Jung didn’t really play the science game, and Jungians continue not to play the science game, but if we squint a little, we can see that some of his ideas have been vindicated by science.
The term ‘archetype’ is often misunderstood as meaning certain definite mythological images or motifs. … The archetype is a tendency to form such representations of a motif. … My critics have incorrectly assumed that I am dealing with “inherited representations”
From Jung’s Approaches to the unconscious (1964), quoted in Jones, Mixed Metaphors and Narrative
Shifts (2003).
In this narrow sense, Jung’s archetypes are uncontroversial: Much reasoning in humans is framed in cognitive science as proceeding by a mechanism of counter-factual simulation called manipulation of mental imagery. Kosslyn, for example, was a pioneer of studies of (visual modality) mental imagery manipulation. The knowledge used in this reasoning, stored in neural degrees of freedom, such as firing biases or synaptic connectivity, may be called abstract, semantic, or conceptual when it is generalized from across many episodic contexts and when its access does not strongly evoke vivid, specific memories of the experiences from which it was generalized (i.e. so much as you can reason about dogs without thinking of the last dog you saw, your dog knowledge may be called semantic rather than episodic). Features of concepts which are typical or discriminative of instances of those concepts are sometimes called prototypical, stereotypical, or (occasionally) archetypal. For science here, the study of cognitive biases relating to typical and discriminative features of conceptual reasoning were pioneered by Rosch. In addition to cached knowledge which is source-situated (exemplar knowledge from episodic memories of individual entities), there may be cached abstract knowledge, such as a mental image of a bird which takes its character from typical or discriminative features of the concept (perhaps specified randomly or unspecified in regard to features which have wide variance across birds), rather than as a cut-and-past mixture of bird-instance parts.
This is all pretty strongly based on visual-modality knowledge of physical objects, and it doesn’t always generalize exceptionally well to other modalities. Like Ullman does nice research on how the declarative / procedural division of memory types maps well onto a lexical / grammatical division of language capacities; a lexicon, Ullman says, is cached linguistic knowledge, and certainly I can recall cached lexical entities, like idioms, which are not episodically situated (“sick as a dog” without thinking of anyone saying that phrase”), yet my introspection suggests that when I think about idioms generally, I’m not usually looking to “sick as a dog” to guide my reasoning. And it’s much harder to find a reasonable interpretation wherein “sick as a dog” is a conjunction of discriminative features of idioms, than it is to understand how a mental image of a bird could have typical features of birds and yet not be identifiable as belonging to a recognized genus.
Other parts of the Jungian archetype construct than “capacity to reason with non-situated mental images having discriminative features” probably haven’t been borne out so well by scientific inquiry.
Is there decent evidence behind the concept of Jungian archetypes?
Jung didn’t really play the science game, and Jungians continue not to play the science game, but if we squint a little, we can see that some of his ideas have been vindicated by science.
From Jung’s Approaches to the unconscious (1964), quoted in Jones, Mixed Metaphors and Narrative Shifts (2003).
In this narrow sense, Jung’s archetypes are uncontroversial: Much reasoning in humans is framed in cognitive science as proceeding by a mechanism of counter-factual simulation called manipulation of mental imagery. Kosslyn, for example, was a pioneer of studies of (visual modality) mental imagery manipulation. The knowledge used in this reasoning, stored in neural degrees of freedom, such as firing biases or synaptic connectivity, may be called abstract, semantic, or conceptual when it is generalized from across many episodic contexts and when its access does not strongly evoke vivid, specific memories of the experiences from which it was generalized (i.e. so much as you can reason about dogs without thinking of the last dog you saw, your dog knowledge may be called semantic rather than episodic). Features of concepts which are typical or discriminative of instances of those concepts are sometimes called prototypical, stereotypical, or (occasionally) archetypal. For science here, the study of cognitive biases relating to typical and discriminative features of conceptual reasoning were pioneered by Rosch. In addition to cached knowledge which is source-situated (exemplar knowledge from episodic memories of individual entities), there may be cached abstract knowledge, such as a mental image of a bird which takes its character from typical or discriminative features of the concept (perhaps specified randomly or unspecified in regard to features which have wide variance across birds), rather than as a cut-and-past mixture of bird-instance parts.
This is all pretty strongly based on visual-modality knowledge of physical objects, and it doesn’t always generalize exceptionally well to other modalities. Like Ullman does nice research on how the declarative / procedural division of memory types maps well onto a lexical / grammatical division of language capacities; a lexicon, Ullman says, is cached linguistic knowledge, and certainly I can recall cached lexical entities, like idioms, which are not episodically situated (“sick as a dog” without thinking of anyone saying that phrase”), yet my introspection suggests that when I think about idioms generally, I’m not usually looking to “sick as a dog” to guide my reasoning. And it’s much harder to find a reasonable interpretation wherein “sick as a dog” is a conjunction of discriminative features of idioms, than it is to understand how a mental image of a bird could have typical features of birds and yet not be identifiable as belonging to a recognized genus.
Other parts of the Jungian archetype construct than “capacity to reason with non-situated mental images having discriminative features” probably haven’t been borne out so well by scientific inquiry.