After advancing his theory of how sense-data create the physical world, Russell abandoned it in future works and reverted to the notion that physical objects could legitimately be inferred from sensory experience. This was partially due to scientific advances in physics and human physiology, which were asserting that perception is, in fact, caused by the effects of the physical world on our sense organs. Russell also sensed other difficulties in his theories on physics and perception. For one, his notion of sensibilia is difficult to establish: what, exactly, does it mean for there to be “unsensed sense-data”? How can there be an element of perception when no one is present to do the perceiving? Russell was unable to adequately describe his system of private and public spaces or to explain how sense-data and sensibilia interact with that system. In later work like The Analysis of the Mind (1921), Russell stops treating sense-data and the act of sensation as separate entities. He does, however, maintain the classic empiricist position that physical objects are not directly knowable; only their sensory effects (what he took to calling percepts, as opposed to sense-data) are available to us. Eventually, Russell abandoned his inquiries into the relationship of matter and perception, though he continued to work in other areas of epistemology.

Critics also found flaws with Russell’s presentation of logical atomism, which was often sketchy in areas. His former student Ludwig Wittgenstein, for example, disagreed with the idea that logical atomism should be tied to empiricism, and in his Tractatus Logico-Philosophicus Wittgenstein treated logical atomism as a purely formal theory. Critics have also taken issue with how Russell defines the “simple” elements of the world’s structure. Russell says that the basic atomic facts of the world with which we can become acquainted are of two kinds: particulars and universals. Particulars are individual instances of sense-data, whereas universals are concepts that apply to many objects. These include qualities (like redness, softness, heaviness) and temporal and spatial relations (before, on top of, next to). Russell contends that particulars and universals are atomic “simples”—that is, they are finite and individual and cannot be analyzed or broken down further. However, it is difficult to see how this definition could apply to a universal like “redness.” The very concept of redness requires us to compare different objects and classify them as similar; this being so, it is impossible for “redness” to be an independent entity.

Other critics have refuted Russell’s logical atomism in myriad ways, many of which are too complex to cover here. Although Russell’s presentation of logical atomism may have proved untenable, it remains an important moment in the history of philosophy. Like the Principia Mathematica—another of Russell’s projects that eventually proved largely unjustifiable—Russell’s defense of logical atomism is spurred by an intense interest in justification. Russell’s work throughout his career can be characterized by an extreme reluctance to believe any proposition without a firm, sound reason to do so. Logical atomism was one of the earliest manifestations of analytic philosophy, which (in its most general sense) holds that philosophy should aspire to the precision and exactitude of the sciences. As Russell was one of the founders of analytic philosophy, his work inspired philosophers to rigorously examine their own assumptions and to avoid taking seemingly self-evident truths for granted. It is this dedication to constant, consistent analysis that is Russell’s greatest legacy to philosophy.