Is Indirect Realism a Convincing Theory of Perception?
I will argue that IDR can respond to Berkeley's criticism but not to the scepticism objection, and is therefore not a convincing theory of perception
- Indirect realism (IDR) is the view that the immediate objects of perception are mind-dependent sense data, which are caused by and represent mind-independent objects - 'Realism' refers to the metaphysical claim that mind-independent objects exist; 'indirect' means mediated — there are three components to perception: the perceiver, a mind-dependent representation, and the mind-independent object itself - The epistemological consequence is that we gain knowledge of mind-independent objects indirectly, through their representation by sense data - A strength of the theory is its alignment with scientific findings: neuroscience suggests perception operates via neural intermediaries — sense data understood as brain states caused by external reality. This gives IDR empirical backing that purely philosophical theories of direct perception lack - IDR is also supported by Locke's argument that some of our ideas must represent and be caused by primary qualities of mind-independent objects — when a grain of wheat is divided it loses its secondary qualities like colour but retains primary qualities like extension, suggesting our ideas of primary qualities do represent something mind-independent - I will consider two criticisms: Berkeley's argument that sense data cannot represent mind-independent objects, and the more serious problem that IDR leads to scepticism about whether mind-independent objects exist at all - I will argue that IDR can respond to Berkeley's criticism but not to the scepticism objection, and is therefore not a convincing theory of perception
Section 1: Berkeley's Argument that Sense Data Cannot Represent Mind-Independent Objects
- The counter is effective against Berkeley's argument as stated — the likeness principle is not a necessary condition for representation, and the linguistic analogy demonstrates this clearly - However, we can construct a stronger version of the objection that does not rely on the likeness principle. Even granting that representation does not require resemblance, we still require some justification for believing our sense data represents anything mind-independent at all. We only know the word 'chair' represents a chair because we have experienced chairs independently of the word. If we have never experienced mind-independent objects directly — as IDR insists — we have no basis for knowing our sense data represents them in any respect. For all we know, sense data might bear no representational relationship to mind-independent objects whatsoever - This stronger version of the argument does not depend on likeness and cannot be dismissed by the linguistic analogy — it leads directly into the scepticism objection
: The argument
- IDR claims that sense data represents mind-independent objects - Berkeley's likeness principle states that for one thing to represent another, it must resemble it — and to justifiably know two things resemble each other, we must be able to compare them directly - But IDR entails a veil of perception: we never directly experience mind-independent objects, only our own sense data - We therefore cannot compare our sense data to mind-independent objects to verify any resemblance - So we can never justifiably claim our sense data represents mind-independent objects - IDR is self-undermining: its claim that we only experience sense data undermines its further claim that sense data represents mind-independent objects
: Response
- The likeness principle is questionable: representation does not require resemblance - Linguistic symbols represent objects without resembling them — the word 'chair' bears no resemblance to a chair yet successfully represents one - These representations are entirely arbitrary in their form yet function perfectly - So mind-dependent sense data could represent mind-independent objects without resembling them, and Berkeley's argument fails
Section 2: Indirect Realism Leads to Scepticism about the External World
- Both of Locke's arguments are weak because they assume only two possible sources of our perceptions: the external world or our own imagination. But there are other possibilities — an evil demon could produce coherent, involuntary perceptions across all senses; God could be the origin of our ideas as Berkeley suggests; or an unconscious part of our own mind beyond our voluntary control could produce orderly multi-sensory experience. Locke has not ruled these alternatives out, so neither argument establishes the external world as the cause of our sense data **Response 3 — Russell: the external world as the best hypothesis** - Russell accepts that we cannot prove the existence of a mind-independent world with certainty, but argues this sets the bar too high - Since no deductive proof is available, the appropriate approach is abductive — we should ask which hypothesis best explains our experience - The sceptical hypothesis — that there is no external world — leaves the regularity and order of our perceptions entirely unexplained. Russell's example: his cat appears on one side of the room, then elsewhere — the realist hypothesis explains this naturally (the cat moved); the sceptical hypothesis offers no explanation at all - The realist hypothesis therefore has greater explanatory power and we are justified in preferring it - Russell's approach is stronger than Locke's because it does not try to prove the external world with certainty — it only claims it is the most reasonable hypothesis, which is proportionate to the evidence available - However Russell's argument is only partially successful. The alternatives Locke failed to rule out apply equally here: Berkeley's idealism — God as the cause of our perceptions — would explain the order and regularity of experience just as well as the realist hypothesis. An evil demon producing systematic, orderly experience would also explain it equally well. Russell cannot claim the realist hypothesis is the best when there are equally powerful rival explanations - More fundamentally, even if we grant that the external world is the best hypothesis, it remains only a hypothesis — it could still be false. Russell's defence means our belief in the external world is only as secure as an abductive inference, which is never certain. The sceptic's claim is not that there is no external world, but that there might not be — and Russell's best hypothesis response does not eliminate that possibility - Direct realism and idealism both claim to avoid this problem more decisively by denying any mediating layer between perceiver and reality altogether — though both face their own serious objections
: The argument
- If we only ever directly perceive sense data and never mind-independent objects, we have no direct evidence that mind-independent objects exist at all - We cannot know that our sense data is caused by mind-independent objects — for all we know it could be caused by our imagination, an evil demon, or another mind entirely - The veil of perception is not merely an epistemic inconvenience; it means IDR cannot even justify the 'realism' it is named for - <mark style="background: #ADCCFFA6;">This is the more crucial objection because it attacks the foundational realist commitment of IDR directly. The previous objection queried the representational relationship between sense data and objects; this one questions whether there is anything on the other side of that relationship at all</mark> **Response 1 — Locke: coherence of the senses** - Locke argues that the involuntary coherence of our senses provides evidence for a mind-independent cause - When we perceive an apple we see, feel, smell and taste it — multiple independent senses reporting the same object simultaneously. The probability that each sense would independently produce the same mistaken report seems very low - This coherence is best explained by there being a real mind-independent object causing all of them **Response 2 — Locke: the involuntary nature of experience** - If sense data were caused by imagination we would have control over it, since we control our imagination - But we do not control our perceptions — we cannot choose not to feel pain or will away what we see in front of us - Therefore sense data is not caused by imagination, and an external world is the best explanation
- Berkeley's likeness principle fails: representation does not require resemblance. The stronger version of the argument — that we cannot justify any representational claim without having directly experienced mind-independent objects — is more troubling, but still only attacks the indirect component of the theory - The scepticism objection is more serious and ultimately succeeds. Locke's responses fail because they do not rule out alternative explanations. Russell's best hypothesis response is stronger and represents the most defensible position available to the indirect realist, but it is only partially successful — it cannot establish the external world over equally powerful rival hypotheses, and leaves scepticism as a live possibility - Theories of perception are meant to show how knowledge can be gained through perception. IDR made two claims it cannot fully defend: that sense data is caused by mind-independent objects, and that it represents them. Both ultimately lead to scepticism - Indirect realism is therefore not a convincing theory of perception