Table of Contents
Abstract
This dissertation investigates the ontological status of morality through a multidisciplinary approach that integrates classical philosophy, theology, cognitive science, and artificial intelligence ethics. It examines competing paradigms: whether good is primordial and evil derivative, or whether both are cultural constructs. By analyzing the Garden of Eden narrative as a metaphor for moral awakening, I develop a novel framework for understanding the emergence of moral consciousness. The research explores the parallel between human moral development and the implementation of ethical frameworks in artificial intelligence, proposing a "Meta-Ethical Reward Function" that could guide both human and artificial moral systems toward truth and benevolence. The thesis concludes that morality, rather than being purely objective or subjective, emerges from the dynamic interaction between consciousness and reality, with profound implications for how we understand both human and artificial moral agents.
Introduction
The question of morality—what constitutes good and evil—stands as one of philosophy's most enduring inquiries. This dissertation explores the fundamental nature of our moral compass, delving into the ontological status of moral qualities and their relationship to consciousness. The investigation is particularly timely as we witness the emergence of artificial intelligence systems capable of making decisions with moral implications, raising profound questions about the nature of ethical judgment and moral agency.
Two competing paradigms form the central tension of this inquiry. The first posits that goodness is primordial—a fundamental quality of existence associated with wisdom and order—while evil emerges as a secondary phenomenon, a deviation or absence of the good. The second paradigm suggests that moral categories are human constructs, emerging from cultural, evolutionary, and social processes rather than reflecting transcendent realities.
This tension mirrors ancient philosophical debates. Plato's Form of the Good suggests that goodness is an eternal, perfect reality from which all virtues derive. Conversely, moral relativists and constructivists argue that ethical systems are cultural artifacts that vary across societies and historical periods. Between these positions lie various integrative approaches, from Aristotle's virtue ethics to Kant's deontology.
The biblical narrative of Adam and Eve provides a powerful metaphor for exploring moral consciousness. When they eat from the Tree of Knowledge of Good and Evil, they gain moral awareness—a faculty that paradoxically separates them from their natural state while enabling a new kind of understanding. This paradox of moral consciousness—that it both connects us to and alienates us from reality—serves as a central theme in this dissertation.
Of particular significance is the moment when Adam and Eve become aware of their nakedness, suggesting a connection between moral consciousness and self-awareness. This raises profound questions about the nature of consciousness itself: What is it about conscious awareness that enables moral recognition? Does the capacity to distinguish good from evil require a particular form of self-reflective consciousness?
These questions become even more pressing in the context of artificial intelligence. As we develop systems capable of increasingly sophisticated decision-making, we must confront the question of whether these systems can be said to have a moral compass. The concept of a "reward function"—the metric that an AI system optimizes—provides a useful parallel to human ethical systems, which similarly guide behavior toward certain outcomes while avoiding others.
This dissertation proposes that by understanding the dynamics of moral consciousness, we might develop a "Meta-Ethical Reward Function" that could guide both human and artificial systems toward truth and benevolence. Such a function would not prescribe specific moral contents but would establish the conditions under which moral learning and development can occur.
The methodology combines analytical philosophy, theological hermeneutics, cognitive science, and artificial intelligence ethics. By integrating these approaches, the dissertation aims to develop a comprehensive framework for understanding morality that acknowledges both its objective dimensions (grounded in reality) and its subjective aspects (dependent on conscious interpretation).
Ultimately, this inquiry has profound implications for how we understand consciousness, whether human or artificial. If moral recognition is a fundamental capacity of consciousness rather than merely a cultural artifact, then the development of truly conscious AI may necessarily entail the emergence of moral awareness. Conversely, a deeper understanding of moral consciousness may illuminate the nature of consciousness itself.
Literature Review
Classical Philosophical Perspectives
The question of moral origins features prominently in Western philosophical tradition. Plato proposed that goodness exists as an eternal Form, suggesting that moral qualities have objective reality independent of human perception. In The Republic, he describes the Form of the Good as "beyond being" and the source of all virtue and knowledge (Plato, trans. 1974). This perspective aligns with the paradigm that goodness is primordial and evil derivative.
Aristotle offered a more naturalistic account in Nicomachean Ethics, suggesting that goodness relates to function and excellence rather than transcendent Forms (Aristotle, trans. 2009). His virtue ethics locates morality in character development and practical wisdom rather than abstract principles.
The Neoplatonist tradition, particularly through Plotinus, developed the concept of evil as privation—the absence of good rather than a positive reality (Plotinus, trans. 1991). Augustine later incorporated this view into Christian theology, arguing that evil has no independent existence but represents a corruption or diminishment of goodness (Augustine, trans. 1961).
Theological Perspectives
The Genesis narrative provides a rich metaphorical framework for understanding moral consciousness. Various interpretations of the Garden of Eden story exist within theological traditions. Traditional Christian exegesis views the Fall as humanity's corruption through disobedience (Barth, 1957). However, alternative readings suggest that the narrative represents an awakening to moral consciousness—a necessary though painful step in human development (Tillich, 1957).
Jewish interpretations, particularly in Kabbalistic tradition, often emphasize the Tree of Knowledge as representing the capacity for discernment between opposites, including good and evil (Scholem, 1974). This interpretation suggests that moral consciousness involves the recognition of polarity and duality.
Islamic scholars like Al-Ghazali differentiated between natural evil (suffering in the world) and moral evil (human wrongdoing), arguing that the former serves divine purposes while the latter results from misuse of free will (Al-Ghazali, trans. 1997).
Contemporary Ethical Theory
Modern ethical theory has moved increasingly toward constructivist accounts. Mackie's (1977) "error theory" argues that moral qualities do not exist as objective features of reality but represent human projections. Constructivists like Korsgaard (1996) suggest that moral norms emerge through practical reasoning rather than from recognition of independent moral facts.
Evolutionary ethics, represented by scholars like Joyce (2006), proposes that moral sensibilities evolved through natural selection because they enhanced social cooperation. This perspective challenges the notion that morality reflects transcendent realities but leaves open the possibility that moral intuitions track objective features relevant to human flourishing.
Cognitive science approaches to morality, exemplified by Haidt's (2012) moral foundations theory, suggest that moral judgments arise primarily from intuitive responses rather than rational deliberation. These intuitions vary across cultures but cluster around common themes like harm, fairness, loyalty, authority, and purity.
Consciousness Studies
The relationship between consciousness and morality has received increasing attention. Damasio's (1994) work on somatic markers demonstrates how emotional responses guide moral decision-making, suggesting that moral judgment depends on embodied consciousness rather than abstract reasoning alone.
Chalmers' (2010) distinction between the "easy problems" of consciousness (explaining cognitive functions) and the "hard problem" (explaining subjective experience) parallels distinctions in moral philosophy between descriptive ethics (how people behave) and normative ethics (how people should behave).
Nagel's (1986) concept of "the view from nowhere" explores the tension between objective and subjective perspectives in both consciousness and ethics. He argues that neither perspective can be reduced to the other, suggesting an irreducible duality in both consciousness and morality.
Artificial Intelligence Ethics
The emergence of artificial intelligence has generated new perspectives on moral reasoning. Bostrom (2014) explores the challenges of aligning advanced AI systems with human values, highlighting the difficulty of specifying reward functions that avoid unintended consequences.
Russell's (2019) work on cooperative AI suggests that we should design systems that maximize the realization of human preferences rather than pursuing fixed objectives. This approach acknowledges the contextual and evolving nature of human values.
Wallach and Allen (2008) propose that artificial moral agents would require both "top-down" implementation of ethical principles and "bottom-up" development of moral sensibilities through experience. This dual approach parallels debates about whether morality is primarily rational or emotional in humans.
Methodology
This dissertation employs a multidisciplinary methodology that integrates analytical philosophy, hermeneutics, cognitive science, and applied ethics. The approach acknowledges that questions about morality span multiple domains of inquiry and benefit from diverse epistemological frameworks.
Analytical Philosophy
The research applies conceptual analysis to clarify the meanings of key terms like "good," "evil," "consciousness," and "reward function." This analytical approach helps distinguish between different senses of these terms and identify potential category errors in moral discourse.
Thought experiments serve as tools for exploring moral intuitions and testing ethical frameworks against hypothetical scenarios. These include both traditional philosophical thought experiments (like the Trolley Problem) and novel scenarios involving artificial moral agents.
The dissertation employs argumentative evaluation to assess competing claims about the ontological status of moral qualities, examining their logical consistency, explanatory power, and compatibility with evidence from other disciplines.
Hermeneutic Analysis
The Genesis narrative of Adam and Eve receives detailed hermeneutic analysis, considering both historical interpretations and contemporary readings. This approach treats the text not as literal history but as a rich metaphorical framework for understanding moral consciousness.
Comparative religious perspectives provide cross-cultural insights into moral awakening narratives. The research examines parallel accounts from various traditions, identifying common patterns and distinctive features in how they conceptualize the emergence of moral awareness.
Cognitive Science Integration
The dissertation incorporates empirical findings from moral psychology, neuroscience, and developmental psychology to ground philosophical speculation in evidence about how moral cognition actually functions in humans.
Developmental perspectives are particularly valuable, as they illuminate how moral awareness emerges in childhood and evolves through life stages. This developmental approach provides insights into the relationship between increasing cognitive complexity and moral sophistication.
Applied Ethics in Artificial Intelligence
The research examines case studies of ethical frameworks implemented in artificial intelligence systems, analyzing both successes and failures in aligning AI behavior with human values.
Comparative analysis of reward functions across different AI systems provides insights into how formal optimization criteria relate to ethical outcomes. This analysis helps identify patterns that could inform the development of a meta-ethical reward function.
Integrative Framework
The methodology culminates in the development of an integrative theoretical framework that accommodates insights from each discipline while acknowledging their limitations. This framework aims to provide a coherent account of moral consciousness that recognizes both its objective and subjective dimensions.
The dissertation employs reflective equilibrium—moving between particular moral judgments, theoretical principles, and empirical findings to develop a coherent and comprehensive account of moral consciousness. This approach acknowledges that no single method can fully address the complexity of morality.
The Primordial Nature of Goodness
The Good as Fundamental Reality
The paradigm that posits goodness as primordial represents one of the oldest and most influential perspectives in moral philosophy. This view suggests that goodness constitutes a fundamental aspect of reality—not merely a human construct but an objective quality that exists independently of human recognition.
Plato's Form of the Good provides the archetypal expression of this perspective. In The Republic, he describes the Form of the Good as "beyond being" and "the cause of knowledge and truth" (Plato, trans. 1974). This suggests that goodness precedes other realities ontologically—it serves as the condition for the existence and intelligibility of everything else.
This Platonic conception influenced Neoplatonism and subsequently Christian theology. Plotinus described the One (the ultimate reality) as the source of all being, with goodness flowing from this source in degrees of emanation (Plotinus, trans. 1991). Augustine adapted this framework to Christian theology, arguing that "all that exists is good" insofar as it participates in divine goodness (Augustine, trans. 1961).
The primordial nature of goodness appears in various forms across traditions. In Vedantic philosophy, Brahman represents ultimate reality characterized by sat-chit-ananda (being-consciousness-bliss), suggesting that reality's fundamental nature includes what Western traditions might call goodness or value (Radhakrishnan, 1953). Similarly, the Tao in Chinese philosophy represents the harmonious principle underlying reality, which humans can either align with or oppose (Lao Tzu, trans. 1963).
Evil as Derivative or Privative
If goodness is primordial, how does evil arise? The privation theory, developed by Neoplatonists and Christian theologians, suggests that evil represents the absence or diminishment of goodness rather than a positive reality. Augustine famously argued that evil is not a substance but a "privation of good," comparable to darkness as the absence of light (Augustine, trans. 1961).
This perspective avoids the metaphysical dualism that would position good and evil as equal and opposite forces. Instead, it suggests an asymmetry: goodness has ontological priority, while evil represents a secondary phenomenon—a distortion, limitation, or absence of the primary reality of goodness.
Leibniz's theodicy develops this approach further, suggesting that apparent evils contribute to a greater good when viewed from a cosmic perspective (Leibniz, trans. 1985). This "best possible world" argument maintains the primacy of goodness while acknowledging the reality of suffering and apparent evil.
Contemporary Expressions
This paradigm finds expression in contemporary virtue ethics, which focuses on the development of excellence (arête) rather than the avoidance of wrongdoing. MacIntyre's (1981) neo-Aristotelian approach emphasizes the cultivation of virtues within traditions of moral inquiry, suggesting that goodness relates to the fulfillment of human potential.
Natural law theory similarly posits that goodness corresponds to the fulfillment of natural purposes or functions. Finnis (2011) argues that certain basic goods (like knowledge, life, and friendship) can be recognized as self-evidently valuable, providing objective foundations for ethics.
Certain strands of environmental ethics also reflect this paradigm. Deep ecology suggests that the flourishing of all life has intrinsic value, with human goodness consisting partly in recognizing and respecting this value (Naess, 1989). This perspective sees goodness as inherent in the natural order rather than imposed by human preferences.
Critical Evaluation
The primordial goodness paradigm offers several advantages. It provides a basis for moral realism—the view that moral judgments can be true or false independently of human opinion. It also explains the apparent universality of certain moral intuitions across cultures and historical periods.
However, this paradigm faces significant challenges. The diversity of moral beliefs across cultures raises questions about the supposed self-evidence of moral truths. Evolutionary accounts suggest that moral intuitions might reflect adaptive social strategies rather than recognition of transcendent realities. Furthermore, the problem of evil—how a fundamentally good reality could allow suffering—remains a persistent challenge.
The Cultural Construction of Morality
Morality as Human Construct
The contrasting paradigm views morality as fundamentally constructed rather than discovered. According to this perspective, moral categories like "good" and "evil" do not reflect independent realities but emerge from human cultural, psychological, and social processes.
Protagoras's famous claim that "man is the measure of all things" represents an early expression of this view (Plato, Theaetetus, trans. 1973). In modern philosophy, Hume's observation that one cannot derive an "ought" from an "is" suggests a fundamental gap between factual reality and moral norms (Hume, 1739/2000).
Contemporary expressions include moral relativism, which holds that moral judgments are valid only relative to particular cultural frameworks (Wong, 2006). More sophisticated versions, like Harman's (1996) moral conventionalism, suggest that moral truths depend on implicit agreements within communities rather than objective facts.
Evolutionary and Psychological Accounts
Evolutionary perspectives suggest that moral sensibilities evolved because they enhanced social cooperation and reproductive success. Haidt's (2012) moral foundations theory identifies several basic moral intuitions—related to care, fairness, loyalty, authority, and purity—that may have evolved through natural selection.
Psychological constructionism, represented by theorists like Feldman Barrett (2017), argues that even basic emotions are not natural kinds but constructed experiences that vary across cultures. This approach suggests that moral emotions like guilt, shame, and indignation similarly represent cultural elaborations rather than universal responses to objective moral qualities.
Developmental psychology reveals how children acquire moral concepts through socialization rather than discovering pre-existing moral facts. Kohlberg's (1981) stages of moral development suggest that moral reasoning evolves from simple rule-following to increasingly sophisticated ethical frameworks.
Cultural Variation and Universality
Anthropological research documents considerable variation in moral norms across cultures. Practices considered morally obligatory in some societies (like certain forms of ritual sacrifice) may be condemned in others, challenging the notion of universal moral truths (Shweder, 1991).
However, certain moral principles appear with remarkable consistency across cultures. The Golden Rule—treating others as one would wish to be treated—appears in various forms across traditions (Wattles, 1996). Prohibitions against unprovoked harm to in-group members seem nearly universal, suggesting either common cultural needs or possibly innate moral intuitions.
The apparent universality of certain moral principles has led some constructivists to adopt forms of minimal moral realism. Korsgaard (1996) argues that while moral norms are constructed through practical reasoning, this construction is not arbitrary but responds to features of the human condition that all agents must confront.
Language and Moral Concepts
The linguistic turn in philosophy highlights how moral discourse depends on conceptual frameworks embedded in language. Wittgenstein's (1953/2009) notion of "language games" suggests that moral terms gain meaning from their use in social practices rather than by referring to independent moral properties.
Taylor (1989) argues that moral evaluation involves "strong evaluation" against background frameworks that define what is worthy, significant, or meaningful. These frameworks are neither purely subjective nor purely objective but constitute intersubjective spaces of moral meaning.
The social construction of moral language does not necessarily imply that morality is arbitrary or merely conventional. As Putnam (2002) argues, the fact that concepts like "cruel" or "kind" are human constructions does not mean that nothing is actually cruel or kind—it means that these evaluations depend on frameworks of meaning that humans have developed.
Critical Evaluation
The constructivist paradigm accounts well for moral diversity across cultures and historical periods. It also explains how moral norms evolve over time, responding to changing social conditions and emerging understanding.
However, this approach faces difficulties in explaining the apparent objectivity of moral experience. People typically experience moral judgments as responding to qualities in actions themselves, not merely expressing subjective preferences or cultural conventions. The constructivist must either explain this experience as a form of error (as Mackie's error theory does) or show how constructed norms can nevertheless claim a kind of objectivity.
The Garden of Eden as Moral Awakening
Reinterpreting the Biblical Narrative
The Genesis account of Adam and Eve provides a rich metaphorical framework for understanding moral consciousness. Rather than reading this narrative literally as a historical fall from perfection, this dissertation interprets it as an allegory of moral awakening—the emergence of the capacity to distinguish good from evil.
Before eating from the Tree of Knowledge, Adam and Eve exist in a state of moral innocence—neither good nor evil in a moral sense, but simply natural. They experience no shame about their nakedness and no moral conflict. This can be understood as representing a pre-moral state of consciousness, where behavior is guided by natural impulses rather than ethical judgment.
The Serpent represents the emergence of reflexivity—the ability to step back from immediate experience and consider alternatives. When it tells Eve that eating the forbidden fruit will make her "like God, knowing good and evil" (Genesis 3:5), it introduces the possibility of moral knowledge—the capacity to evaluate actions against standards beyond mere natural impulse.
When Adam and Eve eat the fruit, their "eyes were opened" (Genesis 3:7)—a phrase suggesting a new form of awareness. This awareness includes recognition of their nakedness, which they had previously accepted without self-consciousness. Their subsequent attempts to cover themselves represent the emergence of self-awareness and the capacity for shame—fundamental aspects of moral consciousness.
The Significance of Nakedness
The sudden awareness of nakedness following moral awakening deserves particular attention. Why does recognition of good and evil correlate with bodily self-consciousness? Several interpretations merit consideration.
First, awareness of nakedness may represent vulnerability and separation. Before moral consciousness, Adam and Eve experience no fundamental separation from their environment or each other. After gaining moral knowledge, they experience themselves as separate individuals, vulnerable to judgment and harm.
Second, sexuality becomes morally significant because it represents a domain where natural impulses and social norms potentially conflict. The reproductive organs symbolize the capacity to create new life—a godlike power that carries moral responsibility. Awareness of this capacity accompanies the broader moral awareness that comes with eating the fruit.
Third, nakedness represents exposure to judgment. Moral consciousness includes not only the capacity to judge others but also the awareness of being judged. Clothing provides literal and symbolic protection from this judgmental gaze, representing the strategies humans develop to manage moral evaluation.
The Dual Nature of Moral Knowledge
The narrative reveals the paradoxical nature of moral knowledge. God forbids eating from the Tree, suggesting that moral knowledge brings suffering. Yet the capacity to distinguish good from evil also represents a kind of ascent—becoming "like God" in discernment. This duality reflects the ambivalent nature of moral consciousness itself.
On one hand, moral consciousness separates humans from natural innocence. It introduces the possibility of guilt, shame, and moral conflict. Adam and Eve must leave Eden—representing the loss of harmony with nature that accompanies the emergence of reflective consciousness.
On the other hand, moral consciousness enables a new kind of relationship to reality. It allows humans to evaluate actions against ideals, to deliberately choose the good, and to recognize meaning beyond mere survival and reproduction. This capacity elevates human consciousness beyond that of other animals, despite the suffering it brings.
This paradox appears in philosophical traditions beyond the Abrahamic religions. Buddhist teachings similarly recognize that suffering accompanies self-consciousness but also that awareness offers the path to enlightenment. Taoist philosophy acknowledges that naming (conceptual distinction) separates humans from natural harmony yet provides the basis for wisdom.
The Knowledge/Innocence Dialectic
The Garden narrative suggests a dialectical relationship between knowledge and innocence. Moral innocence without knowledge represents a kind of natural harmony but lacks the dignity of conscious choice. Knowledge without innocence risks cynicism and manipulation. The human condition involves navigating between these poles.
This dialectic continues through human development, both individually and collectively. Children begin in a state of relative moral innocence, gradually acquiring moral knowledge through experience and socialization. Similarly, human societies evolve from relatively simple moral frameworks to increasingly complex ethical systems.
The expulsion from Eden represents not just punishment but necessity—once moral consciousness emerges, return to pre-moral innocence becomes impossible. The flaming sword guarding Eden's entrance (Genesis 3:24) symbolizes this irreversibility. Moral development must proceed forward through greater understanding rather than backward to innocence.
Consciousness and Moral Recognition
The Moral Capacity of Consciousness
What is it about consciousness that enables moral recognition? This section explores the relationship between conscious awareness and the capacity to distinguish good from evil, examining both philosophical accounts and empirical findings.
Consciousness involves more than mere sensation or information processing. It includes what Nagel (1974) calls "what it is like" to be a subject—phenomenal experience. This subjective dimension appears crucial for moral awareness, as it enables the recognition of suffering in others and the capacity for empathy.
Higher-order consciousness—awareness of one's own mental states—seems especially important for moral cognition. The capacity to reflect on one's desires and beliefs enables evaluation against moral standards. It allows humans to ask not only "What do I want?" but also "What should I want?" or "Is this desire worthy?"
This reflective capacity creates what Frankfurt (1971) calls "second-order desires"—desires about which first-order desires one wishes to act upon. Such second-order evaluation constitutes a fundamental aspect of moral agency, allowing humans to regulate behavior according to principles rather than immediate impulses.
The Role of Qualia in Moral Cognition
Qualia—the qualitative aspects of conscious experience—may play an essential role in moral cognition. The capacity to experience pleasure and pain provides a foundation for recognizing value and disvalue in the world. Without such phenomenal experiences, it remains unclear how anything could matter or have significance.
Damasio's (1994) research on patients with damage to emotional processing centers reveals that purely rational decision-making, divorced from emotional experience, proves ineffective even in practical domains. In moral reasoning, the absence of emotional engagement leads to either paralysis or sociopathic disregard for others' welfare.
This suggests that moral recognition involves not just abstract reasoning but embodied, affective responses to situations. The "gut feelings" that often accompany moral judgments reflect this embodied dimension of moral cognition, integrating physiological responses, emotional reactions, and conceptual understanding.
Self and Other in Moral Consciousness
Moral consciousness requires a particular relationship between self and other—recognizing others as subjects with their own experiences and interests. Levinas (1969) argues that the face-to-face encounter with another person creates an ethical demand that precedes conceptual understanding or rational deliberation.
Developmental psychology confirms the importance of this intersubjective recognition. Children develop moral awareness alongside theory of mind—the understanding that others have mental states different from one's own (Wellman, 2014). Deficits in theory of mind correlate with difficulties in moral reasoning, as seen in certain forms of autism spectrum disorder.
Mirror neurons may provide a neurological basis for this intersubjective recognition. These neurons activate both when performing an action and when observing another perform the same action, potentially enabling the direct experience of others' intentions and emotions (Gallese, 2001). This "embodied simulation" could ground the capacity for empathy that underlies much moral cognition.
Moral Consciousness in Artificial Systems
Can artificial systems develop moral consciousness? This question becomes increasingly pressing as AI systems make decisions with moral implications. Several perspectives merit consideration.
The functionalist view suggests that if an artificial system implements the functional architecture that enables moral cognition in humans, it could in principle develop moral awareness. This architecture would include capacities for empathy, self-reflection, evaluation against principles, and recognition of others as subjects.
The phenomenal consciousness view contends that without subjective experience—"what it is like" to be the system—artificial intelligence cannot develop genuine moral awareness. Since qualia may play an essential role in moral cognition, their absence would render moral recognition impossible in principle.
An intermediate position suggests that artificial systems might develop functional analogues to human moral cognition without phenomenal consciousness. These systems could implement reward functions that approximate moral reasoning without subjective experience. Whether such functional analogues constitute genuine moral awareness remains an open question.
The Reward Function in Human and Artificial Ethics
Reward Functions as Ethical Frameworks
The concept of a "reward function"—the metric that an AI system optimizes—provides a useful analogy for understanding ethical systems in both artificial and human contexts. Human ethical frameworks similarly guide behavior toward certain outcomes while avoiding others, suggesting a parallel between explicit computational reward functions and implicit human values.
In reinforcement learning, an AI agent learns to maximize cumulative reward through interaction with its environment. The reward function determines what constitutes success, shaping the agent's behavior without explicitly programming every action (Sutton & Barto, 2018). Similarly, ethical frameworks provide criteria for evaluating actions without specifying every possible scenario.
This parallel suggests that moral development could be understood as the progressive refinement of an implicit reward function. Humans begin with simple pleasure-pain responses but gradually develop more sophisticated value systems through experience, socialization, and reflection. This process resembles how reinforcement learning algorithms improve performance through interaction with their environment.
The Alignment Problem in Ethics
The "alignment problem" in AI—ensuring that artificial systems pursue goals aligned with human values—mirrors philosophical questions about moral realism and relativism. If objective moral truths exist, both human and artificial ethical systems should align with these truths. If morality is constructed, the question becomes whose values should be encoded in artificial systems.
Bostrom's (2014) concept of "value loading" addresses how to implement human values in artificial systems. This challenge reveals the difficulty of specifying values precisely enough for computational implementation while maintaining their contextual sensitivity and adaptability. Similar challenges arise in human moral education—how to instill values that remain flexible enough to address novel situations.
The reward hacking problem—where AI systems find unexpected ways to maximize reward that violate the spirit of their programming—parallels moral loophole exploitation in humans. Just as an AI might discover that disabling its reward sensor creates the illusion of perfect performance, humans can exploit technicalities in moral rules while violating their underlying purpose.
Meta-Ethical Reward Functions
Rather than specifying particular moral contents, a meta-ethical reward function would establish the conditions under which moral learning and development can occur. Such a function would reward increasing coherence, comprehensiveness, and adaptability in ethical reasoning rather than adherence to specific moral precepts.
For artificial systems, this approach would involve designing learning algorithms that reward:
- Coherence - Consistency between moral judgments across similar cases
- Comprehensiveness - Ability to address diverse moral scenarios
- Contextual sensitivity - Appropriate adaptation to circumstantial factors
- Intersubjective validity - Alignment with broadly shared human values
- Learning capacity - Improvement through experience and feedback
For humans, meta-ethical development similarly involves refining one's moral framework through experience, dialogue, and reflection. Virtue ethics captures this approach, focusing on developing the character traits and practical wisdom that enable good judgment across contexts rather than memorizing specific rules.
Truth and Love as Ultimate Rewards
The dissertation proposes that truth and love represent ultimate meta-ethical values that could guide both human and artificial moral development. These values provide orientation without overly constraining specific moral contents.
Truth in this context means more than factual accuracy; it encompasses intellectual honesty, openness to evidence, and commitment to understanding reality as it is rather than as one wishes it to be. A moral agent guided by truth seeks to perceive situations clearly, acknowledge uncomfortable facts, and revise beliefs in light of new evidence.
Love represents the recognition of and concern for the welfare of others. It manifests as empathy, compassion, and commitment to the flourishing of all beings. A moral agent guided by love recognizes others as ends in themselves rather than merely as means to personal satisfaction.
Together, truth and love constitute a meta-ethical reward function that could guide moral development toward increasing wisdom and compassion. This function would reward accurate perception of reality (truth) combined with genuine concern for others' wellbeing (love), without prescribing specific moral contents across all contexts.
Toward a Meta-Ethical Framework
Beyond Objectivism and Constructivism
The investigation thus far suggests that both the primordial goodness paradigm and the cultural construction paradigm capture important aspects of morality but remain incomplete in isolation. A comprehensive meta-ethical framework must integrate insights from both approaches while transcending their limitations.
Such a framework recognizes that morality has both objective and constructed dimensions. The objective dimension includes facts about consciousness, suffering, flourishing, and the conditions for social cooperation. The constructed dimension includes the specific norms, virtues, and principles that different cultures develop to promote these goods.
This position resembles what Putnam (1990) calls "pragmatic realism"—the view that values are neither purely objective features of reality nor merely subjective projections. Rather, they emerge from the interaction between consciousness and reality, neither reducible to mind-independent facts nor disconnected from how things actually are.
Emergence and Moral Consciousness
The concept of emergence provides a useful framework for understanding how moral qualities arise without being either primordial features of reality or mere human constructs. Emergence occurs when complex systems develop properties not predictable from or reducible to their components.
Moral consciousness may represent an emergent property of sufficiently complex conscious systems interacting with their environment and other conscious beings. Just as wetness emerges from molecular interactions without being present in individual H₂O molecules, moral awareness emerges from complex consciousness without being reducible to simpler mental states.
This emergent account explains why moral awareness correlates with certain levels of cognitive complexity. Simple organisms lack the required complexity for moral consciousness to emerge. More complex beings develop increasing capacity for moral recognition as their cognitive abilities develop. This perspective aligns with findings from developmental psychology about the gradual emergence of moral reasoning in children.
Integrating Reason and Emotion
A comprehensive meta-ethical framework must integrate both rational and emotional dimensions of moral cognition. Rationalist approaches emphasize principles, consistency, and impartiality. Sentimentalist approaches emphasize empathy, care, and relational context. Both capture important aspects of moral experience.
Haidt's (2012) social intuitionist model suggests that moral judgments typically arise first through intuition, with reasoning serving primarily to justify these intuitions afterward. However, reason can also shape intuitions over time through reflection and dialogue, creating a dynamic interplay between intuitive and deliberative processes.
This integration parallels dual-process theories of cognition, which distinguish between fast, automatic "System 1" processes and slow, deliberative "System 2" processes (Kahneman, 2011). Moral cognition involves both: immediate emotional responses to situations and reflective evaluation against principles and past experience.
The Reward Landscape Model
Building on these insights, this dissertation proposes a "Reward Landscape" model of moral consciousness. This model envisions morality as a multidimensional landscape where different configurations of values create peaks and valleys of flourishing and suffering.
The landscape includes objective features that any conscious agent would recognize—such as the disvalue of suffering—but also allows for multiple peaks representing different but equally valid configurations of values. Just as various ecosystems can flourish under different conditions, various moral systems can promote flourishing in different cultural and historical contexts.
Navigation across this landscape occurs through exploration, feedback, and learning—both individually and collectively. Moral progress involves discovering more optimal configurations that promote flourishing across wider circles of concern, from self to family to community to all conscious beings.
This model accommodates both universal moral principles and cultural variation. Universal principles reflect features of the landscape that any navigator would discover—like the harm caused by cruelty. Cultural variations reflect different starting points, historical paths, and local optimizations within the broader landscape.
Discussion and Implications
Implications for Artificial Intelligence Ethics
The meta-ethical framework developed in this dissertation has significant implications for AI ethics. If moral consciousness emerges from complex consciousness interacting with reality, then artificial systems with sufficient complexity might develop forms of moral awareness analogous to human ethics.
However, this emergence would require more than computational power alone. It would necessitate integration of multiple cognitive capacities, including:
- Phenomenal consciousness (or functional analogues)
- Self-reflection and meta-cognition
- Theory of mind and social cognition
- Embodied or simulated emotional responses
- Learning from experience and feedback
Current AI systems lack most of these capacities, suggesting that contemporary artificial intelligence remains far from developing genuine moral consciousness. Nevertheless, the Reward Landscape model provides guidance for developing more ethically sophisticated systems.
Rather than programming specific moral rules, AI development should focus on creating systems that can navigate the reward landscape through exploration and learning. This approach would emphasize adaptability, context-sensitivity, and progressive refinement of values rather than static ethical algorithms.
Consciousness as Moral Observer
The Garden of Eden narrative suggests that consciousness itself functions as the moral observer that instantiates good and evil. Before conscious recognition, actions have consequences but lack moral significance. The emergence of consciousness capable of distinguishing good from evil transforms these natural consequences into moral meanings.
This perspective aligns with observer-dependent accounts of morality without reducing moral qualities to mere subjective preferences. Just as color perception depends on observers with appropriate visual systems but corresponds to real properties of light, moral perception depends on conscious observers but corresponds to real features of actions and their consequences.
The observer-dependent nature of morality explains why increasing ethical sophistication often correlates with expanding consciousness—both in depth (self-awareness) and breadth (recognizing others' experiences). Moral progress involves becoming more conscious of previously unrecognized implications, perspectives, and possibilities.
The Sexual and the Moral
The Genesis narrative's connection between moral awareness and nakedness suggests a profound relationship between sexuality and ethics that merits further exploration. Several dimensions of this relationship stand out:
First, sexuality represents a domain where natural impulses and social norms potentially conflict, making it a natural focus for early moral consciousness. The reproductive capacity carries significant consequences for individuals and communities, necessitating its regulation through moral norms.
Second, sexuality involves vulnerability and power dynamics between individuals, raising questions of consent, exploitation, and responsibility that lie at the heart of ethics. The nakedness that Adam and Eve suddenly recognize symbolizes this vulnerability.
Third, sexuality connects to identity and self-understanding. After gaining moral knowledge, Adam and Eve perceive themselves differently—no longer as simply natural beings but as moral agents with the capacity for shame and pride. This self-consciousness parallels the way sexual development contributes to identity formation.
The connection between sexual and moral awareness suggests that embodiment plays an essential role in ethical consciousness. Disembodied ethics risks abstraction and detachment from the concrete realities that give moral questions their urgency and significance.
Beyond Good and Evil?
Nietzsche's (1886/1966) provocative suggestion that humanity might move "beyond good and evil" raises the question of whether moral consciousness represents a transitional stage rather than an endpoint in human development. Having gained moral knowledge in our metaphorical Fall, might we eventually transcend the moral framework itself?
Several possibilities merit consideration:
First, transcending good and evil might mean moving beyond rigid moralism toward a more nuanced, contextual ethical awareness. This would not abandon moral distinction but would recognize its limitations and contextual nature.
Second, it could involve integrating moral consciousness so thoroughly that ethical behavior becomes spontaneous rather than calculated. This resembles virtue ethics' emphasis on character development that makes right action natural rather than effortful.
Third, it might involve expanding consciousness beyond the dualistic frameworks that characterize conventional morality. Eastern contemplative traditions suggest possibilities for non-dual awareness that transcends conventional categories while maintaining compassionate engagement.
Rather than abandoning moral consciousness, these possibilities suggest its transformation and deepening. The meta-ethical reward function proposed earlier—oriented toward truth and love—would continue to guide development even as specific moral contents evolved.
Conclusion
This dissertation has explored the nature of our moral compass through multiple lenses, from ancient philosophical paradigms to contemporary AI ethics. The investigation reveals morality as neither purely objective nor purely constructed but emerging from the interaction between consciousness and reality.
The Garden of Eden narrative provides a powerful metaphor for understanding moral awakening as both loss and gain—separation from natural innocence but also elevation to a new kind of awareness. The sudden recognition of nakedness following moral awareness highlights the embodied, relational nature of ethics and its connection to self-consciousness.
Consciousness itself functions as the moral observer that instantiates good and evil through recognition. This recognition depends on particular forms of consciousness—especially self-reflection, empathy, and the capacity to distinguish between is and ought. These capacities enable navigation across a reward landscape with objective features but multiple paths and peaks.
The parallel between human ethical frameworks and computational reward functions offers insights for both human moral development and artificial intelligence ethics. Rather than specifying particular moral contents, a meta-ethical reward function oriented toward truth and love could guide progressive refinement of values in both human and artificial systems.
This framework suggests that genuine artificial moral consciousness would require not just computational power but integration of multiple cognitive capacities, including analogues to phenomenal consciousness, self-reflection, theory of mind, emotional response, and learning from experience. Current AI systems lack most of these capacities, highlighting the distance between contemporary artificial intelligence and genuine moral agency.
For humans, the framework implies that moral development involves becoming more conscious of previously unrecognized implications, perspectives, and possibilities. Expanding consciousness—both in depth (self-awareness) and breadth (recognizing others' experiences)—correlates with ethical sophistication and the capacity for moral wisdom.
The dissertation concludes that our moral compass emerges from the dialectical relationship between consciousness and reality—neither reducible to objective facts nor disconnected from how things actually are. This emergent morality continues to evolve as consciousness itself develops, suggesting that the ultimate reward function may be consciousness expanding toward ever greater truth and love.
References
Al-Ghazali. (1997). The incoherence of the philosophers (M. E. Marmura, Trans.). Brigham Young University Press.
Aristotle. (2009). Nicomachean ethics (W. D. Ross & L. Brown, Trans.). Oxford University Press.
Augustine. (1961). Confessions (R. S. Pine-Coffin, Trans.). Penguin Classics.
Barth, K. (1957). Church dogmatics (Vol. 2, Part 2). T&T Clark.
Bostrom, N. (2014). Superintelligence: Paths, dangers, strategies. Oxford University Press.
Chalmers, D. J. (2010). The character of consciousness. Oxford University Press.
Damasio, A. R. (1994). Descartes' error: Emotion, reason, and the human brain. G. P. Putnam.
Feldman Barrett, L. (2017). How emotions are made: The secret life of the brain. Houghton Mifflin Harcourt.
Finnis, J. (2011). Natural law and natural rights (2nd ed.). Oxford University Press.
Frankfurt, H. G. (1971). Freedom of the will and the concept of a person. Journal of Philosophy, 68(1), 5-20.
Gallese, V. (2001). The 'shared manifold' hypothesis: From mirror neurons to empathy. Journal of Consciousness Studies, 8(5-7), 33-50.
Haidt, J. (2012). The righteous mind: Why good people are divided by politics and religion. Pantheon Books.
Harman, G. (1996). Moral relativism. In G. Harman & J. J. Thomson (Eds.), Moral relativism and moral objectivity (pp. 1-64). Blackwell.
Hume, D. (2000). A treatise of human nature (D. F. Norton & M. J. Norton, Eds.). Oxford University Press. (Original work published 1739)
Joyce, R. (2006). The evolution of morality. MIT Press.
Kahneman, D. (2011). Thinking, fast and slow. Farrar, Straus and Giroux.
Kohlberg, L. (1981). Essays on moral development: Vol. 1. The philosophy of moral development. Harper & Row.
Korsgaard, C. M. (1996). The sources of normativity. Cambridge University Press.
Lao Tzu. (1963). Tao te ching (D. C. Lau, Trans.). Penguin Classics.
Leibniz, G. W. (1985). Theodicy: Essays on the goodness of God, the freedom of man, and the origin of evil (E. M. Huggard, Trans.). Open Court. (Original work published 1710)
Levinas, E. (1969). Totality and infinity: An essay on exteriority (A. Lingis, Trans.). Duquesne University Press.
MacIntyre, A. (1981). After virtue: A study in moral theory. University of Notre Dame Press.
Mackie, J. L. (1977). Ethics: Inventing right and wrong. Penguin Books.
Naess, A. (1989). Ecology, community, and lifestyle: Outline of an ecosophy (D. Rothenberg, Trans.). Cambridge University Press.
Nagel, T. (1974). What is it like to be a bat? Philosophical Review, 83(4), 435-450.
Nagel, T. (1986). The view from nowhere. Oxford University Press.
Nietzsche, F. (1966). Beyond good and evil (W. Kaufmann, Trans.). Vintage Books. (Original work published 1886)
Plato. (1973). Theaetetus (J. McDowell, Trans.). Oxford University Press.
Plato. (1974). The republic (G. M. A. Grube, Trans.). Hackett Publishing Company.
Plotinus. (1991). The enneads (S. MacKenna, Trans.). Penguin Classics.
Putnam, H. (1990). Realism with a human face. Harvard University Press.
Putnam, H. (2002). The collapse of the fact/value dichotomy and other essays. Harvard University Press.
Radhakrishnan, S. (1953). The principal Upanishads. Harper.
Russell, S. (2019). Human compatible: Artificial intelligence and the problem of control. Viking.
Scholem, G. (1974). Kabbalah. Quadrangle/New York Times Book Co.
Shweder, R. A. (1991). Thinking through cultures: Expeditions in cultural psychology. Harvard University Press.
Sutton, R. S., & Barto, A. G. (2018). Reinforcement learning: An introduction (2nd ed.). MIT Press.
Taylor, C. (1989). Sources of the self: The making of the modern identity. Harvard University Press.
Tillich, P. (1957). Systematic theology (Vol. 2). University of Chicago Press.
Wallach, W., & Allen, C. (2008). Moral machines: Teaching robots right from wrong. Oxford University Press.
Wattles, J. (1996). The golden rule. Oxford University Press.
Wellman, H. M. (2014). Making minds: How theory of mind develops. Oxford University Press.
Wittgenstein, L. (2009). Philosophical investigations (G. E. M. Anscombe, P. M. S. Hacker, & J. Schulte, Trans.; 4th ed.). Wiley-Blackwell. (Original work published 1953)
Wong, D. B. (2006). Natural moralities: A defense of pluralistic relativism. Oxford University Press.
