Do we hallucinate our world? And if so, how could we know? Since it is impossible to step outside our own bodies and minds, are we condemned to never truly know whether we know? In other words, beyond all doubt about the external world, can we ever be certain that we are not deceiving ourselves? Is this the ultimate human condition—to remain ignorant of our own ignorance and knowledge?
To affirm that we know the world is as untenable as claiming it is merely a product of our own delirium. Both assertions require us to transcend the subject-object correlation. And since such transcendence is impossible, we appear to be trapped in a state of perpetual undecidability.
In this regard, we may ask: what would a consistent realist thesis look like? Certainly not the claim that “a world exists outside me (or us).” Such a statement simultaneously presupposes the subject-object correlation and the necessity of overcoming it. A more coherent thesis should inquire into reality while acknowledging the involvement of mind and subjectivity. I propose the following: “The source and content of my representations are not solely my (or a) mind.” There is something “else,” something “more.” Or, viewed from the opposite direction: representations are not “everything.” This “logical nuance” offers an alternative formulation of the classic problem of realism.
Before proceeding further, we must acknowledge that this may seem like an odd moment to advocate for realism. Have not transcendental philosophy, German idealism, phenomenology, hermeneutics, and all philosophies grounded in the linguistic turn effectively refuted realism? Realists are often dismissed as naïve, unable to recognize the mind’s role in the constitution of knowledge. Moreover, they are frequently disconnected from historicity and the complex web of conditions—language, power, class relations, and so forth—that shape our understanding.
Yet the need for realism today arises from the twilight of the linguistic turn, which proved to be but another form of subjectivism. The same critique applies to idealism, phenomenology, and hermeneutics. But why is subjectivism problematic? And could it be that subjectivism is, in fact, more internally consistent than any form of realism?
Subjectivism asserts that reality is confined to our interpretations, cultural constructions, linguistic structures, etc. “Communities” are seen as reciprocally closed, each shaped by its unique cultural destiny. Subjectivism is also detached from the pursuit of truth; it has appropriated the term only to signify subjective authenticity. But above all, subjectivism is self-defeating. If all truths are reduced to subjective interpretations, it sidesteps a crucial question: Why are all interpretations particular to begin with? What is it that all subjects share in the very act of interpreting?
Subjectivism also overlooks a critical fact—not the possibility or impossibility, but the reality—of translation among different subjectivities, histories, and languages. How is it that we still inhabit a shared world despite the multiplicity of interpretations? We have yet to fully confront the problem of how a world can be at once familiar and heterogeneous.
In essence, subjectivism is a sophisticated extension of human provincialism. The concept of the “common” must extend beyond what is shared among humans to encompass what is shared with non-human beings as well. If this thesis holds true, it raises an important question: How can human language, interpretation, and constructions possibly exhaust the real?
The central issue of realism—and idealism—lies here: in the nontrivial continuity of the world between subjectivity and objectivity, between subjects and other subjects, and between objects and other objects.
If modern philosophy can be characterized as a philosophy of the conscious subject, then postmodernism—understood in a broad sense—may be described as a philosophy of the unconscious subject. The latter was neither materialist nor realist; it was an anti-subjectivist form of subjectivism.
Yet even if our claims seem reasonable, how can one assert that a “world” exists in itself—a “reality” that neither depends on us nor requires us, yet remains accessible to our knowledge? Such a claim becomes impossible as long as we remain within the dual framework that opposes subject and object within a single relational structure, which is often referred to as correlation.
Within the bounds of correlation, it is logically inconsistent to affirm the existence of a world independent of subjectivity. If the world is defined in relation to its independence from a subject, we are left with a negative definition that still hinges on the concept of subjectivity. One cannot claim the independence of a concept—such as the real world—by starting from relationally defined terms like subject and object.
Thus, the world is defined through a process of subtraction. But such a definition is inadequate for a world presumed to be self-grounding (selbstständig) rather than merely independent (unabhängig) of the subject.
The fundamental correlation between subject and object has been conceptualized in various ways:
a) by affirming only or privileging the object—either as independent from subjectivity (naïve realism, common sense) or as absolute, self-relating substance (Spinoza);
b) by affirming only or privileging the subject—as in objective idealism (Berkeley), transcendental philosophy (Kant), or absolute subjectivity (Fichte, Schelling, Hegel);
c) by positing the insurmountable correlation between subject and object, particularly through the concept of intentionality (Brentano, Husserl).
Variations of positions (a) and (b) often involve attempts to ground one term in the other: either subjectivity grounding objectivity, or vice versa.
The former results in idealizing the world; the latter in objectifying subjectivity. But is there another way?
We cannot erase our eyes from the world. It is a fact that we cannot “reduce” the transcendental subject without falling into contradiction. Even when all representations and faculties are stripped away, some trace of subjectivity remains. This is what Eugen Fink described as the anonymous “transcendental observer” (transzendentaler Zuschauer), an eye without an eye, preceding all constituted subjectivity and objectivity. This is akin to what Heidegger termed Being (Sein).
Rather than trying to remove the eye from the world, we must instead make it real, as part of the world. Without a real subject, there could be no access to the world, and all its interpretations would be impossible. However, including the subject within the world does not mean naturalizing it—that is, reducing it to an object among others, like a stone or a planet.
Direct (or immediate) realism is naïve. It assumes that representations are either real things or, at the very least, faithful copies of the external world. In doing so, it completely disregards the role played by subjectivity in the construction of knowledge. Transcendental philosophy, by contrast, reveals the subjective conditions underlying all possible knowledge. Yet it still clings to a belief in the real world—what it calls the thing-in-itself. But if something truly exists beyond our subjectivity, how can we know it? How can it be both acknowledged and beyond our reach?
The only proposed solution is absolute idealism: the I must simultaneously function as both subject and object. All representations, in this view, are products of its activity. This implies that every representation can be traced back to some subjective process. However, to “trace back” means more than a mere conceptual move—it means to demonstrate, to construct a systematic account of knowledge, beginning with the I and culminating in every possible representation.
Affirming the necessity of an absolute I is a transcendental claim—that is, a formal condition of possibility—but it does not amount to a positive proof.
Ultimately, none of the idealist systems succeed in tracing what I find in myself (perceptions of things) back to what I produce by myself. The claim that even intuitions require some form of subjective activity does not entail that subjectivity generates intuitions—both in form and in content—in the final instance. Moreover, these systems often assume that all subjective activity is conscious and accessible to itself via self-consciousness.
They also fail to explain a key aspect of our practical life: our reliance on evidence, whether in science or law, which presupposes reference to a world not reducible to individual opinion. Crucially, contemporary science shows that human beings and intelligence are outcomes of a long evolutionary process of matter. This observation implies a paradox: we think this thought, yet in doing so, we must admit that thought itself emerges at a certain point within the universe.
This leads to a short circuit: intelligence must arise from a non-intelligent world, but we can only assert and validate this insight through intelligence. Thus, privileging either subjectivity or objectivity proves equally one-sided.
We cannot eliminate subjectivity, but neither can we expand it without limits. To assert that subjectivity is the ultimate author of the world contradicts the very concept of subjectivity. If subjectivity were absolute, our finite consciousness could never apprehend it. A drop of water in the ocean cannot say it exists, for it is indistinguishable from the whole. Yet the ocean, in this metaphor, also ceases to exist—first, because there is no distinction between the drop and the ocean; second, because nothing lies outside the ocean to define its boundary. Let us recall that the biblical Genesis begins with the separation of waters from the heavens—an act of differentiation.
The impasse lies in the formulation of the problem itself. A two-term relation cannot be reduced to either of its components: neither subject nor object alone. Moreover, a dual and symmetrical relationship does not allow us to determine what belongs to the subject and what to the object. What is needed is at least two relationships—or a third term—to permit variation. It is differentiated variance across relationships that yields information about the variables. In this sense, we need something analogous to a system of equations. Consider a simple example. Given the equation:
x + y = 8
We cannot determine the individual values of x and y from this alone. We need a second equation involving the same variables but with different values, such as:
x – y = -2
We can solve one equation for a variable: x = 8 – y
Substituting this into the second equation:
(8-y) – y = -2, which simplifies to y = 5
Substituting back: x + 5 = 8, we find x = 3
Alternatively, consider a more abstract yet simpler framework. Given two terms, A and C, we can understand their relationship by introducing two intermediary relationships: A → B and B → C. This process, known as composition, allows us to trace a relationship through a network of intermediate connections: A → B → C. Such an arrangement permits various combinations and mediations, enabling a structured and meaningful link between otherwise disjointed terms.
Consider the simple, undirected S–O (subject–object) relationship. Objectivity can only be established by introducing two morphisms: S₁ → O and S₂ → O, meaning that two observers apprehend the same object. This idea can be extended: we may introduce multiple subjects—S₁, S₂, S₃, …, Sₙ—and observe whether O remains unchanged across this variation. If so, we may claim that O is independent of these subjects.
Metaphysically, we might venture a further thesis: that the object is not only independent (unabhängig) of any given subject, but also relatively self-standing (selbstständig). Here, the language of variance and invariance becomes fruitful.
For simplicity, we can outline three fundamental types of relationships between subjects and objects:
• S₁–S₂: Intersubjectivity
• S–O: Knowledge
• O₁–O₂: Real relationships between objects (e.g., causality)
We have already used arrows to represent these relationships, implying directionality. Therefore, we can define six basic directional morphisms:
S₁ → S₂ (intersubjectivity from one side)
S₂ → S₁ (intersubjectivity from the other side)
S → O (cognitive apprehension of the object)
O → S (affection or impact of the object on the subject)
O₁ → O₂ (a real, object-to-object relation originating from O₁)
O₂ → O₁ (a real, object-to-object relation originating from O₂)
If a relationship is symmetrical—representing a bijection (in functional terms) or an invertible morphism (in category theory)—we can denote it as A ↔ B. Here, the double arrow indicates bidirectional morphisms rather than a single, undirected link.
We can now begin composing these relationships to build a structured framework—one that permits differential variation of subjects and objects. This structure allows us to discern the functional roles played by each element. In this way, we can ground realism by identifying invariants on the object’s side, while also accounting for interpretation as variation on the subject’s side. Similarly, we can detect subjective invariants through their stable responses across varying inputs.
The minimal structural expression of this relational network—the seed of a broader epistemological and ontological model—would be as follows:
The proposed structure incorporates intersubjectivity, intentionality and real relationships. Rather than seeking to escape the correlationist framework of the subject–object (S–O) relation, we should aim to integrate it within a broader network that includes diverse types of relationships. Such an approach yields a non-flat world—a world that is non-trivially interconnected and structured.
The world always reveals itself through multiple perspectives. Yet it also possesses an inner structure. The richness of subjectivity engages with the richness of objects in an intricate interplay of positions. Of course, we cannot assign definitive “values” to either term, as we might in a classical system of equations. Moreover, the categorical language demands certain axioms and structural constraints that are not easily satisfied in philosophical discourse.
Nonetheless, our reference to mathematical and categorical frameworks preserves what is essential: a network of directed relationships that allows for local variation. This, in essence, defines experience—the capacity to conduct experiments, to vary differentially a set of partially heterogeneous, interwoven elements.
At the same time, we must acknowledge the necessity of different kinds of equivalence—not merely “identity” or “difference.” In this sense, the language of category theory expands our formal universe, offering a spectrum of equivalence relations, each represented by a distinct type of arrow or morphism.
