Word of the Week: Phenomenalism

George Berkeley - the founding father of phenomenalism

Note

This entry follows on from a previous entry on 'realism', which talked about the attitude that we can clearly and distinctly perceive the world as it really is, rather than as Trump wants it to look like! As you'll see, 'phenomenalism' is virtually the opposite. What do you think: are you a realist or phenomenalist? Please give your view in the comments section, and I'll do my best to respond!

Phenomenalism is undoubtedly one of the most radical responses to the problem of perception. In its rejection of the common-sense principles of naïve realism, phenomenalism is second only to idealism in its scepticism of the power of our senses at gaining access to the external world. This is both an advantage and a disadvantage for phenomenalists, who eliminate the problems of realism but at the cost of advancing counter-intuitive assumptions about the nature of perception. I contend that, despite phenomenalism’s clear advantages in contrast with naïve realism, the theory fails to clearly and simply model our sense experience. If phenomenalism is to survive, the theory should either be modified to taken into account its flaws or rejected in favour of one of the less radical solutions to the problem of perception (e.g., Russell’s representative realism). In reaching this conclusion, I shall address the following issues: firstly, the merits of the phenomenalist solution to the problem of perception; secondly, the difficulties with accepting the ‘traditional’ phenomenalism of Berkeley, Mill and Ayer; thirdly, the equally problematic ‘adverbial’ modification to phenomenalism. Then I shall reiterate my belief that phenomenalism, although giving a compelling critique of naïve realism, fails to provide the simplicity that realism, whether naïve or representative, provides.

Firstly, it is worth considering the problem that phenomenalists are trying to solve. Whilst our common-sense notion of perception suggests that, as John McDowell notes, our senses have ‘Openness to the world’ and the ordinary (physical) objects therein, it is clear that this is not always the case. Whilst we would like for our senses to be both open to the primary qualities of objects and aware of their existence apart from our own minds, we do not always have such direct perceptual contact with objects. In illusion, as Paul Snowden and C.D. Broad note, one is not aware of an ordinary object, or objects made up of physical matter. Since the same account of experience must apply to both veridical (correct) and illusory (incorrect or misleading) experiences, the conclusion that one is never perceptually aware of ordinary objects seems inevitable. Moreover, hallucinatory experience of an ordinary (material) object is not a case of awareness of an ordinary object. For instance, if one is dreaming of a flying multi-coloured teddy bear, this does not mean that we are actually aware of a multi-coloured teddy bear; rather, we are only aware of an idea of a teddy bear. As veridical and hallucinatory experiences are to be given the same account, it seems that one is never perceptually aware of ordinary objects in veridical experience, since hallucinatory perception is subjectively indistinguishable from veridical perception. That is, if we are dreaming, we do not necessarily know that we are dreaming, so there is nothing obviously distinguishable between dream and reality. This is the problem of perception: we do not directly perceive ordinary objects in illusions and hallucinations, so we do not directly perceive ordinary objects in veridical experiences either.

Thus the problem of perception leads to a rejection of the common-sense, or naïve, assertion that we directly perceive the world. Phenomenalism attempts to solve this problem by introducing a new theory. Phenomenalists accept the arguments from illusion and hallucination as premises, and move from these premises to the conclusion that, as we do not directly perceive ordinary objects, we must perceive non-ordinary objects. These non-ordinary objects are ideas in our mind. For phenomenalists, these perceptual ideas can be called ‘sense data’, or the immaterial objects of sensation. In this way, we are indubitably and directly acquainted with sense data even when we are not directly acquainted with material ordinary objects. Instead of having awareness of ordinary objects, we are, as G.E. Moore and Howard Robinson contend, aware of non-ordinary sense data. Phenomenalism solves the problem of perception by rejecting naïve, or common-sense, realism, since it accepts the commonality between dreaming and being awake by admitting sense data as the omnipresent objects of perception. The problem of not being aware of ordinary objects does not apply for phenomenalists, since only non-ordinary objects are perceivable. So the chief merit of phenomenalism is its ingenious solution to the problem of perception.

However, the non-ordinary nature of sense data is simultaneously the greatest strength and most fundamental weakness of phenomenalism. George Berkeley unintentionally elucidated this weakness in phenomenalism. As Peter Millican notes, Berkeley read John Locke as saying that secondary qualities, or the qualities within objects which cause sensations in the subject (perceiver), are the sensations themselves. In other words, Berkeley took the phenomenalist position to its logical conclusion of denying the distinction between perceiver and perceived. He therefore viewed all qualities of object, be they primary (e.g., extension, solidity and other quantifiable properties) or secondary (e.g., sight, smell, or other obviously sensory properties), as sensations. Berkeley therefore maintained that a sentence such as ‘I see a camel’ states ‘I see an idea of a camel’ (my example). Hence Berkeley is regarded both as an ‘idealist’ and a ‘phenomenalist’. Objects do not really exist in themselves, because only ideas (sense data) are real for Berkeley. This led Berkeley to a total denial of the material world in favour of the idealist maxim that, regarding objects, ‘their esse is percipi’, or ‘their essence is to be perceived’ and not to exist. This leads to a complete scepticism of all claims to the existence of matter, which controversial amongst materialists (or physicalists) both in Berkeley’s day and today. Matter, it seems, is common sense for most people, and central to the doctrines of thinkers from Aristotle to Russell. Without matter, it is not clear how objects could really exist as simply sensations, since non-ordinary sense data are fleeting, whereas ordinary objects seem to be at least moderately permanent. This, therefore, represents a key problem for Berkeley’s phenomenalism, regarding how a phenomenalist can account for the endurance of objects when unperceived.

Berkeley himself solved this dilemma by postulating the existence of a deity who can perceive objects even when they are not perceived by anyone else. This explains why objects exist when unperceived by humans, since they still exist in the mind of God. As Ronald Knox acknowledged in his lyrical homage to Berkeley, even when a tree in a Quad is unperceived by any living thing, ‘the tree will continue to be, since [it is] observed by yours faithfully, God’. However, this presupposition of the existence of God fails to apply Ockham’s razor somewhat, since the postulation of a deity generates more epistemic assumptions than it explains. Put simply, God is an unnecessary element of a perceptual theory, since the existence of God, as with all things, should not be taken for granted. Therefore, the problem remains as to how phenomenalism explains the fact that material objects, even when unperceived, appear to endure. As Bertrand Russell reminds us, it would seem odd if, upon a cat’s passing from one part of the room to another, having been hidden from our view for some duration of its journey, someone said that the cat vanished. But this is what Berkeley would have to contend without postulating a deity: that the cat, when unperceived, truly vanished and, seemingly by magic, reappeared at another part of the room subsequently. In fact, it is quite clear that the more parsimonious model would accept that the cat, whether as an idea or ordinary object, actually endured. So if the cat endured when unperceived, and if objects in general exist when unperceived, the question remains: what are these objects, or sense data, which exist when unperceived?

J.S. Mill responded to this question by supposing that objects, rather than being physical objects or fleeting ideas, represent enduring possibilities of perception. Objects are indeed material, but matter, for Mill, ‘may be defined [as] a Permanent Possibility of Sensation’. Mill has solved the problem of enduring substance in the very definition of his theory, since matter is defined as ‘something which exists when we are not thinking of it; which existed before we had ever thought and would exist if we were annihilated’. Arguably, this solves the problem of enduring substance without resorting to realist claims about reality, since all we can perceive is a ‘set of appearances’. All properties of objects are, to use Locke’s terminology, ‘secondary’. There are no primary qualities, since such qualities could exist regardless of their ability to cause sensations in us. Secondary qualities, by contrast, are such that they necessarily cause sensations within us, such that ‘all attributes which we ascribe to objects consist in their having power of exciting one or another variety of sensation in our mind’. Only sense data exist for Mill, therefore, since only sense data can be perceived. But sense data are not physical themselves, and nor can we infer from sense data to unobserved non-mental objects. The only realm is the mental, or phenomenal. Nevertheless, Mill’s case is incomplete. He wants to say that objects are simultaneously temporary sense data and enduring possibilities of sense data. This is a paradox which is somewhat lacking in explanation. Mill does not answer the question regarding why, if sense data only temporarily exist in the mind of the perceiver, sense data have the power to reappear whenever observed. His account is therefore paradoxical, as it does not account for why permanent possibilities of sensation are ‘permanent’ or ‘possible’. A realist could just as easily reply that this permanency is derived from objects existing independently of the mind, lending little support to the phenomenalist case.

A.J. Ayer tried to bypass the problem with a different version of phenomenalism—namely, linguistic phenomenalism. Instead of trying to ask what objects are, Ayer asked what our statements about objects were, and thus found a way of dodging Mill’s paradoxes. He argued that ‘every empirical statement about a physical object’ is ‘reducible to a statement, or set of statements, which refer exclusively to sense-data’. Sense data, for Ayer, are not actual, material things. Sense data are Berkeleian ideas, similar to Mill’s permanent possibilities of sensation, but are linguistic rather than factual. Ayer therefore referred to the endurance of unperceived material objects, such as Russell’s disappearing cat, as reducible to a set of linguistic statements about possible sense data. To say that the cat exists when unperceived is to say that, were one in a suitable position, one would have a sense datum of a cat. However, this is complicated by Ayer’s denial of the linguistic significance of matter, since it is not possible to say that one is standing a suitable angle to observe the cat, as this presupposes the existence of ‘I’ and the spatial dimensions necessary for observation to take place. Instead, for Ayer, one must use conditional statements purely about sense data. Applying this to Russell’s cat, one would have to say that, were one to experience a sense datum of oneself being in a certain position, one would have the sense datum of a cat. This, however, makes the laws of physics seem coincidental, since an apple falling to the Earth is reducible to a statement about a sense datum of an apple, followed by a sense datum of an apple in a lower position, followed by a sense datum of an apple on the ground. Apart from being counter to the realist assumptions of science, this also over-complicates all talk about the world, as a phenomenalist can only talk about possible and actual sense data, rather then enduring physical objects. For each physical object, a phenomenalist refers to an infinite number of possible sense data. By failing Ockham’s parsimony (simplicity) test, Ayer’s phenomenalism demonstrates is impracticability.

Moving on to the third and final part of this discussion, it is worth considering the adverbial modification to phenomenalism and its usefulness. The theory is, amongst other things, a response to general problems with traditional phenomenalism, such as Roderick Chisholm’s ‘problem of the speckled hen’. According to this problem, a phenomenalist claims that all statements, including statements about number, are reducible to statements about sense data. However, the statement that a hen has ‘48 speckles’ is not thus reducible, since one’s estimation may be incorrect. The number ‘48’ is qualitatively indistinguishable from the numbers 47, 49 or 50. This forced Ayer into the position of denying the possibility of counting using sense data alone, since inductive inference (or rational thinking outside of mere acquaintance with sense data) is also needed to enumerate things. This has led adverbial phenomenalists to ask whether sense data exist at all, due to their lack of utility in the basic practice of counting. Adverbialists, such as Chisholm, C.G. Ducasse and Ned Block, contend that secondary qualities are instantiated, not as sense data, but as adverbs. These adverbs, as Gilbert Harman notes, are the ‘mental paint’ of perceptual experience, and allow us to perceive without the need for immaterial sense data or material objects. For an adverbial phenomenalist, to say ‘I see a pink gate’ is to say ‘I am seeing pinkly and gately’, since there is no object, be it a non-ordinary sense datum or ordinary object, to be perceived. There is only our experience and its intrinsic qualities, which are analogous to adverbs. However, adverbialism fails on two grounds. Firstly, it cannot explain why objects exist when unperceived, since seeing a pink gate recurrently is analogous to ‘sensing pinkly and gately’ at several different occasions, with total coincidence and no physical necessity. The fact that Russell’s cat reappears after travelling through a tunnel or emerging from a cave is, for an adverbial phenomenalist, a mere coincidence. Secondly, as Frank Jackson notes, to see a brown square and a green triangle is simply ‘sensing brownly and squarely and greenly and triangularly’. But this removes the association between ‘brown’ and square’, such that, on the adverbialist account, a brown square could be mistaken for a green triangle, and a green triangle could be mistaken for a brown square. So adverbialism cannot cope with multiple sensations at once.

Overall, phenomenalism is praiseworthy for its ingenious solution to the problem of perception, since it restricts our perception to non-ordinary (immaterial) phenomena rather than ordinary (material) objects. However, traditional phenomenalism fails at explaining why objects persist when unperceived: Berkeley posits a deity without first providing a coherent argument for such a being; Mill begs the question by failing to explain why sense data are ‘permanent possibilities of sensation’; and Ayer’s linguistic phenomenalism restricts all talk of physics to talk of possible sense data, making Newton’s celebrated laws seem like coincidences rather than the products of necessity (or, as we should say in today’s ‘Quantum age’, at least high probability). The attempt to rescue phenomenalism through adverbialism is similarly brittle when scrutinised properly. All in all, phenomenalism is insufficient at explaining the patterns of our perceptual experience. Perhaps realism should come to the rescue.