempts to operationalize or empirically test ideas from analytical psychology are still fairly uncommon.

Two studies that did seek to provide an empirical test of the notion of archetypes are therefore noteworthy, see [2,10]. Rosen et al. [2] found that participants could not reliably identify the proposed associated meaning of symbols deemed to be archetypal when they relied only on resources available to consciousness. However, when participants were presented with pairs of symbols and meanings to learn in a paired-associate recall procedure, they showed significantly better recall of those pairs in which the archetypal symbols were matched with their associated archetypal meanings than those in which the associated meaning did not correspond to the archetypal meaning. In interpreting their results, the authors theorized that the presentation of the symbol and the associated meaning mobilized prior, implicit associations encoded in memory which under normal conditions are not available to conscious recall. The results of this initial study were subsequently replicated by Huston [11] and Bradshaw and Storm [12].

Although these results may be viewed as lending empirical support to the notion of the existence of collective unconscious (archetypal) memory, they may also reflect linguistic or cultural characteristics of the population tested (native speakers of English in the United States and Australia). To determine whether the obtained effect is not unique to this population it is important to conduct studies with native speakers of other languages, and in other cultural contexts. This was the aim of the present study. In this study we developed a German language adaptation of the materials used by Rosen et al. and tested participants residing in the German-speaking part of Switzerland. It was hypothesized that if certain symbols truly have underlying, perhaps universal, “archetypal” meanings, then they should be significantly better recalled if they are paired in a memory task with those meanings than if they are paired with other meanings unrelated to the archetypal ones.

Before proceeding with a description of our study a brief background discussion of archetypes as developed by Jung is in order.

1.1. Archetypes

Unlike Freud, Jung believed that the dynamic unconscious was not just the seat of sexual and aggressive instincts and repressed wishes. Through his work with the word association test, the study of myths and fairy tales, and of fantasy products of psychotic patients, Jung reached the conclusion that there was a layer of the unconscious which contains images, patterns of behavior and modes of perception accessible to the whole of the human race (and to the animal world, as well). He named these specific patterns of perception and behavior which crystallize in consciousness in the form of symbols archetypes (the word archetypos was used by Plato for his ideas and Jung knew this as was pointed out by Barnes [13]). Jung and suggested that archetypes were “empty and purely formal” ([14], p. 79, par. 155) “a possibility of representation given a priori” ([14], p. 79, par. 155). Further on, Jung stressed that “the representations themselves are not inherited” ([14], p. 79, par. 155). In this sense, Jung believed that the archetype-as-such is unknowable and “irrepresentable” ([15], p. 213, par. 417); rather, it affects consciousness mainly from its “ability to organize images and ideas” ([15], p. 231, par. 440). In Jung’s view, the archetype “can be named and has an invariable nucleus of meaning—but always only in principle” ([14], p. 80, par. 155). Anything we say about the archetype remains a visualization which is made possible by the current state of consciousness at a given moment. Archetypes for Jung are numinous (that is, highly emotionally charged) and are associated with strong affective responses. Furthermore, the archetype was thought by Jung to have a “psychoid nature” ([15], p. 215, par. 419), which he described as follows: ”the archetype describes a field which exhibits none of the peculiarities of the physiological and yet, in the last analysis, can no longer be regarded as psychic, although it manifests itself psychically” ([15], p. 215, par. 420). In other words, as conceptualized by Jung, archetypes-as-such while being universal are unknowable or unconscious, but can have a profound impact on consciousness and the life of the individual. They do not belong just to the psychic sphere and seem to be given a priori as a possibility or as a form without content.

It has been noted that Jung’s account of archetypes is multifaceted. For example, Roesler [9] pointed out that we can speak of at least four different definitions of the archetype in Jung’s writing. The first is a biological definition, according to which the archetype was considered as an inborn pattern of perception and behavior. The second definition is an empirical-statistical one based on Jung’s work with the word association test, according to which the archetype is the nucleus of the categories of complexes noted by him in different individuals. A third definition views archetypes as transcending any particular time, place or individual and whose real nature can never become conscious. Finally, there is a cultural-psychological understanding of the archetype which differentiates between the archetype-as-such and its concrete manifestations which are culturally determined [9]. Although depending on the theoretical orientation there can be significant overlap between these definitions, the research reported here investigates primarily the first, biological, definition of the archetype but it is also compatible with the third definition.

Contemporary researchers have tried to reformulate the theory of the archetype to make it more compatible with notions in modern science. Among one of the most well formulated approaches is a model which theorizes that what Jung might have meant with the archetype is similar to the contemporary cognitive semanticists’ notion of image schemas [3–5,16–18], that is, a structure of sensorimotor experience that captures a “dynamic, recurring pattern of organism-environment interactions” ([19], p. 136), that can be—“recruited for abstract conceptualization and reasoning” ([19], p. 141). Image schemas are thought to be “preverbal and mostly nonconscious” ([19], p. 144). Jean Knox [3] first proposed a connection between the notion of an image schema and the archetype-as-such. In this sense the archetype is looked at as an early achievement of development resulting from the qualities of the brain as a dynamic system and the interactions between the individual (biological and psychological) and the environment (social, cultural and physical). This understanding of the archetype uses a dynamic systems approach to the development of cognition and action. This approach to cognition and action relates to the process of formation of preverbal image schematic representations in the infant’s brain which are largely determined by the history of the brain as a system, i.e., are based on the experience the system has in the physical world and the ability of the brain as a dynamic system to self-organize [20]. Later on, this pre-verbal neuronal activation pattern serves as a foundation for the development of conceptual thought—categories and concepts. In themselves these neuronal activation patterns constitute attractor states for the dynamic system of the brain.

The idea of the image schema also finds support in contemporary research on embodiment where embodiment is defined as the meaning of symbols to an agent and the reasoning about meaning and sentence understanding which “depends on activity in systems also used for perception, action and emotion” ([21], p. 4). Neuroimaging studies support the idea that sensory and motor systems are involved in concept understanding and retrieval [22]. Thus, image schemas can be understood as neuronal activation patterns which encode embodied experience in the world. They function automatically, i.e. unconsciously, and underlie concepts, narrative and ritual [23], all qualities which can be attributed also to archetypes.

Varela, Thomson and Rosch [24] propose a slightly different approach to cognition and action, namely, an enacted cognition approach to the study of mental processes and representations. According to this approach, cognition is “enaction: a history of structural coupling that brings forth a world” ([24], p. 172); this view seems consistent with most of the above mentioned ideas. Varela et al. go a step further to suggest that “the cognitive system projects its own world, and the apparent reality of this world is merely a reflection of internal laws of the system” ([24], p. 172).

Among Jungian scholars, George Hogenson [25] looked into the connection between archetypes and mirror neurons and proposed understanding the archetype as an “elementary action pattern” ([25], p. 325), which sounds similar to some of the ideas of the enacted cognition approach of Varela, Thomson and Rosch. Other Jungian scholars stress in their re-interpretation of the nature of the archetype non-linear dynamics which underlie both the functioning of the brain as a system and some aspects of the archetype related to, for example, synchronicity, enantiodromia, or the therapeutic relationship looked at as a dynamic open system. Hogenson proposed that the archetype could be understood as an “iterative moment in the self-organization of the symbolic world” ([26], p. 279). Saunders and Skar have suggested that the archetype is an emergent structure which derives from the self-organizing properties of the brain (a notion very similar to the theory of the image schema) [27]. McDowell stressed that the archetype was a pre-existing principle of the organization of personality [28], while van Eewynk [29,30] looked at archetypes as strange attractors of the dynamic system of the psyche whose non-linear dynamics underlie individuation and the therapeutic relationship.

Perhaps one of the most controversial aspects of the notion of archetypes is that of innateness. How do we understand innateness and what was actually meant by Jung when he stated that archetypes are a priori given to us? Furthermore, how do we understand the innateness of archetypes in an age in which the meanings of symbols are not likely to be transmitted genetically?

While there are still proponents of the idea that archetypes are transmitted genetically (see for further information the review by Roesler [1]), many consider discussions of nature versus nurture to be obsolete and stress the interactionist nature of human development [1,4,9,17,25,31] or point out psychological factors in evolution in the argumentation against a purely genetically transmitted innateness [32]. The innate aspect of the archetype can also be looked at as predisposition to a genetic condition which needs certain environmental cues to find expression in the sense of epigenetics as described by Roesler [1,9] and Rosen [31,33]. In the light of new discoveries it might well be the case that this epigenetic process which provides the link between environment and genome and determines which genes are being active and which are deactivated might even be more important than the genes themselves and may provide the link between biological substrates—genome and cultural heritage—behavior, habits etc. [34]. The Jungian scholar Pietikanen [35] suggested a radical departure from the discussion about innateness and proposed that with the help of a Cassirerian approach archetypes could be understood as “culturally determined functionary forms organizing and structuring certain aspects of man’s cultural activity” ([35], p. 325).

Regarding inborn behavior and archetypes there appears to be empirical support for innateness in experimental psychology for a range of phenomena including the deep structure of language [36], early attachment patterns [37], the idea of “basic emotions”, language acquisition mechanisms, and a face recognition program [1,9]. Roesler [1] points out Seligman’s concept of “preparedness to learn” as a further example of innateness that can be applied to archetypal theory. Similarly, Erik Goodwyn [8,38] uses in defense of innateness findings from evolutionary psychology and neuroanatomy.

We can also say that controversies concerning innateness and the archetype reflect broader controversies in psychology at large. While approaches such as the dynamic systems approach, cognitive semantics, embodiment and enacted cognition as approaches in the study of cognitive processes enjoy widespread popularity, there are also many scholars who conduct experimental work in connection with innate mechanisms. The experimental work of developmental psychologists such as Spelke provides data which supports the hypothesis of multiple innate mechanisms with which infants are equipped at birth. Spelke suggests that “perception, thought, value and action depend on domain-specific cognitive systems” and “each system has its own innate foundations and evolutionary history” ([39], p. 204). For example, in a recent study Izard, Sann, Spelke and Steri [40] report findings that support the assertion that infants at birth are equipped with abstract, numerical representations. Yet other cognitive scientists do not readily accept the notion that there are innate foundations for cognitive capacities, particularly for certain capacities, such as language. It, thus, seems that cognitive science at large is still grappling with questions concerning innateness.

The debate around the nature of the archetype is further enriched by archetypal psychology which sees the place of the archetype in imagination and stresses the transcendental nature of the archetype [1,9]. Although this approach to the archetype might not resonate with many mainstream psychologists, there are tendencies in contemporary studies of consciousness which are compatible with the ideas of archetypal psychology. The Hameroff and Penrose quantum theory of consciousness [41], the idea that consciousness “emerges as natural processes” that involve quantum phenomena “unfold[ing]” [42], and the hypothesis that the brain does not produce consciousness but serves the purpose of receiving and transmitting information which exists from beyond it [43] can all be seen to resonate with some of the basic ideas of archetypal psychology concerning the archetype. Furthermore, the notion of synchronicity—meaningful coincidences—based on an acausal connection principle, which Jung developed in exchange with Wolfgang Pauli and Albert Einstein, and which can be seen as an expression of a constellated archetypal field at work [6,44], finds in recent days, support through discoveries in complexity theory and the dynamics of complex adaptive systems [7].

Given all these ideas how are we to understand the archetype? Are archetypes transmitted biologically or are they transmitted by culture as Roesler [1] asks? Can we understand the collective unconscious in terms of subliminal transmission and inter-individual neuronal format as Roesler [1] proposed or is it a form of archetypal memory as Rosen et al. [2] suggested? However we reformulate the theory of the archetype and the collective unconscious most Jungian scholars would agree that the basis of the archetype and the collective unconscious is both innate and environmental. The differences are more in terms of degree and the role of each of the two factors.

While the above developments in psychology provide much food for thought, finding a way to test notions about archetypes, however this notion is formulated, would be instructive. We thus turn to two previous empirical studies which attempted such a test and found empirical support in favor of the existence of something akin to archetypes, henceforth termed the archetype hypothesis.

1.2. Previous Research

Apart from the above mentioned theoretical discussions concerning the nature of the archetype a few scholars have sought to empirically test the hypothesis of archetypes and archetypal memory.

As mentioned above, Rosen et al. [2], as well as Huston, Rosen and Smith [45], Bradshaw and Storm [12] and Maloney [10] examined this in the domains of memory and preferences.

Maloney [10] asked a community sample of 151 participants to rate their preferences to images containing archetypal themes and factor analyzed the responses. The images included the archetypal themes of the mother and the hero in both anthropomorphic (e.g., woman gazing lovingly at a child  for the positive mother, Hercules for the positive hero) and non-anthropomorphic (e.g., the cave as a symbol of the Great Mother, the heraldic lion as a symbol of the hero) form. Both positive and negative aspects of these themes were examined. The study used an unconstrained Q-sort method. Participants were presented with sets of six images and asked to rate their responses to three questions in respect to the images using a limited set of possible answers. The analysis demonstrated a stable three-factor structure underlying responses to the question “If I were to keep this image with me forever, I would be”. Factor 1 contained images related to a quest theme—the positive hero, the non-anthropomorphic hero, the non-anthropomorphic mother, according to the author. Factor 2 was reported to contain images related to an attachment theme—positive mother. Factor 3 was interpreted as being related to a conflict theme. The author thus concluded that “archetypal structure underlies adult affective responses” ([10], p. 110). Furthermore, Maloney concluded that the images alone were not enough to evoke an archetypal structure, they had to be viewed in a certain way so that the structure was triggered which in the design of his study was achieved through the question that the subjects had to answer. Only the question which required most active participation on the part of the participants in assessing the images yielded significant results.

A different experimental paradigm was developed by Rosen, Smith, Huston and Gonzales [2]. Rosen and colleagues argued that a natural extension of Jung’s own early studies with the Word Association Test would be the study of associations on the basis of symbols. They developed an inventory of forty symbols and forty associated words which were intended to correspond to the symbol’s archetypal meanings—The Archetypal Symbol Inventory (ASI). Furthermore, they designed a cognitive psychological experimental paradigm to test the hypothesis that archetypal symbols were strongly associated to these proposed underlying meanings and that the association lies beyond conscious retrieval under ordinary conditions. Rosen et al. conducted a series of three experiments with undergraduate students in psychology at a large university in southwestern U.S. The first two experiments tested participants’ conscious knowledge of the symbols and their meanings. When they were shown each of the ASI symbols, and asked to guess the meaning of each symbol, American participants could not come up with the designated meaning of the symbols. Even more surprisingly, when they were given the 40 ASI symbols with a randomly ordered list of the meanings, participants were unable to match symbols to their correct meanings above the level of chance. These results show that participants were not consciously aware of the meanings of the symbols. The third experiment was a paired-associate learning task in which students (divided into two groups) were first shown all forty symbols. Each group was given half of the symbols matched with the proposed associated meanings and the other half with symbols and meanings mismatched (the particular pairings were counterbalanced across the two groups). After a one minute rest participants were shown only the symbols and were asked to remember and write down the word they initially saw paired with the symbol. It was found that students learned and recalled significantly better the words whose meanings corresponded to the proposed meanings of the archetypal symbols than those that were unrelated to the purported meaning of the symbols. From the list-learning research literature (e.g., [46,47]) it is known that pairs of strongly associated words are learned better than less associated pairs. This gave ground to the authors of the study to conclude that archetypal symbols are strongly associated to the proposed related meanings and that the association is unconscious.

Huston, Rosen and Smith [45] proposed a mechanism to explain the observed effects in the original Rosen et al. study and a second variation of the research [11]. They suggested that when a symbol  was presented paired with its associated “archetypal” meaning priming occurs which facilitates later recall. The correctly paired symbol with its proposed related meaning also triggers an emotional response which contributes to the “activation and constellation of an archetypal image” ([45], p. 147). The constellated archetypal image and the associated meaning presented to participants together led to priming of memory for the association and facilitated later recall. The mechanism proposed by the above authors is still in the realm of hypothesis and needs to be experimentally tested.

In a recent study Bradshaw and Storm [12] conducted three experiments based on the Rosen and Smith paradigm using 30 out of the original 40 symbols from the ASI in a sample of 237 students and members of the general public in the state of Victoria, Australia. The sample consisted of predominantly Australian/New Zealander citizens (81%) and was predominantly English native speaking (around 86%). The other countries/regions represented were respectively, Britain (3%), Europe (4%), Asia (7%), America (North and South 2%) and Other 3%. The authors replicated the results of Rosen and Smith in the free association task (Experiment 1) and detected in the forced association task (Experiment 2) seven out of 30 symbols which could be consciously known by the participants. For the rest of the symbols there was no statistical evidence in the forced association task for conscious knowledge. The authors modified the paired-associate learning task used in the third experiment of the paradigm. To additionally control for intermediate effects they presented four randomized versions of symbol-word sets, i.e. instead of two counterbalancing conditions they had four. Furthermore they modified the timing in the list learning task giving participants 8 seconds in  the learning phase as opposed to 5 seconds in the original paradigm and 20 seconds in the recall phase as opposed to 8 seconds in the original paradigm. As stimuli the authors used a set of pictures and drawings of the symbols predominantly downloaded from Internet instead of the original images from the ASI. There was no explanation given for the above modifications. The results replicated the findings of Rosen et al. [2] and Huston [11]. Matching words with the symbol that they are associated with, benefitted learning and subsequent recall of the words. The authors reported a statistically significant difference between the different versions of the main experiment. There was a statistically higher recall rate for both matched and mismatched recall in one of the versions. This was partially explained by the age difference between the participants in this version (M = 23 years) and one of the other versions (M = 30 years). No information is available about the mean age in the other groups, as well as the means and standard deviations for matched and mismatched recall in the different groups. Furthermore, the authors detected increased difficulty in learning and recall of mismatched pairs with increased age in their sample (mean age 27, SD = 11 years). No significant interaction between country and ethnicity and performance was found on any of the tasks in all three experiments. This is not surprising since as noted above the sample consisted of predominantly Australian/New Zealander citizens (81%). The number of participants from other countries of origin was very small. As such it could be argued that the sample size of the individual ethnic groups (distributed across the 6 different conditions) was too small to detect any meaningful difference. There is also no information available about how the different ethnic groups or counties of origin were represented across the different experimental conditions. Furthermore, the experiment was carried out in English. All participants, even those who were not native English speakers (14% or less since the authors did not control for language which the participants consider to be their native language) used English as the experimental language. In this sense, it cannot be ruled out that the effect which the authors report (no difference in performance between the different ethnic groups, as well as the significant effect of  matching on learning and recall) can be explained by characteristics specific to the English language.

Following its publication the Rosen et al. study led others to wonder how robust or generalizable the findings were. Jill Gordon [48] posed the question whether the images used by the team could be considered to be archetypal before additional, cross-cultural, research is conducted using the same paradigm. Similarly, Gordon stressed the importance of conducting cross-cultural studies to determine whether the images used really had the qualities of archetypal images, namely, whether these were “forms that provoke more or less similar or even identical associations from a majority of people” ([48], p. 229). Raya Jones argued in a similar fashion that the results observed by Rosen et al. could be explained either in terms of “cultural convention” or as “artifacts of the statistical procedure” ([49], p. 707).
Read the whole paper.
Post a Comment