I believe that mythical stories can help us grasp large, cohesive organizations that possess agency—aka ogregores. However, there’s a wrinkle. Most if not all stories involve human-like characters. On the other hand, ogregores aren’t human and don’t behave like them. Animist narratives which portray non-human agents without using anthropomorphism and psychological research into the causes of anthropomorphism may point a way around the dilemma.
Ogregores aren’t Human
Let’s start with how ogregores’ umwelt—the way they perceive and experience the world. An entity’s umwelt is determined by the capabilities of its perceptual systems. There are differences between ogregores’ umwelts and our own. Their sensors are widely distributed since their operations are spread out geographically, from vendors and manufacturing plants to warehouses, data centers, and sales outlets. Also, unlike our primary senses of sight, smell, sound, taste and touch, ogregores primarily perceive and process numbers: product prices, stock prices, inventory levels, sales volumes, profit margins, cash flows, sales cycle lengths, customer churn rates, average revenue per user, cost of capital, web site traffic, marketing ROI, production efficiency, employee turnover rates, customer service request and clearance rates, etc.
Unlike animals, ogregores are not compact, solid blobs with clear spatial boundaries. Their components are spread out and connected by communication links. The closest biological analogy might be “superorganisms” like bee hives and ant hills, though ogregores have a much lower spatial density and large companies can have more employees most hives. (A typical honeybee hive contains 20,000–80,000 bees; however, colonies of the ant Formica yessensis can exceed 300 million. The largest human organizations are the defense ministries/departments in India and the U.S. with 3 million people.) Ogregores do not move as animals (including insect swarms) do; instead, they add or remove distinct, distant components like factories, offices, or outlets.
Ogregores don’t have structures corresponding to human physiology, so I assume that they do not experience pleasure or pain. Since they respond and adapt to favorable or adverse environmental signals, one can probably ascribe sensations to them that correspond, for example, to key performance indicators falling below a threshold which leads to resource reallocation, or indicators like a stock price or analyst reports leading to a stock buyback program or changes in hiring. However, it seems unlikely that they experience the affective quality of a sensation or experience, generally categorized as pleasant, unpleasant, or neutral.
Organizations do have life stages like inception, growth, and dissolution but these events are much more diverse than human equivalents. Organizations can be created by legislation, by incorporation, or by splitting off from another entity. Growth can be measured in many ways, including number of employees, sales, or profits rather than size in biological organisms. Dissolution can occur through bankruptcy, hostile takeover, or merger—in which case, it’s not clear whether it’s growth or dissolution. However, organizations don’t reproduce sexually. There are some biological analogs since they can split (like binary fission in bacteria) and merge (like slime mold cells forming a fruiting body) but their “reproduction” very different from that of humans.
They don’t have human-like social structures either. Because they don’t reproduce sexually or use that as a basis for structuring the world, they don’t have families. Ogregores form partnerships (anthropomorphism alert) but they don’t have friendships or mutual affection. They compete, but don’t have enemies (though their human constituents may think of other ogregores and their members that way).
In summary, ogregores don’t have humanlike shape or motion. They don’t have humanlike perceptions, emotions, physical traits, or social structures and therefore they almost certainly lack human motivations and intentions. However, they do have motivations and intentions: at the very least, a drive to survive, and more generally purposes that are given to them at inception or that they acquire over time, such as growth or profitability. Here, perhaps, is a way around the problem of telling stories about entities that are profoundly non-human.
Anthropomorphism and Narrative
According to a summary of the academic literature, anthropomorphism—the attribution of exclusively human traits to non-human entities—is deeply embedded in narrative, and many scholars argue it is largely unavoidable, especially in storytelling and language. Some suggest that it improves memory and understanding.
One can make a strong case that most myths employ anthropomorphism. Across cultures, gods look and behave like humans. They love, hate, scheme, and quarrel. Zeus is lustful, Odin seeks wisdom, and Shiva dances. Myth is usually a form of story, and narrative thrives on characters with agency, intentions, and emotions. Anthropomorphizing gods turns abstract cosmic processes into narratable personalities.
However, there are some counter-examples and caveats. In animist and Indigenous traditions, rivers, mountains, and winds often appear as forces with potency or efficacy rather than personalities. They act, but not with humanlike motives. For example, Papa and Rangi (the Māori earth and sky) exert forces of separation without dialogue or psychological scheming. The agency of the wind or water appears as causal power (it moves, erodes, nourishes, destroys) rather than as the fruit of inner motives. In Inuit traditions, the wind is sometimes addressed as a being that governs weather, dangerous if disrespected, but not consistently described with human traits.
Some cosmologies are abstract or structural and don’t invoke anthropomorphic gods. The Mesopotamian Tiamat initially exists as a chaotic oceanic presence, more a material substrate than a character. In some Dogon cosmology (as reported by Marcel Griaule, though scholars debate its accuracy), primordial beings like Amma are portrayed as cosmic structurers rather than personal agents. Their “actions” are often transformations or emanations of order rather than emotionally motivated choices. In the same vein, some mythic beings (e.g., Oceanus in Hesiod, Wakan Tanka in Lakota traditions) appear in stories but act more as conditions and structures than characters. In Judaism, Islam, and strands of Hinduism and Buddhism, gods or absolute principles are ultimately non-anthropomorphic.
Nurit Bird-David argues that agency in animistic systems should be understood through a network of relationships between beings, both human and non-humans rather than in terms of essences (‘Animism’ Revisited, 1999).
In summary, myths include a spectrum of representation: forces (Tiamat as waters); minimal agents (rivers blocking paths); to anthropomorphic characters (Zeus, Kali, Thor). Most mythologies include all these layers, even within the same text. There is a continuum from non-human agency → partial personification → full anthropomorphism. While forces may enter myth as conditions, anthropomorphism seems unavoidable when long, dialogic stories (conflicts, reconciliations, love affairs) have to be sustained.
Anthropomorphism and Agency
One can argue that agency detection is a precursor to anthropomorphism: once we detect an agent, we are then more likely to attribute human-like qualities to it. There is moderate though debated evidence for a Hyperactive Agency Detection Device (HADD) in people (Consensus.app). This idea proposes that people are predisposed to attribute agency—even when none exists—to ambiguous stimuli. This supposedly explains the universal tendency to believe in supernatural agents like gods and spirits. However, the neural and psychological mechanisms remain poorly defined.
A widespread understanding of agency presupposes intention: if we detect an agent, we ascribe beliefs, desires, and intentions to it. Full-blown anthropomorphism goes further than recognizing intentional agency, however, attributing exclusively humanlike characteristics, emotions, motivations, or physical traits to non-human entities.
There is a middle ground, however, between merely detecting agency and fully ascribing exclusively humanlike traits. One can ascribe intentional agency without anthropomorphizing, but anthropomorphism always involves ascribing intentional humanlike agency. Scholars describe this space in terms of minimal, non-human agency attribution, such as saying “the triangle chased the circle” in Heider & Simmel’s classic 1944 paper, An Experimental Study of Apparent Behavior.
Evading the dilemma
To recap the problem: I ascribe agency to ogregores but deny that they are humanlike. At the same time, I propose that we use mythic stories to better grasp ogregores. However, storytelling entails anthropomorphism, but ogregores aren’t human.
A further difficulty is that there is some evidence that organizations are anthropomorphized. A preprint by Pramesti et al. (2025) provides systematic review of the literature. Ashforth, Schinoff, & Brickson (2020) argue that anthropomorphism is both pervasive and surprisingly important in organizational life. Kervyn, Fiske, & Malone (2012) show that consumers perceive brands in the same way they perceive people. Andersen (2008) gives examples where scholars of organization argue that they’re humanlike, and attempts to disprove the anthropomorphism of organizations. Morris, Sheldon, Ames, & Young (2007) found that financial journalists describe the market in agentic, emotional terms and that exposure to such metaphors influenced investment choices.
One way out is to question the premise that storytelling requires anthropomorphism, for example by drawing on the ways animist traditions depict non-human agency. (I suggested back in 2023 that animism offered a way to think of ogregores as nature sprits than gods.) Entities can be narrated as doing rather than feeling. Ogregore characters can have goals and reasons without having emotions. In animist traditions, a river blocks the hero’s path, punishes arrogance, or carries a child away, without explicit motivation. We speak the same way today, saying that “the algorithm decided” or “my computer doesn’t want to cooperate” without imagining a literal human mind behind it. The appendix lists some ways we might speak about technologies and organizations by ascribing agency but not humanity.
It is also worth noting that the degree of attributed anthropomorphism is highly variable. A highly-cited 2007 paper predicts that people are more likely to anthropomorphize when (1) no other detailed framework, like scientific or experiential knowledge of the nonhuman agent is available; (2) in situations loaded with unpredictability or threat; and (3) when human contact is lacking and social support can by generated by anthropomorphizing pets, gadgets, or supernatural beings (Epley, Waytz, & Cacioppo, 2007). The right circumstances reduce the need for anthropomorphism and thus the need for it in stories about ogregores.
A related source of optimism is that the human mind is supple and plastic, for all its inborn propensities. Music and mathematics describe nonhuman-shaped worlds in largely non-verbal ways. Of course, science regularly relies on narrative structures like explanatory arcs (e.g., “selfish genes” in evolutionary biology) and weak anthropomorphism, i.e., imputing agency but not uniquely human motivations or using playful or metaphorical language (e.g., “molecules recognize each other” and “the model forgets rare cases”; more in the appendix) even as it aspires to objectivity. While science doesn’t fully avoid anthropomorphism, scientists insist that such metaphors are heuristic figures of speech.
I believe that one can reduce the need for, and thus risk of, anthropomorphism in ogregore myths by, among other things, activating knowledge about organizations’ non-human but still agentic characteristics, and by using language that describes generic behavior (aggressive, lethargic, alert, watchful, etc.) and avoids anthropomorphic traits (jealous, angry, thoughtful, considerate, sympathetic, devious, vindictive, etc.). One can focus on attributes they share with humans such as purposes, perceptions, and adapting to their environment, without invoking a human nature.
Despite the evidence given above for anthropomorphizing organizations, I’ve been struck by how blind we are to their agency. Our Hyper-Active Agency Detection mechanism seems to fail when it comes to the agency of ogregores, perhaps because they’re not human enough to trigger anthropomorphism consistently. If nothing else, they don’t exhibit humanlike motion and morphology, which Epley at al. suggest is a key trigger for anthropomorphism.
However, the dilemma may be a paradox that cannot be resolved. This is illustrated by a tension at the heart of the Biblical and Qur’anic traditions. The Hebrew/Islamic God is explicitly transcendent and imageless, but narratives often ascribe human-like emotions and behaviors to it. Humans struggle to avoid anthropomorphism when narrating non-human agency. Theology forbids it, rationality frowns on it, but storytelling demands it.
Thanks to Susan Tonkin for feedback on a draft
Appendix: Modern turns of phrase that treat entities as agents but not human persons
In these examples of contemporary parallels to rivers or winds in mythic storytelling, I use verbs of decision, intention, (in)action, but not emotion, affect, or mood.
- “The algorithm boosted this post” (but not “loved this post”)
- “The recommender system punishes clickbait” (but not “hates clickbait”)
- “My printer refuses to work” (but not “is sullen today”)
- “The network won’t let me connect”
- “The chatbot flagged my request as unsafe” (but not “is in a bad mood today”)
- “The model doesn’t like long inputs” (but not “frowns on long inputs”)
- “The market punished investors” (but not “took against speculators”)
- “The stock market rallied” (but not “is feeling frisky today”)
- “The government decided to intervene” (but not “lost its patience”)
- “The legal system doesn’t tolerate that” (but not “is averse to that”)
- “Twitter doesn’t like external links” (but not “fears external links”)
- “The game won’t let me save here”
- “Excel rejects inconsistent formulas.” (but not “hates inconsistency”
- “The car refuses to start” (but not “is cranky this morning”)
- “The elevator won’t come” (but not “doesn’t want to deal with us”)
- “The grid failed last night” (but not “had a nervous breakdown”)
- “The train decided to skip my stop” (but not “is thinking about other things”)
- “The satellite lost track of the storm” (but not “is feeling poorly today”)
- “GPS refuses to find my location” (but not “is angry at me”)
Examples of weak anthropomorphism in scientific language:
- Genes “compete” and “want to replicate”
- “The cell is trying to repair itself”
- “Molecules recognize each other”
- “Natural selection favors efficiency”
- “Species adapt to survive in harsh climates” (adaptation is mechanistic, not intentional)
- “Black holes eat stars”
- “Electrons want to reach the lowest energy state”
- “Neural nets learn patterns”
- “The model forgets rare cases”
No comments:
Post a Comment