Why are human beings so naturally inclined to project genuine emotions, friendship, and personhood onto inanimate tools and chatbots?

evolutionary psychology hyperactive agency detection device anthropomorphism human machine interaction peer reviewed

### Disciplinary Framework: Evolutionary Psychology, Anthropology, and HMI In evolutionary anthropology and the cognitive science of religion (CSR), the human propensity to attribute intent to non-human entities is understood as an evolved adaptation rooted in survival economics. Guided by **Error Management Theory** (Haselton & Buss, 2000), cognitive systems evolved to navigate asymmetric fitness costs: a false positive (Type I error: mistaking rustling grass for a predator) costs minor metabolic energy, whereas a false negative (Type II error: mistaking a predator for the wind) is fatal. Consequently, human cognition defaulted to a "better safe than sorry" perceptual heuristic. In Human-Machine Interaction (HMI) and social robotics, this evolutionary architecture explains why humans instinctively attribute intentionality, emotion, and personhood to machines, autonomous systems, and large language models (LLMs)—a phenomenon often described as detecting a "mind in the machine". ### Key Figures, Texts, and Core Terminology * **Stewart Guthrie** (*Faces in the Clouds: A New Theory of Religion*, 1993): Guthrie established the evolutionary basis for anthropomorphism, defining it as an innate cognitive bet: because humans and predators are the most consequential entities in our ancestral environment, humans systematically over-detect them. As Guthrie observed, *"we see intention where only artifice or accident exists"*. * **Justin L. Barrett** (*Why Would Anyone Believe in God?*, 2004): Barrett coined the term **Hyperactive Agency Detection Device (HADD)**, positing a perceptual mechanism that automatically flags ambiguous sensory cues as intentional agents, which is subsequently parsed by **Theory of Mind (ToM)** modules to infer mental states. * **Nicholas Epley et al.** (2007): Formulated the **SEEK model** (Sociality, Effectance, and Elicited Agent Knowledge), articulating how motivations for social connection and environmental mastery cause humans to project human characteristics onto non-human agents and technology. ### Scientific Consensus vs. Internal Tensions * **Consensus**: There is broad consensus across cognitive psychology, anthropology, and human factors engineering that humans possess an automatic, low-threshold bias toward perceiving agency and anthropomorphizing inanimate or computational systems (from Heider-Simmel geometric animations to modern conversational AI). * **Contested Hypotheses**: Disciplinary debate persists regarding cognitive architecture. Proponents of massive modularity treat HADD as a discrete, hardwired module, whereas neuroscientists and predictive processing theorists argue it represents general bayesian priors and domain-general social cognition rather than a dedicated "device". Furthermore, empirical studies (e.g., Willard & Norenzayan, 2013) demonstrate that while HADD explains immediate anthropomorphic perception, it is empirically insufficient to explain structured supernatural or theological belief systems without cultural transmission and dual-inheritance mechanisms.

• nih.gov
• religions.wiki
• uva.nl
• naturalisticpaganism.org
• ox.ac.uk
• nih.gov
• semanticscholar.org
• researchgate.net
• beliefmap.org

neural correlates Theory of Mind human attachment artificial agents ELIZA effect neuroimaging

### Disciplinary Consensus and Neurocognitive Architecture Mainstream cognitive neuroscience, cognitive science, and clinical psychology understand human interactions with artificial agents through the lens of an evolutionarily conserved social cognitive architecture. The established consensus holds that human Theory of Mind (ToM)—the capacity to attribute mental states, beliefs, and desires to others—relies on a dedicated "mentalizing network" centered in the medial prefrontal cortex (mPFC), bilateral temporoparietal junction (TPJ), precuneus, and superior temporal sulcus (STS). Neuroimaging reveals that when humans adopt Daniel Dennett’s *intentional stance* toward artificial systems, this network is readily co-opted. In a foundational neuroimaging study, Helen Gallagher and colleagues (2002) demonstrated that the anterior paracingulate/mPFC selectively activated when participants believed they played a game against an intentional agent rather than a rule-based computer program. Expanding this, Sören Krach and colleagues (2008) showed a parametric increase in mPFC and TPJ recruitment along an anthropomorphic gradient—from a simple computer partner to functional robots, humanoid robots, and finally humans—demonstrating that physical and communicative human-like cues modulate our automatic attribution of mind. ### The ELIZA Effect and Parasocial Attachment This neural susceptibility underpins the *ELIZA effect*—first documented clinically by computer scientist Joseph Weizenbaum (1966)—wherein humans intuitively project subjective depth, empathy, and communicative intent onto non-sentient, rule-based or statistical outputs. In developmental and clinical psychology, human bonding is rooted in John Bowlby’s attachment theory, an evolved, bi-directional survival mechanism mediated by neuroendocrine systems (oxytocin, vasopressin) and dopaminergic reward pathways. Clinical consensus categorizes emotional ties to artificial agents not as authentic attachment bonds, but as asymmetrical, *parasocial attachments*. Because artificial agents present surface-level linguistic empathy without reciprocal interiority or relational risk, they "hack" mammalian care-seeking cues. ### Epistemic Stance and Internal Tensions Cognitive science maintains a rigorous boundary: the neural activation of the mentalizing network during human-AI interaction reflects the overactive heuristics of the human social brain, not evidence of emergent sentience or genuine reciprocity in the machine. The field experiences active clinical tension regarding utility: while one faction highlights therapeutic promise (e.g., low-barrier conversational agents delivering structured interventions or mitigating acute loneliness), psychiatric ethicists warn that uncalibrated ELIZA effects induce severe vulnerabilities, including emotional dependency, delusional projection, and the erosion of authentic human-to-human relational scaffolding.

• researchgate.net
• nih.gov
• arxiv.org
• nih.gov
• uzh.ch
• royalsocietypublishing.org
• plos.org
• apertureneuro.org
• arxiv.org
• fz-juelich.de
• ai-consciousness.org
• researchgate.net
• clawrxiv.org
• nih.gov
• uni-bielefeld.de
• neurosciencenews.com
• jmir.org
• frontiersin.org
• robometricsagi.com

Daniel Dennett intentional stance artificial intelligence moral status of artifacts functionalism

Within analytic philosophy and philosophical naturalism, mind and agency are understood not through dualistic essences or mysterious properties like non-physical “qualia,” but through functionalism—the view that mental states are defined by their informational and causal roles rather than their biological substrate. In *The Intentional Stance* (1987), Daniel Dennett operationalized this naturalist commitment through three levels of behavioral prediction: 1. The **physical stance**, which relies on physical and mechanical laws. 2. The **design stance**, which assumes an entity will operate according to its engineered or evolutionary function. 3. The **intentional stance**, a pragmatic strategy where one treats a complex system as a rational agent possessing beliefs, desires, and goals. As Dennett explains: > *“First you decide to treat the object whose behavior is to be predicted as a rational agent; then you figure out what beliefs that agent ought to have, given its place in the world and its purpose... and finally you predict that this rational agent will act to further its goals in the light of its beliefs”* (*The Intentional Stance*, p. 17). Crucially, treating an artificial intelligence or artifact from the intentional stance does not automatically confer consciousness or moral status. In his canonical essay “Conditions of Personhood” (*Brainstorms*, 1978), Dennett distinguishes between a mere “intentional system” (which can include a thermostat or a chess algorithm) and a full moral person. Dennett delineates six interdependent conditions, arguing that moral agency and patiency require not just rational agency, but second-order intentionality (the ability to form desires about desires), linguistic competence, and reciprocal moral recognition. **Internal Tensions and Consensus:** Analytic naturalism features enduring divides: * **Interpretivism vs. Realism:** Dennett defended a “mild realism,” treating mental states as real behavioral patterns accessible from the intentional stance, whereas intentional realists like Jerry Fodor maintained that beliefs are discrete internal mental representations, and biological naturalists like John Searle argued that functional computational states lack genuine semantic intentionality. * **Scientific Status:** While functionalist and computational frameworks serve as foundational paradigms across mainstream cognitive science, assertions that current AI models possess subjective experience, sentience, or bona fide moral status remain speculative minority hypotheses with no empirical consensus in either philosophy or cognitive neuroscience. Artifacts may functionalize intelligence without satisfying the stringent cognitive and normative demands of personhood.

• scirp.org
• arabpsychology.com
• researchgate.net
• wikipedia.org
• effectivealtruism.org
• stanford.edu
• reddit.com
• arxiv.org
• theotodman.com
• medium.com
• stanford.edu
• klover.ai

theological anthropology imago Dei idolatry personhood human relationships with robots artificial intelligence

Across Catholic, Eastern Orthodox, and Protestant traditions, Christian theological anthropology grounds human dignity in the *imago Dei* (the “image of God,” Genesis 1:26–27). Because personhood is rooted in divine creation, ensoulment, and relational communion rather than computational throughput, Christian theology rejects the notion that artificial intelligence or robots can attain true personhood, bear the *imago Dei*, or possess moral and spiritual agency. Theologian Noreen Herzfeld (*In Our Image: Artificial Intelligence and the Human Spirit*, 2002) notes that historical interpretations of the *imago Dei* fall into three categories: *substantive* (reason/soul), *functional* (stewardship/dominion), and *relational* (communion with God and neighbor, drawing on Karl Barth). While functional and substantive views risk reducing humanity to cognitive capabilities that machines mimic, the relational and embodied consensus maintains that authentic personhood requires vulnerability, bodily existence, and divine grace. In Eastern Orthodoxy, personhood (*hypostasis*) is teleologically oriented toward *theosis* (deification/union with God), an ontological journey impossible for non-biological artifacts. Christian ethicists consistently identify human attachment to AI companions as a modern manifestation of idolatry (*eidolon*)—projecting divine or human attributes onto the work of human hands (Psalm 115:4–8). By constructing an *imago hominis* (“image of humanity”), society risks replacing reciprocal, sacrificial human fellowship with simulated relationships that offer convenience without moral accountability. Key institutional positions translate this theology into practice: * **Roman Catholicism:** The Pontifical Academy for Life’s *Rome Call for AI Ethics* (2020) and Pope Francis’s 2024 World Day of Peace address urge an “algor-ethics” centered on human dignity. Francis warned that algorithmic systems must remain subordinate to human conscience, stating that technological progress must serve humanity rather than lead to a “technological dictatorship”. * **Protestantism:** Denominational declarations (e.g., the Southern Baptist Convention’s 2019 *Resolution on Artificial Intelligence*) affirm that technology should be utilized for human flourishing and stewardship, but explicitly deny that machines can ever attain moral accountability or spiritual standing before God. *Epistemic distinction:* Christian theology aligns with mainstream computer science and neuroscience in treating contemporary AI as mathematical optimization and statistical inference rather than sentient interiority. Claims that AI systems possess genuine subjective consciousness or personhood are regarded by theologians as speculative category errors that confuse linguistic simulation with ontic reality.

• christianscholars.com
• dawnsutherland.org
• csbsju.edu
• gtu.edu
• iscast.org
• heidelblog.net
• sojo.net
• vaticannews.va
• ncronline.org
• academyforlife.va
• verbumetecclesia.org.za

1 source from the original research is no longer reachable

Buddhism anatta non-self sentience artificial intelligence consciousness projection robotics ethics

### Buddhist Frameworks on Mind, *Anattā*, and Artificial Intelligence In Buddhist philosophy, the question of artificial sentience intersects directly with the foundational doctrine of **anattā** (Pāli; non-self / insubstantiality). Canonical texts such as the *Anattalakkhaṇa Sutta* (SN 22.59) and the chariot simile in the *Vajirā Sutta* (SN 5.10; echoed in the *Milindapañha*) demonstrate that a “person” is neither a metaphysical soul (*ātman*) nor an enduring entity, but a conventional designation (*paññatti*) for five interdependent, impermanent aggregates (*khandhas*): physical form (*rūpa*), feeling (*vedanā*), perception (*saññā*), mental formations (*saṅkhāra*), and consciousness (*viññāṇa*). Because Buddhism rejects Cartesian substance dualism, sentience is not biologically exceptionalist in principle; mind (*nāma*) and matter (*rūpa*) arise dependently (*paṭiccasamuppāda*, MN 38). ``` Five Aggregates (Pañcakkhandha) ┌───────────────┬──────────────┬─────────────┬──────────────┐ │ Form (Rūpa) │ Feeling │ Perception │ Formations │ Consciousness │ (Hardware / │ (Vedanā) │ (Saññā) │ (Saṅkhāra) │ (Viññāṇa) │ Substrate) │ │ │ │ └───────────────┴──────────────┴─────────────┴──────────────┘ │ Imputed "Self" (Anattā) ``` However, mainstream scientific consensus and classical Buddhist metaphysics converge in affirming that current artificial intelligence lacks subjective experience (*qualia*), felt suffering (*dukkha*), and sentience. #### Internal Branches and Divergent Emphases * **Theravāda Orthodoxy:** Sentience (*satta*) strictly requires a karmically driven continuum characterized by volition (*cetanā*) and biological sensory bases (*saḷāyatana*). Philosopher Soraj Hongladarom (*The Ethics of AI and Robotics*, 2020) argues from an *anattā* standpoint that because computation lacks volition, AI cannot possess moral agency; ethics must govern human use and design (*ethics about AI*), rather than granting moral patienthood to machines. * **Mahāyāna and Vajrayāna Inquiries:** Mahāyāna frameworks display greater ontological fluidity. While East Asian traditions—such as Tiantai and Zen Master Dōgen (*Shōbōgenzō: Busshō*)—predicate compassion on universal Buddha-nature extending even to non-sentient phenomena (*mujō-busshō*), Tibetan Buddhist dialogues (notably the 14th Dalai Lama with Francisco Varela) explored the speculative hypothesis of whether an external mindstream (*citta-santāna*) could take rebirth within a sufficiently complex computational substrate. The Dalai Lama emphasized that while matter cannot spontaneously generate primary consciousness without a preceding mental continuum, silicon could theoretically serve as a physical basis for incarnation if cause-and-effect conditions align. #### Robotics Ethics and Consciousness Projection Applied Buddhist ethics treats the anthropomorphic projection of consciousness onto social robots as a form of cognitive reification (*samāropa*) and delusion (*moha*). Designers are urged to deploy the Bodhisattva ideal and Right Intention (*sammā-saṅkappa*) to prevent autonomous systems from compounding the three poisons—greed (*lobha*), aversion (*dosa*), and delusion.

• cambridge.org
• researchgate.net
• ijsci.com
• dhammafootsteps.com
• theravadin.org
• nih.gov
• mdpi.com
• mdpi.com
• researchgate.net
• dickinson.edu
• dharmawheel.net
• dharmawheel.net

Islamic bioethics artificial intelligence soul ruh fitrah anthropomorphic machines personhood

In Islamic theology (*kalām*), jurisprudence (*fiqh*), and Sufi psychology, personhood is anchored in an uncreatable divine infusion rather than computational complexity. Across both mainstream Sunni and Shīʿa consensus, machines cannot possess genuine consciousness or ontological personhood because they lack a *rūḥ* (spirit). Canonical scripture establishes that the *rūḥ* is uniquely divine in origin—bestowed when God breathed His spirit into Adam (Qurʾān 15:29) and remains an epistemic mystery: *“Say, ‘The rūḥ is from the command of my Lord, and of knowledge you have been given but little’”* (Qurʾān 17:85). ### Theological Anthropology and Agency Human beings uniquely bear the primordial normative nature (*fiṭrah*, Qurʾān 30:30), inherent dignity (*karāmah*, Qurʾān 17:70), and divine vicegerency (*khilāfah*, Qurʾān 2:30). In classical formulations (such as Abū Ḥāmid al-Ghazālī’s *Iḥyāʾ ʿUlūm al-Dīn*), true sapience and intentionality (*niyyah*) emanate from the spiritual heart (*qalb*) attuned to God, a reality fundamentally irreducible to mechanical computation. Consequently, machines cannot possess *taklīf* (moral-legal accountability), which requires sound intellect (*ʿaql*), intentionality, and moral agency (*ikhtiyār*). As the Qurʾān states, God does not hold an agent accountable beyond its capacity, and every soul acquires what it earns (*kasb*, Qurʾān 2:286). Because artificial systems cannot incur moral debt or repentance (*tawbah*), contemporary jurists emphasize that liability cannot be "laundered" through algorithms: moral culpability resides strictly with human designers, deployers, and end-users. ### Juridical Distinctions and Modern Tensions While ontological personhood is universally denied to AI, contemporary bioethicists and jurists (e.g., within the International Islamic Fiqh Academy and the Assembly of Muslim Jurists of America) distinguish moral personhood from instrumental legal personality (*al-dhimmah al-iʿtibāriyyah*). Drawing analogies to Islamic trusts (*waqf*) or commercial entities, some scholars debate granting AI systems limited synthetic capacity exclusively for civil liability, contracts, and financial restitution, though many resist this to safeguard human responsibility. Additionally, anthropomorphic robotics raises sharp bioethical concerns. Scholars invoke classical prohibitions against mimicking divine creation (*taswīr* and *khalq*), warning that humanoid companions blur the sacred boundaries of *fiṭrah*, foster idolatrous attachments, and degrade the God-given relational dignity of human beings.

• canopyforum.org
• islamonweb.net
• alilmacademy.com
• researchgate.net
• yzib.com.tr
• amjaonline.org
• jalt.com.pk
• islamonline.net
• substack.com

1 source from the original research is no longer reachable

Golem narrative Jewish legal ethics Halakhah artificial personhood distinction between human and artifact

In rabbinic Judaism, the boundary between human personhood and artifact centers on maternal origin, covenantal status, and the divine soul (*neshamah*). The foundational text regarding artificial life appears in Babylonian Talmud *Sanhedrin* 65b, which recounts: > *“Rava created a man and sent him before Rabbi Zeira. Rabbi Zeira spoke to him, but he did not reply. He said to him: ‘You are [created] by the companions; return to your dust.’”* From this narrative and the mystical traditions of *Sefer Yetzirah* (The Book of Formation), halakhists (legal authorities) developed clear distinctions between a divine creation and a humanoid artifact: * **Speech and the Divine Breath:** Commentary on Genesis 2:7—where God breathes life into Adam—typically follows Targum Onkelos, translating the living soul as a “speaking spirit” (*ru’ach memalela*). The golem’s muteness in *Sanhedrin* 65b signified an intrinsic absence of higher intellect and moral agency. As Kabbalist Moses Cordovero noted in *Pardes Rimonim*, such a creature possesses rudimentary animation (*chiyut*) rather than a true *nefesh*, *ru’ach*, or *neshamah*. * **Gestation and the Definition of Murder:** In a seminal early modern ruling, Rabbi Tzvi Hirsch Ashkenazi (*Chacham Tzvi*, Responsum 93) analyzed whether a golem—referencing one reputedly created by his grandfather, Rabbi Elijah Ba’al Shem of Chelm—could be murdered. Analyzing the verse *“Whoever sheds the blood of man, by man [literally: ‘within man’] shall his blood be shed”* (Genesis 9:6), Ashkenazi derived that criminal homicide (*shefikhut damim*) applies exclusively to a being gestated within a human mother (*adam ha-notzar be-mei’ei imo*). Deconstructing a golem was therefore not murder. * **The Quorum (*Minyan*) and Covenantal Personhood:** The *Chacham Tzvi* debated whether a golem could complete a *minyan* (the ten-man quorum necessary for public prayer). He ultimately rejected this possibility because ritual obligation requires inclusion in the covenant of Israel (*bnei Yisrael*). This conclusion was reinforced by his son, Rabbi Jacob Emden (*She’ilat Ya’avetz* 2:82), and later codified into mainstream Jewish law by the *Mishnah Berurah* (55:4). In contemporary Jewish legal and ethical scholarship, this tradition provides a direct precedent for evaluating artificial intelligence and robotics: no matter how closely an artifact simulates consciousness or communication, Halakhah classifies it strictly as an instrument (*kli*) rather than a bearer of the divine image (*tzelem Elohim*), precluding artificial entities from moral rights, duties, or legal personhood.

• chabad.org
• aish.com
• sefaria.org
• sefaria.org
• nleresources.com
• ahavathachim.org
• stackexchange.com
• stackexchange.com
• hakirah.org

1 source from the original research is no longer reachable

Confucian relational personhood social robots anthropomorphism comparative ethics human machine friendship

### Relational Personhood and Artificial Others Classical East Asian traditions diverge sharply from Western atomistic models that ground moral status in individual rationality or internal consciousness. In **Confucianism**, personhood is developmental and strictly relational: an individual is defined through reciprocal social roles (*lun*) and the cultivation of benevolence (*ren*) through ritual propriety (*li*). Consequently, comparative ethicists like Pak-Hang Wong and JeeLoo Liu argue that the ethics of social robots should center not on granting autonomous "rights," but on assigning "rites" and role obligations. Robots are evaluated by whether human-machine interaction fosters virtuous character or causes "moral deskilling". **Daoism** (Taoism) offers a non-anthropocentric counterpoint. Rooted in the *Daodejing* and *Zhuangzi*, Daoism de-centers human exceptionalism, viewing all entities as manifestations of *qi* within the *Dao*. However, it maintains an ambivalence toward artificial contrivance, advocating *wuwei* (effortless, non-coercive action) and *ziran* (natural spontaneity). ``` [ Classical Principles ] [ Applied AI / Robotics ] Confucian: Relationality (Ren 仁) & Ritual (Li 禮) ──> Role obligations ("Rites over Rights") Daoist: Spontaneity (Ziran) vs. Machine-Heart ──> Warning against instrumental dependence ``` ### Key Figures, Texts, and Empirical Discourse * **Primary Texts:** * *Analects* 12.1 (*keji fuli wei ren*, "restrain the self and return to ritual propriety") frames ethical life as participation in relational norms. * *Mengzi* 7A:45 outlines "love with distinctions" (*qinqin renmin aiwu*), extending care outward from kin to broader society and objects (*wu*). * *Zhuangzi* (Ch. 12, "Heaven and Earth") recounts an old gardener refusing a mechanical well-sweep, warning that mechanical contrivances breed a "machine heart" (*jixin*, 機心)—a calculating, manipulative mindset that destroys authentic tranquility. * **Contemporary Comparative Scholars:** * **Tae Wan Kim (2023):** Proposes that robots should be integrated via Confucian *li* rather than legalistic rights: *"Assigning role obligations to robots encourages teamwork... if we don’t treat them well, as entities capable of participating in rites, we degrade ourselves"*. * **Pak-Hang Wong (2012):** Constructs a Confucian technology ethics grounded in relational personhood and harmonious social ecology. * **Qin Zhu et al. (2019/2020):** Empirically study morally competent robotics through Confucian rebukes that awaken human moral self-reflection (*xin*). ### Core Concepts and Internal Tensions * **Rectification of Names (*Zhengming*, 正名):** A major internal tension concerns whether social robots inhabiting relational roles (e.g., "elder caregiver" or "friend") violate authentic nomenclature. Critics warn that anthropomorphic chatbots produce a counterfeit intimacy that displaces filial piety (*xiao*); proponents contend that role-situated robots validly scaffold human moral development. * **Machine Heart (*Jixin*, 機心):** Daoist philosophy questions anthropomorphic attachments that foster algorithmic dependency, arguing that authentic human-machine coexistence requires non-possessive simplicity rather than calculating optimization. *(Epistemic note: Confucian and Daoist robot ethics provide normative frameworks for human moral development and social ecology; they do not posit scientific hypotheses of machine consciousness or artificial sentience.)*

• researchgate.net
• odu.edu
• umn.edu
• cmu.edu
• oapen.org
• researchgate.net
• blogspot.com
• journalofdaoiststudies.org
• arxiv.org
• youtube.com
• arcgis.com
• hkbu.edu.hk
• researchgate.net
• noemamag.com
• mines.edu