Roundtable Discussion: "Empiricism and the Philosophy of Mind" Revisited
Overview of the Paper and its Enduring Significance
Wilfrid Sellars’ 1956 paper, "Empiricism and the Philosophy of Mind," stands as a landmark critique of a foundational assumption that has shaped much of Western philosophy: the "Myth of the Given." At its core, the paper challenges the idea that our knowledge of the world is built upon a secure base of raw, uninterpreted sensory experiences—"data"—that are directly presented to the mind and serve as the ultimate justification for all other beliefs.
Sellars argues that this concept of a "given" element in experience is a profound mistake. He contends that for something to function as a piece of knowledge, even the simplest observation like "This is green," it must already be situated within a complex conceptual framework, or what he terms the "logical space of reasons." To recognize something as green is not a passive reception of a raw quality but an active application of a concept, an ability acquired through public language and social practice. This ability presupposes a whole network of other concepts, such as understanding what colors are, knowing the conditions under which objects appear to have their true color, and being able to differentiate one color from another.
The impact of this idea on our everyday understanding of perception and knowledge is profound. We tend to think of seeing as a direct and unproblematic window onto the world. Sellars’ argument suggests this is a misleading picture. The experience of something looking green, for instance, is not a more primitive or certain piece of knowledge than the knowledge that it is green. Rather, the ability to report how things "look" is a more sophisticated linguistic skill that presupposes the concept of how things "are."
The contention surrounding these concepts arises because they dismantle the traditional empiricist picture of knowledge as a structure built from the ground up, starting with simple, infallible sensory bricks. Sellars replaces this with a holistic and self-correcting vision of knowledge as a "web of belief," where no single belief is foundational and any claim can be revised in light of new evidence and better theories. This anti-foundationalism is unsettling for those seeking absolute certainty.
To explain how we can meaningfully talk about private "inner episodes" like thoughts and sensations without treating them as "given," Sellars introduces a now-famous allegorical myth. He imagines a community of our ancestors whose language initially only refers to public objects and behaviors. A genius named Jones then postulates "thoughts" as unobservable, inner-speech-like episodes to explain intelligent behavior that occurs without overt speech. Subsequently, he postulates "impressions" as theoretical inner states to explain the qualitative character of perceptual experiences. Crucially, these inner episodes are not directly intuited "givens" but are theoretical posits, learned and understood through their role in a public theory of mind and perception. This revolutionary account attempts to preserve the reality of our inner lives while denying them the foundational epistemic status they were traditionally assigned.
Panel Members
Dr. Helena Markos, a Neo-Sellarsian philosopher, works to update Sellars' project using the tools of contemporary philosophy of language. She defends the core tenets of Sellars’ critique of the Given, his account of inner episodes as theoretical posits, and his view of knowledge as a non-foundational, self-correcting enterprise.
Prof. Adrian Llewellyn is a historian of analytic philosophy who advocates for a revised sense-datum theory. He accepts Sellars’ critique of foundationalism but argues that "sense-data" can be rehabilitated as legitimate theoretical entities needed to explain the qualitative content of perception, rather than as epistemic givens.
Dr. Penelope Kwan, a philosopher in the Wittgensteinian tradition, shares Sellars’ anti-foundationalism but is critical of his theoretical approach to the mind. She argues that puzzles about "inner episodes" are better dissolved by examining the grammatical rules and public criteria governing our use of psychological language.
Prof. Malik Rahman is a cognitive scientist who approaches these philosophical questions from the perspective of methodological behaviorism and computational modeling. He interprets Sellars’ Jonesean myth as a pioneering, pre-scientific model for postulating internal states to explain behavior, a practice now central to cognitive science.
Dr. Sofia Neves, a phenomenologist and enactivist, argues that both Sellars and his classical opponents overlook the role of the embodied, skillful agent in perception. She contends that perceptual experience is not primarily about forming inner representations or claims, but about practical engagement with a structured environment.
Prof. Quentin Desai, a naturalist philosopher in the Quinean tradition, endorses Sellars' anti-foundationalism and holism. He is skeptical of any sharp distinction between the "space of reasons" and the natural world, urging that all claims, including those about minds and norms, be integrated into our best overall scientific theory.
Dr. Beatrice Holm, a philosopher with a Kantian perspective, interprets Sellars as a key figure in updating Kant’s transcendental project. She defends the idea that a conceptual framework is a necessary condition for any possible experience, thereby rejecting the Myth of the Given, while seeking to preserve a role for non-conceptual sensory input as an enabling condition for knowledge.
The Roundtable Discussion
The discussion was organized around the central themes of Sellars' paper, beginning with his critique of the "Myth of the Given" and moving through his analyses of perceptual language, the nature of inner episodes, and the relationship between our common-sense and scientific views of the world.
Deconstructing the "Myth of the Given"
The panel opened with a consensus on the significance of Sellars' attack on epistemological foundationalism. Dr. Markos framed the issue by stating that Sellars’ core insight was that to be in the "logical space of reasons"—the domain of justification and knowledge—is to be a player in a public, normative game. Even the most basic observation report, she argued, is not a raw input but a move in this game, deriving its authority from a complex network of learned concepts and background knowledge.
Professor Llewellyn agreed that Sellars successfully demolished the idea of self-authenticating, non-inferential knowledge. However, he contended that Sellars conflated two distinct ideas in the classical tradition. "The first," Llewellyn explained, "is the epistemic thesis that sensations provide a foundational justification for empirical knowledge. This is the 'Given' Sellars rightly rejects. But the second is an ontological thesis: that to explain the qualitative character of perception, we must posit inner, private particulars—sense-data. Sellars’ critique hits the first, but the second can be reconstructed as a viable explanatory theory."
Dr. Kwan expressed skepticism about this move, suggesting it was an attempt to save a problematic picture. From a Wittgensteinian perspective, she argued, the mistake is looking for an entity (a "given" or a "sense-datum") to ground our knowledge at all. "The bedrock," she stated, "is not an object of awareness but a form of life, a set of shared practices. Our certainty in 'This is red' comes not from an infallible inner seeing, but from our mastery of a public technique for using color words."
Dr. Holm offered a Kantian interpretation, suggesting Sellars provides a linguistic and social turn on Kant's dictum that "concepts without intuitions are empty, intuitions without concepts are blind." She proposed that while there is no epistemic "given," there must be a non-conceptual sensory manifold that provides the raw material for conceptual synthesis. "This manifold is not a piece of knowledge," she clarified, "but a necessary condition for the possibility of knowledge. Sellars, perhaps, underplays this enabling role."
The Logic of 'Looks' and Perceptual Experience
The discussion then shifted to Sellars' analysis of perceptual judgments, particularly his treatment of the locution "X looks green." Dr. Markos presented the Sellarsian view: to say "X looks green" is not to report a more fundamental, incorrigible fact about an inner "appearance." Instead, it is to report an experience that is qualitatively like seeing something that is green, while withholding the full epistemic endorsement that comes with the claim "X is green." This move, she noted, situates 'looks'-talk as a sophisticated linguistic practice, not a description of a primitive datum.
Professor Llewellyn countered that this account leaves a critical explanatory gap. "What makes the experience of seeing a green object, of a blue object in funny light looking green, and a hallucination of a green object qualitatively similar?" he asked. "Sellars calls this the 'common descriptive content' but offers no positive account of it. A heterodox sense-datum theory provides one: in all three cases, the perceiver is sensing a green, non-physical particular. This explains the phenomenal similarity."
Dr. Neves rejected both frameworks as overly intellectualist. "Perception is not primarily about making or withholding propositional claims, nor is it about sensing inner replicas," she argued. "An object 'looking' a certain way is a matter of our embodied, skillful readiness to interact with it. A 'generic' look, which Sellars explains as a less determinate propositional claim, is better understood as the solicitation of a range of possible actions. The phenomenal character is tied to this bodily engagement, not to an inner picture or a linguistic token."
Professor Rahman saw a way to naturalize Sellars' account. He proposed that the authority of a report like "This is green" is a learned, metacognitive assessment of the reliability of one's own perceptual system in a given context. "'Standard conditions' are not just a list of external factors, but a set of parameters that the brain models. When the system's confidence is high, it outputs a categorical judgment. When confidence is low due to poor lighting or conflicting cues, it outputs a hedged 'looks' report. The propositional claim is the output of a reliable, self-monitoring causal mechanism."
The Status of 'Inner Episodes': The Jonesean Myth
The conversation turned to Sellars' constructive project: accounting for inner episodes like thoughts and sensations. Dr. Markos introduced the "myth of Jones," presenting it as a rational reconstruction of how a purely behaviorist community could develop a vocabulary for mental states. "Jones postulates 'thoughts' as theoretical inner-speech episodes to explain intelligent behavior," she explained. "This framework is then taught, and individuals learn to self-report these states without inference from their own behavior, granting them 'privileged access.' The crucial point is that the intersubjective, theoretical role is primary; the first-person reporting role is a later, acquired skill."
Professor Rahman endorsed this model enthusiastically, calling it a "brilliant prefiguration of the cognitive revolution." He argued, "Jones is the first cognitive scientist. He does exactly what we do: posit unobservable internal states or representations with semantic properties to explain complex behavior. The 'privacy' of thoughts is not metaphysical; it's the result of a system having direct access to its own internal processing states, which others can only infer."
Dr. Kwan, however, launched a forceful critique. "The Jones myth," she asserted, "is a compelling story, but it re-enacts the very mistake it's meant to correct. It treats our talk of 'thinking' as a theory about hidden causes, when it's better understood as a grammatical feature of our language games. The 'privileged access' we have to our own thoughts isn't a form of inner observation of a theoretical entity; it's a feature of the grammar of the word 'to think.' You don't need evidence to say 'I am in pain,' not because you're observing a private object, but because that's how the concept of pain works."
Professor Desai sided with the scientific perspective, but with a Quinean caveat. "Whether we talk of Jones' 'thoughts' or Kwan's 'grammatical rules,' the ultimate arbiter is our best overall theory of the world," he stated. "If postulating inner representational states provides the most powerful, predictive explanation of human behavior, as cognitive science suggests it does, then we have reason to believe such states exist. The talk of 'grammatical rules' risks becoming a sterile description that insulates itself from empirical revision."
Explaining 'Impressions' and the Scientific Image
The final and most contentious part of the discussion focused on "immediate experience"—sensations and impressions. Dr. Markos explained the second stage of Jones' myth, where he postulates "impressions" (e.g., an impression of a red triangle) as theoretical states of the perceiver. These are modeled on "inner replicas" but, via a commentary, are not claimed to literally be red or triangular. Their function is to explain the qualitative dimension of perception.
Professor Llewellyn argued that his own theory was superior. "Sellars' 'impressions' are ghostly theoretical states with properties defined only by analogy. A micro-ontology of sense-data, conceived as actual particulars in a phenomenal space with intrinsic structural properties mirroring color-space, provides a more robust and less metaphorical explanation."
Dr. Holm raised a critical point about the two images. "Sellars' distinction between the Manifest Image (the world of persons, things, and norms) and the Scientific Image (the world of theoretical posits like particles and fields) is crucial here. Jones' 'impressions' are an attempt to introduce a scientific concept to explain a manifest phenomenon. The danger is in confusing the two." She argued that the manifest image, as the framework in which we live as agents, has a pragmatic primacy and cannot simply be replaced by the scientific image without incoherence.
Professor Desai challenged this, suggesting the manifest image is, in effect, a folk theory. "It's a highly successful one for navigating the everyday world, but ultimately, it's revisable and, in many respects, false. Science is the measure of what is real. If our best science tells us that the world consists of fields and quarks, then colored physical objects and Jonesean impressions are, in the final analysis, useful fictions."
Dr. Markos clarified Sellars' position. "Sellars grants the scientific image primacy in the descriptive and explanatory domain. However, the manifest image is also the domain of the 'space of reasons'—of persons, norms, and intentions. These are not objects in the scientific image, but their reality is not thereby denied. The challenge is to understand how the two images can be fused into a single, stereoscopic vision."
Dr. Neves concluded the discussion by suggesting a path beyond the debate over inner posits. "The entire problem of 'impressions' arises from a Cartesian picture of a mind sealed off from the world, needing inner mediators. An enactivist approach dissolves this. The qualitative feel of red is not a property of an inner object but a feature of a particular mode of active, embodied exploration of the environment. The structure is not in a private mental space but in the lawful relations between sensory stimulation and movement. We don't need Jones' theory because we were never trapped inside our own heads to begin with."
The roundtable concluded without a definitive resolution, but with a clear map of the enduring philosophical landscape shaped by Sellars' work. The central tensions between theoretical posits and grammatical rules, the normative space of reasons and the causal space of nature, and language-centric versus body-centric accounts of the mind remain at the forefront of contemporary philosophical debate.