Shimada, S., Tanaka, S., Morioka, S., Rode, G., Roy, J.-M., & Rossetti, Y. (2026). Narrative embodiment: A conceptual framework linking the narrative self and the embodied self. New Ideas in Psychology, 83, 101285. https://doi.org/10.1016/j.newideapsych.2026.101285
Shimada and colleagues (2026) propose “narrative embodiment” as a conceptual framework linking two aspects of the self: the narrative self and the embodied self. Their central proposal is that “character,” borrowed from Paul Ricoeur, functions as an intermediary between narratives about who we are and relatively stable behavioral dispositions. Narratives may therefore alter behavior by changing character, while embodied experiences may feed back into character and ultimately change the narrative self. The authors apply this framework to phenomena ranging from identification with fictional characters and stereotype priming to virtual-reality avatars and rehabilitation.
The article is interesting partly because it raises a broader question about how theoretical articles should be evaluated. Not all theories have the same epistemic status. In particular, it is useful to distinguish speculative theories from explanatory theories.
Speculative and explanatory theories
At one end of a continuum are speculative theories. These theories propose mechanisms or constructs that could potentially explain observations, but the empirical phenomena they are intended to explain have not themselves been firmly established. Freud’s psychoanalytic theories provide a familiar historical example. Concepts such as repression, the id, ego, and superego offered potentially interesting explanations of human behavior, but the empirical foundations for many of these explanations were weak and the theories were sufficiently flexible to accommodate many different observations.
At the other end are theories constructed to explain credible empirical findings. Perception research provides many examples. Psychophysical phenomena can often be demonstrated repeatedly under highly controlled conditions. Researchers may disagree about why a particular perceptual phenomenon occurs, but there is little disagreement that the phenomenon itself occurs. In this situation, theories compete to explain reliable data. The theoretical problem is not “Does this phenomenon really exist?” but “What mechanism explains it?”
This distinction can be represented on a five-point continuum:
- Highly speculative — the theory attempts to explain reported findings whose credibility or replicability has not been established.
- Mostly speculative — some credible empirical evidence exists, but important parts of the empirical foundation remain uncertain.
- Mixed — the theory integrates both well-established findings and substantially less certain findings.
- Mostly explanatory — the major empirical phenomena are supported by strong and reasonably replicable evidence.
- Explanatory theory of credible findings — the phenomena to be explained are firmly established; the principal uncertainty concerns their explanation rather than their existence.
An important implication of this distinction is that the number of empirical citations in a theoretical article does not necessarily make the theory empirically grounded. A theory can cite dozens of experiments and nevertheless remain highly speculative if those experiments are unreliable. What matters is the credibility of the phenomena that the theory attempts to explain.
Narrative embodiment
On this continuum, I would rate the narrative-embodiment framework 1 out of 5.
This does not mean that the theory is necessarily false. In fact, several aspects of it are intuitively plausible. It seems entirely possible that people’s narratives about themselves influence their behavior and that behavioral experiences subsequently influence how people think about themselves. The authors also make a constructive attempt to translate their framework into empirically measurable variables such as identification, self-efficacy, body image, and behavioral expression.
The problem is that the article does not critically evaluate the credibility of the empirical phenomena used to motivate the theory. Instead, published findings are typically treated as facts that require explanation.
The clearest example appears in the section on stereotype effects. The authors cite the famous study by Bargh, Chen, and Burrows (1996), in which participants exposed to words associated with old age reportedly walked more slowly after leaving the laboratory. Shimada et al. describe the result as follows:
“This demonstrates that stereotypes can shape behavior even without conscious awareness.”
The word “demonstrates” is important. The Bargh finding is not presented as a historically influential but controversial result. It is presented as empirical evidence for the proposed theoretical framework.
Yet the credibility of this finding has been questioned for well over a decade. Doyen et al. (2012) conducted a larger replication using automated measurement of walking speed. Their first experiment found essentially no difference between the elderly-prime and control conditions. The study included 120 participants, compared with 60 across Bargh et al.’s two original elderly-priming experiments, and used infrared sensors rather than manual stopwatch measurements. (PLOS)
Doyen et al.’s second experiment also showed that the outcome depended on experimenters’ expectations. The predicted slowing effect appeared when experimenters had been induced to expect primed participants to walk more slowly. Doyen et al. therefore concluded that priming alone was insufficient to reproduce the original walking-speed effect under their conditions. (PLOS)
Whatever one’s final judgment about behavioral priming, these findings clearly make the empirical status of the Bargh effect relevant to any theoretical review published in 2026. Yet Shimada et al. cite Bargh et al. without mentioning Doyen et al. or the broader controversy surrounding the replicability of behavioral priming.
This is not merely a minor omission in the reference list. It illustrates a fundamental problem in building psychological theories from the published literature. If significant findings are accepted at face value, virtually any collection of published results can provide apparent empirical support for a theoretical framework. After the replication crisis, however, the existence of a published finding cannot be equated with the existence of a credible phenomenon.
A more detailed examination of the history of the elderly-priming effect shows why this distinction matters. The evidence was much less consistent than the textbook version of the finding suggests, and later large preregistered studies provided additional negative evidence. A meta-analysis shows that the existing evidence is so weak and inconsistent that the true effect can range from no effect to a very strong effect without any known moderators (“Elderly Priming: Did It Ever Work?” ReplicabilityIndex)
A theory in search of credible phenomena
The same concern applies more broadly to the article. The authors draw on identification with fictional characters, stereotype effects, the Proteus effect, avatar embodiment, self-efficacy, narrative therapy, and other literatures. Some of these phenomena may ultimately prove more robust than others. Indeed, the authors cite meta-analytic evidence for the Proteus effect and occasionally acknowledge inconsistent findings. But there is no systematic attempt to distinguish highly credible phenomena from literatures characterized by small studies, selective reporting, or replication problems.
Consequently, the empirical literature functions primarily as illustration rather than as a set of established facts that constrain the theory.
This distinction is important because a genuinely explanatory theory should face constraints imposed by reliable observations. Suppose, for example, that repeated high-powered experiments established that identification with a fictional character reliably changed several character-consistent behaviors that had never been directly primed. A theory would then be needed to explain why these effects occur together. Shimada et al.’s concept of “character” might provide one possible explanation.
Indeed, one of the most promising ideas in the article points in precisely this direction. The authors suggest that activating one aspect of a character should affect other behaviors associated with the same character. Experiencing Superman’s ability to fly, for example, might subsequently increase helping behavior or bravery. If such cross-behavior effects were demonstrated reliably, especially under conditions that distinguished character identification from simpler priming or expectancy explanations, the narrative-embodiment framework would become substantially more explanatory and less speculative.
At present, however, the direction of inference is largely reversed. Published findings are assembled into a framework first, while the reliability of those findings is mostly taken for granted.
Conclusion
Narrative embodiment is an imaginative theoretical proposal. It integrates philosophical ideas about narrative identity and embodiment with several areas of psychological research and generates potentially testable hypotheses. For a journal called New Ideas in Psychology, this type of conceptual speculation is entirely appropriate.
But it is important to distinguish an interesting idea from an empirically established explanation.
On a continuum from speculative theory to explanation of credible empirical findings, I would rate the article:
1 / 5 — highly speculative.