Laurian Kertesz ú.æ. 322039595 ùðä á;, úåàø ùðé
The purpose of the present essay is to defend the position of Hylas against Philonous in what I consider to be a key passage in the "Three “Dialogues” The "Three Dialogues" have the aspect of a game between Hylas and Philonous. Berkeley's (demonstrative) intention is to portray Hylas as the obvious loser of the game. My conviction is that Hylas, if the loses the game, does so because of Philonous’ psychological pressure and rhetorical tactics. More specifically, Hylas starts by raising valid points, espouses acceptable intuitions that he does not pursue them to the end. The reason of Hylas' defeat is not his lack of justice, but rather his lack of courage in the face of what I shall prove to be Philonous' sophistical attacks.
My essay is not a historical one. I am not primarily concerned with matters of textual accuracy, but rather with the coherence of the arguments I discuss. That's why the name "Berkeley", as I shall use it, does not (necessarily) refer to the actual philosopher Berkeley, but rather to a hermeneutical construct. I shall concentrate on an argument that I believe that the actual philosopher Berkeley really held and that he presents in the "Three Dialogues". It is not my purpose to discuss whether, and how,
As I understand Berkeley’s immaterialism, its central thesis is the rejection of the existence of a substratum, of a substance of sensible properties. The “substratum” may be understood in two different ways: (1) a propertiless and unsensible support of (sensible) properties; (2) an unsensible property that underlies the sensible properties. As I shall argue in the second section, Berkeley’s rejection of substratum (2) proceeds via the sophism of equating mental acts and intentional objects.
I shall concentrate on Berkeley’s attack against the existence of a substratum as understood in the first of the above ways.
The first section of my essay will point to an assumption that I believe can be detected in Berkeley's philosophy, namely the assumption of the uniformity of perceptions. This assumption says that mental terms denote homogenously psychological entities belonging to the same natural kind, or the same type of psychological entities. This assumption is necessary if Berkeley is to avoid present-moment solipsism.
The second section will concentrate on Berkeley’s main argument against the existence of the matter (that is, of a substratum of properties). This argument consists in equating the mental acts that are directed to an object with the object itself. This argument, to be valid, involves two additional assumptions, namely what I shall call the IMPORTATION SCHEMA (according to which the inference from “I conceive a seen object” entails “I conceive that I see an object”) and the assumption of EMPATHY, which says that access to a mental state is only possible by experiencing that mental state itself. The conclusion that I will draw is that Berkeley’s argument would be valid only if instead of the self understood as the mind of an empirical person he would use the transcendental subject. That is, it is true that esse est percipi by the transcendental subject, although it is not true that esse est percipi by any empirical mind.
The third section of my essay will have a structure similar with the second one. In it I will concentrate on what I shall call Berkeley’s modal fallacy, namely the inference from I perceive an object as having sensible quality S to I perceive an object that necessarily has sensible quality S. Like in the second section, I will conclude that this inference would be valid only from the point of view of the transcendental ego.
1. Then Assumption of the Uniformity of Perceptions
This (pre-Wittgensteinian) assumption says that a mental term M denotes tokens of the same type, or types of the same natural kind no matter the individual to whom it applies, and that the term M could be used by an individual in isolation with equal right (and identical results) as an individual belonging to a common practice of using psychological predicates. I would dare to say (although I am not sure) that this assumption (of the uniformity of mental states denoted by a certain term) characterizes all writers of the British empiricism, and is responsible for its psychologistic trend.
The role that this assumption plays in Berkeley's philosophy is to ensure the possibility that objects do exist even in the absence of my perceiving them. Berkeley needs the assumption of the uniformity of human (and divine) perceptions in order to make plausible the rejection of a substratum. In the absence of the assumption of this uniformity, the only source of objectivity would be the (Lockean) matter, the existence of a support of sensible properties that is not itself a property; that does not vary with the variation of properties. For Berkeley, the assumption of the uniformity of human perceptions means, most importantly, that my perceiving object x and other person's perceiving object x are one and the same mental state. I think that this contention is vital for Berkeley, since – had he abandoned it, or doubted it – then he would no longer have been able to hold that the objects are the human perceptions of objects; he would have had either to embrace an empiricist version of solipsism ("the objects are the way I, my empirical mind, perceives them") or the thesis that objects are what causes their being perceived in several minds, or what is common in the ways they are perceived by several minds. But he explicitly rejects the former possibility (empirical solipsism), since he accepts the existence of other minds; and he rejects the latter possibility since it would commit it to ascribing an ontological status to the material substratum of sensible qualities, as something that is in objects beyond their being perceived by any human mind. And the rejection of a substratum of sensible qualities is the crux of Berkeley's philosophy.
If Berkeley abandoned the assumption of the uniformity of human perceptions and remained consistent in his rejection of a substratum, he would fall in present-moment solipsism (and atheism) and would have absolutely no possibility to distinguish between objectivity and illusion. To make my point more explicitly: if Berkeley had any doubt that my perception of, say, a horse and other person's perception of a horse are identical (or at least commensurable) then the following argument could be produced: "The perceptions of the other person are unknown to me I may, after all, ascribe no determinations to them. Then what reason do I have to believe other persons have perceptions at all?”. Berkeley presented the analogous of this argument in his rejection of the matter (see especially the Second Dialogue between Hylas and Philonous, where Hylas is put in the impossibility of ascribing any objective determination to matter, being left with an “unknown something”.".
Curiously or not, this problem, of the "other minds", that became famous in 20th century philosophy did not trouble the classical empiricists. When Berkeley addresses it, he gives it a rather simplistic solution, pointing to the argument by analogy. The argument by analogy is however reputedly confused. The main accusation against it is that it is not clear how far it can be pushed. It is not clear, in other words why I cannot ascribe mental states (and perceptions) to a table or to a stone .
What is more worthy to note is that Berkeley was indeed aware of a form of criticizing the argument from analogy: more specifically, he criticizes it to the extent at which such an argument could be invoked to ground the existence of an unsensible substratum by analogy with the sensible appearance of objects. His argument, indeed, is: “It cannot be said that matter is resemblant to the sensible qualities (ideas) since nothing but an idea (a sensible quality) may resemble another idea (another sensible quality). The general form of this argument is: “starting from an entity x of a certain kind, I cannot conclude the existence of another entity that resembles x since an x can only resemble an x”. This inferential schema is, in my opinion, not valid and/or arbitrary: indeed, if it is valid it can be used against any argument by analogy: for example, it can be used to prove that no one has mental states at all: “I am acquainted only with my mental states. But nothing can resemble my mental states except other mental states of mine. Therefore it cannot be said that other people have mental states identical with my mental states, even less so that htye have mental states (type) identical with my mental states.” The fact that Berkeley does not use this argument proves, again, his unfairness with respect of the cause of the matter. He is unfair because he uses a premise (that of the uniformity of human perceptions) that could have been rejected with arguments of the form he uses against matter.
Another assumption that is shared by Berkeley and, most probably, by all the philosophers of the British empiricism is that of the transparency of the mental. By this I mean the conjunction of the theses of phenomenality ("If I have mental state M, then it seems to me that I have mental state M") and of first-person authority ("If it seems to me that I have mental state M, then I have mental state M"). The intricate logical relation between phenomenality and first-person authority cannot be studied in this essay, so I will use interchangeably "phenomenality', "transparency of the mental" and "first-person authority".
.The Problem of the I and the Problem of the Substratum: The Sophism of2 Equating the Act and its Object:
One of the central arguments of Berkeley against the existence of matter, i.e. of a propertiless substratum of sensible qualities, is that, since we have, and may have, access to sensible objects via our senses (more generally: we may have access to objects only via our mental acts) then it is impossible to separate between the acts and the objects. Let's take a closer look at this argument, as it occurs in the First Dialogue between Hylas and Philonous:
"Philonous: I am content to put the whole upon this issue. If you can conceive it possible for any mixture or combination of qualities, or any sensible object whatever to exist without the mind, then I will grant it actually to be so
Hylas: If it comes to that, the point will soon be decided. What more easy than to conceive a tree or a house existing by itself, independent of, and unperceived by, any mind whatsoever? I do at this present time conceive them existing after that manner
Philonous: How say you, Hylas, can you see a thing which is at the same time unseen ?
Hylas: No, that were a contradiction
Philonous: Is it not as great a contradiction to talk of conceiving a thing which is unconceived?
Hylas: It is
Philonous: The tree or house therefore which you think of, is conceived by you.
Hylas: How should it be otherwise?
Philonous: And what is conceived, is surely in the mind.
Hylas: Without question, that which is conceived, is surely in the mind.
Philonous: how then came you to say, you conceived a house or tree existing independent and out of all minds whatsoever?
I quoted the above paragraph for two reasons. First, it makes mention of “conceiving” as a mental faculty parallel to that of “Seeing” and is thus more representative for idealism in a proper sense (purged of its empiricist connotations). Secondly, it is a particularly strong idealistic thesis. If we pay attention to the premise “What is conceived, is surely in the mind”, we cannot help reading it in the following key: conceiving is a mental operation; via the premise of the transparency of the human mind, it follows that we have access to the way it (the human mind) works. Moreover, we may have access only to our mental operations – it is, after all, we, our mind who is to have access to anything if access is possible; and if we could have access to the objects that our mental operations are directed to in the absence of our mental operations, then why would we need mental operations at all? There are no objects different from the mental operations that are directed at them: objects are only the acts, the operations. This is the true meaning of esse est percipi: it is an intensional meaning, stating the equivalence between the concepts of being and being perceived: esse is necessarily percipi. This is an idealistic thesis that says more than “everything that is, is percived”, or even than “everything that is, is necessarily perceived”. This latter (weak) idealistic thesis recognizes the existence of objects that are different from the ways of having access to them, although it does not recognize the existence of objects that are altogether inaccessible to us. The strong idealistic thesis, that I ascribe to Berkeley, says that objects are not only necessarily accessible to us: they are nothing else but our access to them. (There are no lighted bodies, but only light) .
The argument can be generalized to versions of linguistic idealism: Is it not a contradiction to speak of something that is ineffable, i.e. unspeakable of? Of course there is: If I speak about something, obviously it is speakable of, therefore it is not ineffable. But how can I depict in language anything besides the way I speak about it? Obviously I cannot. Therefore it is not only the case that everything that is, is speakable of; it is moreover the case that everything that is, is nothing over and above my way (our way) of speaking about it). Not only that everything that is has its reflexion in language; there is only language. (I wonder whether Wittgenstein of the first or later period would have accepted such a strong idealistic thesis; but this is, of course, a separate question).
Berkeley obviously considered the argument of Philonous in the above quotation to be a valid one. My purpose is to prove that it is not.
In order to see this, let's consider the following antinomy: On the one hand, it is easy to imagine a visual object that is seen by noone, a fortiori a visual object that is not seen by me. On the other hand, such an achievement may seem impossible for the trivial reason that if I imagine any visual object at all I need to imagine myself as seeing that object. Therefore – someone may conclude – it is only apparently easy, it is in fact impossible to imagine a visual object that is seen by noone.
Whether this last argument is a good or a bad one it remains to be debated in the following pages. I believe that Berkeley really believed it is, and that he believed in the importation of the "I", such that the following scheme of reasoning is valid:
(1). I conceive a visual object
(2). Therefore I conceive a seen visual object
(3). Therefore I conceive I see a visual object
I shall criticize these steps. First, the transition from 1 to 2 depends on the lack of a propertiless substratum of objects. If there were a substratum of objects, then it would obviously be possible to imagine a visual object that is not seen: such an object would be imagined as heard, or touched. What would guarantee that it is the same object that I imagine would be that there is a something (the object) beyond the sensible properties of it. (“The substratum” in this context could be conceived also as a principle of unification of the properties of the objects: this principle of unification would be different from the properties that it unifies, that it “holds together”). To the extent that Berkeley admits the transition from (1). to (2)., his reasoning is circular: his purpose is to prove that there is nothing in objects beyond their sensible properties, but this is precisely what he assumes.
Yet, the transition from “I imagine a visual object" to "I imagine a seen visual object" is the crucial step in Berkeley's argument. It could be reformulated as
“If I conceive an object that may be accessible to a certain sense, I imagine an object that is accessible to that sense”. Insofar as this is a fallacious inferential schema (and I am sure that it is), it involves a modal confusion. I will discuss it in the third section of my essay. For the moment, let's notice that the transition from (1) to (2) in the above argument is invalid – since it is possible that an object that may be accessible to a certain sense S be accessible to other sense S' without being accessible to S; and that if it were valid, then its premise should be taken in conjunction with the assumption that an object accessible to sense S is nothing over and above its being accessible to sense S; or it is essentially accessible to sense S (if it becomes inaccessible to S, then it becomes another object). If either of these assumptions were true, then we could do away with a substratum of the properties of an object. But the lack of a substratum of the properties of an object is the core of Berkeley's immaterialism and is precisely what he wants to prove.
We could imagine that Berkeley (more properly speaking: Philonous) would be ready to accept the above criticism and weaken his position in the following way: "Perhaps it is not true that whenever I conceive an object that may be accessible to sense S, I conceive that that object is actually presented to sense S; but at least it is true that whenever I conceive that an object is actually presented to sense S, I conceive that it is me who has access to it via my sense S. So, if I conceive that any object is actually seen, I conceive that it is me who sees it".
This inference is a purely solipsistic one (and to the extent at which Berkeley made it, this proves that he was, despite his best intentions, committed to solipsism). It depends on the following inferential schema, which reduplicates the pronoun I from a direct context into an oblique one:
IMPORTATION SCHEMA: “If I conceive that an object is perceived by anyone, I conceive that I perceive it”.
At face value, this principle seems obviously false: it seems very plausible that I can imagine that someone (say, John) sees a tree in a place very distant from the place where I am, so that I conceive a visual object and simultaneously I conceive that I don’t see it. In this case, it is true that I conceive the tree, the tree’s being seen and myself not seeing the tree. This simple mental experiment seems to me to support the conclusion that I may conceive that x is perceived without conceiving myself to perceive it in such an intuitive way that it is impossible for any philosopher to convincingly refute it. In other words, the importation of the “I” from “I conceive that sensible object x is F-ed”, where “to F” is a psychological verb (denoting for example a sense) to “I conceive that I F x” is false. The scheme of importation of the "I” seems so obviously false that it becomes interesting to discern the reasons why Berkeley could have accepted it, and perhaps the qualifications together with which some reformulation of it could be acceptable.
First, let us not forget that Berkeley makes use of the assumption of the uniformity of human (and divine) perceptions, that I presented in the first section of this paper. This assumption may favor somewhat of the following inferential schema: (call it EMPATHY, for lack of a better term):
EMPATHY: If I am certain that someone else has mental state M, then I have myself mental state M.
(It is a historical matter whether Berkeley really believed in EMPATHY. As a matter of historical interest, Adam Smith fully endorsed EMPATHY in his Theory of Moral Feelings). EMPATHY is itself independent of the assumption of the uniformity of human perceptions. Yet there is a logical link between them, that may be realized via the assumption of phenomenality (the transparency of the mental). This becomes clear once we realize that EMPATHY is a generalization of PHENOMENALITY to a situation where the personal identity of the subject who is experiencing a mental state does no longer matter. Indeed, EMPATHY is a subjectless PHENOMENALITY.
The thesis of phenomenality (or of the transparency of the human mind) says that whenever I have a certain mental state, nothing remains hidden to me, so to speak, in that mental state. That mental state is known to me under all its aspects. From here it is only a small step to concluding that whenever I refer (in language) or conceive mental state M, I experience it, it is fully present to me.(As it were, when I refer to/conceive a mental state, I at least have access to an aspect of it. But whoever has access to an aspect of a mental state M has access to all its aspects). Thus, EMPATHY is nothing but a generalization of the transparency of the mental to a situation when the identity of the subject who experiences that mental state does no longer matter; EMPATHY is transparency adapted to a subjectless context.
I am not myself committed to EMPATHY, but I think that the arguments that favour it, in conjunction with the assumption of the uniformity of human perceptions (all mental states denoted by an identical term are themselves identical), lead to the conclusion that whoever refers to, or conceives, or – more generally – has access to a mental state of anyone else, has himself that mental state. Thus, phenomenality and the uniformity of the human perceptions may be some premises of an argument that concludes EMPATHY. And if EMPATHY is true, then it is true that it is impossible of conceiving myself to have mental state M without actually having it. (What EMPATHY says is that if I conceive of anyone having mental state M, then I have mental state M; a fortiori if I conceive of myself having M, then I have M). Therefore, EMPATHY supports the IMPORTATION SCHEMA: indeed, EMPATHY can be expressed as saying that If I conceive of any mental state experienced by anyone, then I experience that mental state – from which the IMPORTATION SCHEMA directly results – since perceiving an object is an instance of experiencing a mental state.
But we saw in the above lines that IMPORTATION SCHEMA is false, at least according to the most basic intuitions of common-sense. Since EMPAHY correctly supports the IMPORTATION SCHEMA, the conclusion that results from here is that EMPATHY is itself false. On the other hand, we saw that EMPATHY is somewhat better supported by intuition than the IMPORTATION SCHEMA. It remains to investigate whether EMPATHY could be qualified so as to yield a version of the IMPORTATION SCHEMA that will be itself more qualified and more acceptable.
The qualification that I have in mind consists in drawing a distinction that Berkeley was insensitive to, namely that between the transcendental subject and the empirical subject. (Again, as a matter of historical interest: this distinction became a common place in philosophy after Kant and the German idealism; it is explicitly drawn by Husserl and by the Wittgenstein of the Tractatus. This distinction was entirely absent from the writings of the British empiricists, leading to the collapse of the concept of self into a “bundle of perceptions" in Hume). By "transcendental ego” I mean something that is such that is inescapable from the description of the world, the condition of possibility of the objective reality. On the other hand, the empirical ego is merely the mind of an empirical person, a “bundle of perceptions” that has a determinate and limited existence in time and has access, by its faculties of perceiving, to a limited portion of the reality. There are many empirical egos, as many as there are creatures endowed with minds; but there is only one transcendental ego. The distinction between the transcendental and the empirical ego was formulated in its clearest terms in Wittgenstein’s Tractatus, where the transcendental ego is described as the limit of the world , an extensionless point that presupposes the very possibility of describing the world but that does not belong ton the world and cannot be described in meaningful language. Of course, Wittgenstein’s solipsism (in the Tractatus) may not be confounded with the trivial present-moment solipsism (“only my present mental states do exist”) and the reason why it may not be counfounded with it is that it involves the transcendental ego, and not the empirical one. The transcendental ego may be conceived as the guarantee of the possibility of any empirical mind's in general to have access to reality – be it sensorial, intellectual or linguistic.
What does this have to do with the previous discussion about the IMPORTATION SCHEMA and EMPATHY? My suggestion is that EMPATHY, while being a problematic assumption about empirical egos, is evident and rather redundant in the case of the transcendental ego. Indeed, the transcendental ego is such that it may not be absent from the mention of any object that belongs to the world. To use somehow bombastic terms: Whenever something is present in general, it is present to the transcendental ego. Whenever something is seen, or conceived, or described in language, it is I who see/conceive/describe it in language – not the empirical I (my actual, limited, psychological faculties) but rather the transcendental I. The reason why EMPATHY is evident and redundant in the case of the transcendental ego is that there is only one transcendental ego at all. It makes no sense to speak of anyone’s conceiving or describing anything without the transcendental ego conceiving or describing it.
Now, the next step is to consider the relevance of the transcendental ego for the IMPORTATION SCHEMA. That schema said that:
If I conceive and object being perceived by anyone, then I conceive that I perceive it.
The problem is how this schema should be understood in order to make it true. The plausible suggestion is that the last occurrence of the “I” should be understood as standing for the transcendental ego, while the other ones may be understood as referring to the empirical (any empirical) mind. Since the transcendental ego is the condition of the possibility of sensorial, intellectual or linguistic access to any object whatsoever, it is implicit in any such access, had by any empirical ego.
The distinction between the transcendental and the empirical ego may shed light on the distinction (or lack thereof) between mental acts and the objects to which they are directed. We saw that Berkeley obliterates this distinction by holding a strong version of idealism, according to which it is not only the case that all objects are accessible to the subject, but it is moreover the case that objects are nothing over and above their being accessed by the subject. What I hold is that, from the point of view of the transcendental ego, it is indeed correct to equate objects with the subject's having access to objects: the objects are the way they are presented to the subject. I can use a spatial metaphor to support this conclusion: in order to grasp the difference between an object and a way of access to that object, one needs to stand outside the object and the way of access to it. Thus: this distinction is important if and to the extent at which the object can exist independently of my access to it. If "my access to the object" is understood in the sense of an empirical mind's access to it, then it is indeed true that my way of access to the object should be distinguished from the object itself – for, indeed, it may be the case that the object exists unperceived, unaccessed by me. But if "my access to the object" is understood in the sense of the transcendental subject's access to the object, then such an outer perspective makes no sense: a perspective that is outer to an empirical mind is possible, but a perspective that is outer to the transcendental subject is impossible.
To put this in other words: an object accessible to a mental faculty should be distinguished from the mental faculty itself insofar as the object could be accessible to other mental faculties and to other subjects. But – once we accepted, as I did – that nothing may remain inaccessible to the transcendental subject, it follows that the transcendental subject's ways of access to objects are objects themselves.
. Thus, the conclusion of the present section is that Berkeley's reasoning is a sophism but it would be valid if instead of the empirical mind we would understand the transcendental ego. The transcendental ego is "present" whenever something is imagined). If so, then Berkeley's mistake lies in that he does not distinguish between the transcendental and the empirical ego.
If this distinction is not drawn, then Berkeley's idealism amounts to the solipsism of the present moment: indeed, Philonous’ argument (as presented in the quotation I gave) could be generalized to prove the conclusion that whenever something is perceived (or conceived), it is perceived (conceived) by my own’s mind now. Why not the temporal importation from the “I am conceiving right now that I saw the tree” to “I am conceiving of myself seeing the tree right now”? On Berkeley’s premises, this inference would have all reasons to be considered valid; after all the uniformity of human perceptions and EMPATHY should work equally well between a moment of time and other moments of time: that is, my mental state M now and my mental state M in the past are type-identical; and when I conceive that I had mental state M in the past I am really having it now). Given the distinction that I drew between the transcendental and the empirical ego, it is of course true that the transcendental ego has an omnitemporal point of view on the world: it has access, so to speak, to all regions of time and space. But it is of course false that my (or anyone’s) empirical ego has such an access .
Berkeley’s assumptions of the uniformity of perceptions and of EMPATHY is an awkward and incomplete step, towards the elucidation of the concept of the transcendental ego. It is a step in this direction since the assumption of the uniformity of human perceptions suggests the existence of a transcendental ego whose faculties consist in what is common to all faculties of all empirical egos. The transcendental ego could indeed be understood as that which performs the perceiving, describing in language in the way that is common to all acts of perceiving and describing in language by any human being. On the other hand, Berkeley's assumption of the uniformity of human perceptions and the assumption of EMPATHY is an awkward step in this direction, since it is plainly (intuitively) false that whenever any person S sees an object and he is known, believed, conceived by someone else etc as seeing that object, then the second person (the knower/believer/conceiver that S sees the object) has the same mental state, the same perception as S .
The sophism of confounding the act and the intentional object appears appears, in its turn, in a clearer light once we keep in mind the distinction between the transcendental and the empirical ego. This is a sophism since the importation of the I is not valid if the I is understood in an empirical sense, as a particular human mind. It is plainly false that a seen object is merely my seeing it – since it is possible that I not be seeing it and it still be a seen object – that is, seen by other people. (The difference between the intentional act and the intentional object is necessary to the extent at which the intentional act is dispensable). On the other hand, any intentional object really is the transcendental ego's way of access to it – for the simple reason that the access by the transcendental subject is not dispensable. To use a (Tractarian) Wittgensteinian spatial metaphor: in order to assess the difference between the intentional act and the intentional object one needs a standpoint that is outer to the intentional act; one needs to see or "grasp" the object in such a way that it becomes apparent that it is not exhausted by the intentional act. Such a standpoint is available to the transcendental ego as it is available to any empirical mind that is different from mine. (The standpoint of the transcendental ego is however more general since it includes the access not only to the way the object appears to other minds than mind, but also to the way the object could appear to any mind whatsoever).
3. Berkeley’s modal fallacy (the assumption of modal uniformity)
There is another simple reason why I believe that Berkeley's idealism is false (and that I believe anyone would deem it as false). The reason is that I believe it is possible that my perceptions of the sensible objects be fallacious yet the objects remain the same objects. Berkeley’s version of idealism makes no room for the intuition that our senses may deceive us yet we succeed in referring to real objects. That is, it makes no sense for Berkeley to accept that it is possible that (a) I perceive an object as cold; (b) that object is not actually cold; (c) yet I perceive the very same object that I perceived as cold. If he could accept such a possibility, then he would be forced to believe in “matter”, in a substratum of the properties (of the sensible qualities) of the object that is altogether different from those properties. For Berkeley, that is, there is no distinction between perceiving a cold object and perceiving an object as cold. The reason is that any cold object is essentially, necessarily cold, whereas an object that may be perceived as cold without being cold is an object that could be perceived as not-cold by other perceivers, therefore it is an object that is not essentially cold.
Insofar as this distinction seems (for good reasons) to have a solid intuitive basis and is (I take) correct, Berkeley is guilty for not making the distinction between how an object appears to me and how it could have appeared (to me or to anyone else). For the distinction between perceiving a cold object and perceiving an object as cold is obviously a modal distinction related to the manner of that object’s appearing: the situation when I perceive an object as cold but I don’t perceive a cold object is precisely the situation in which the object appears to me as cold, but does not necessarily appear so to other perceivers. Berkeley’s not making this distinction is ascribable to a modal fallacy that I ascribe to him: that fallacy consists in equating how an object appears to me and how that object must appear (to me or to anyone else to whom it appears).
This is somehow related to the assumption of the uniformity of human perceptions that I presented in the first section of my essay. As Berkeley assumes the uniformity of (actual) human perceptions of the same objects, so he assumes the uniformity of all perceptions (possible or actual that a sensible objects might elicit. There is a nice symmetry between saying that an object appears to me in the same way as it appears to God and to any other human being and saying an object appears to me in the same way as it COULD appear to me/to anyone I think that Berkeley accepts both assumptions. In both cases, the assumption of the uniformity of human perceptions is what allows him to do away with a substratum of an object's qualities, as I explained in the first section of this essay: if an object is and can be nothing but the way it appears to me, and to other people, then it is pointless to look for what supports the different ways it appears – the re are no such different ways. Likewise: if an object is essentially, i.e. it must be the way it appears, then it is pointless to look for something that supports the way it appear and the way it could appear – there are no such different ways. Like in the former case, I believe that Berkeley did have some reasons to accept this conclusion, but his reasons are not correct.
Let’s notice the following seeming paradox: (it is actually a sophism, but it has the appearance of a genuine paradox): It is nonsensical and self-refuting to say: "I am not speaking right now". But it is not nonsensical (it is even true) to say: I might not be speaking right now". This has however the appearance of a paradox, since sentence "I might not be speaking right now" would be true precisely in the circumstance in which I might not be speaking right now. That is, "I might not be speaking right now is true precisely when "I am not speaking right now" might be true. This last sentence however is never (and can never) be true. So how could "I might not be speaking right now" be true? A modal sentence of the form It is possible that not-p (M-p) is true just if and only if not-p may be true. But with our example we seem to be in a situation where M-p is true, yet –p may never be true.
This is not a genuine paradox (as shall try to prove in the following lines) yet it invites itself to a sophistical solution. It receives the appearance of a paradox when one is obliged to make the following option: either accept that M-p is itself necessarily false, that is "I might not be speaking right now" is necessarily false; or accept that –p may sometimes be true, i.e. that it sometimes may be true to say "I am not speaking right now". The sophistical solution that invites itself will be the following: since it is, whatever it may take, impossible to admit that "I am not speaking right now" may sometimes be true, then let's abandon the intuition that "I might not be speaking right now" is true. Therefore, "I might not be speaking right now" is necessarily false. Therefore I am always and necessarily speaking: I must always speak, since - isn’t it? – "I might not be speaking right now" is self-contradictory, exactly like "I am not speaking right now".
That something is wrong with this argument is, I hope, evident. The simplest reason why it is a sophism is that it fails to distinguish between the categories of the necessary and of the a priori. The truth is that I am not speaking right now is not necessarily false, although it is a priori false. Therefore, I might not be speaking right now is true, because – like any sentence of the form M-p – its –p may be true: that is, because I am not speaking right now may be true. What the above argument (the appearance of a paradox) proves is that I am not speaking right now could not be uttered when it is true: that is, if it were true it would be inexpressible. (If human beings had another means of communicating whose expressive power would be similar to that of verbal language, then there would be no problem with uttering I am not speaking right now using that means of communicating; actually, there is nothing paradoxical about writing the sentence I am not speaking right now. There might be something paradoxical about writing or uttering I am not writing or speaking right now, but this only suggests that there might be another means of communicating, different from speaking and writing, in which the above sentence could be expressed and be true). The reason why I mentioned this sophism is that it resembles Berkeley’s reasoning about esse est percipi.
Let’s examine other illustration, closer to Berkeley: I am not referring to John right now is a priori false, at least for the obvious reason that the very act of verifying it assumes identifying John as the reference of my word John. However, I might not be referring to John right now is true. It is true because "I am not referring to John right now" is not necessarily false, although it is a priori false. This again has the appearance of a paradox: for indeed "I might not be referring to John right now" would be true when, and only when, I am not referring to John might be true. But I am not referring to John might never be true since it may never be uttered and be true. Therefore, I might not be referring to John may never be true: it is necessarily false. Therefore, I am necessarily referring to John: I MUST refer to John, and I always refer to John. (Isn’t it true, if there were a moment of time t when I did not refer to John, then the sentence I am not referring to John is true if I pronounce it at t; but it is obvious that there is no moment of time such that the sentence I am not referring to John is true at it). Again, this reasoning is fallacious because it does not pay due attention to the distinction between the necessary and the a priori: I am not referring to John is a priori false, but it is by no means necessarily false.
Finally, a properly Berkeleyan example: I regard a window, point ostensively to it and say "I might not perceive this window". Now, what I say is intuitively true. (I think that this no philosopher would dare to deny this intuition or if he would be committed to deny it, there would be something gravely wrong with his argument). However, we stumble onto the same seeming paradox if we notice that I am not perceiving this window is a priori false (when the indexical this is understood as an ostensive gesture) and if we are tempted to believe it is necessarily false. This sentence is false whenever it is utterred (at least if the utterance is accompanied by pointing to the window). Berkeley commits the sophism of concluding to the necessary falsity of I might not be perceiving this window right now from the a priori falsity of I am not perceiving this window right now. The consequence of this sophism is that "I am perceiving this window right now" becomes necessarily true, i.e. it becomes true that I must perceive this window right now
This is a very counter-intuitive consequence (and its being counter-intuitive is enough to prove that the process whereby it was obtained is a sophism), but my thesis is that Berkeley is committed to it. And it is easy to notice why he is committed to it. If he weren't committed to it, then he would be obliged to admit that some objects which are perceived are not necessarily perceived, i.e. they are such that they could be unperceived; that is, esse is not necessarily percipi. To admit that esse is not necessarily percipi may seem inocuous for the man in the street, yet Berkeley is bound to reject such a weakening of his central idealist thesis: he may not accept that some objects are such that they MIGHT not be perceived. Surely, what he wants to say is not merely that, as a matter of a long-term luck, all sensible objects happened to be perceived. If this were what he wanted to tell us, then he would not make a philosophically interesting thesis but at most would reveal a curious physiological circumstance.
Like in the case of the sophism that led to when something is perceived, it is perceived by me, which I analyzed in the second section of this essay, it remains to investigate how Berkeley was led to the conclusion that whenever something is perceived, it is necessarily perceived and how this conclusion might be reformulated so as to appear more acceptable. The only way Berkeley can support his thesis that For any x that I perceive, I must perceive it is, again, to deny the existence of a propertiless substratum of the sensible qualities of objects. Indeed, if there were a propertiless substratum of an object’s properties, then it would be possible to correctly identify an object without correctly identifying any of its properties. (It would be possible that the following situation obtains: I perceive an object as cold; that object is not cold; yet I perceive precisely that object. But, as I pointed in the beginning of this section, Berkeley cannot accept such a situation).
As we saw, denying the existence of the substratum amounts to considering all properties of an object essential, and essential to that object only. Thus, whenever I perceive an object, I perceive all its essential properties. Moreover, it is impossible (not only false) that the object should appear to anyone anytime under a different guise from that under which I am perceiving it right now. Thus hyper-essentialism, which is the logical consequence of the denial of the substratum is what makes Berkeley's implicit thesis that "whenever something is perceived I must perceive it" more acceptable.
In the light of the above considerations, I dare to ascribe to Berkeley another formulation of the above thesis: When it is possible for me to perceive an aspect of an object, I perceive a possible aspect of it. (This formulation may seem a gratuitous play of words; in fact, it captures the ideas that to perceive is equivalent in Berkeley with to be able to perceive (I regret the inexistence in English of an infinitive for the verbs "can" and "may"); the existence of the sensible object x is equivalent with the possibility of sensible object x to exist. To put this conclusion in less bombastic terms: for Berkeley I might perceive the tree is equivalent with I perceive how the tree might have been.
Like in the case of the sophism that led to whenever something is perceived, it is me who perceives it, the sophism that leads to whenever something is perceived, it is necessarily perceived the way it is perceived becomes valid if due attention is paid to the distinction between the transcendental and the empirical ego. That is, the transcendental ego is indeed such that all possibilities of an object’s appearing are accessible to it. The transcendental ego – if this concept has any merit in it – is a super-modal one as it is super-temporal. Whenever something is possible, its possibility of appearing is open to the transcendental ego. As applied to the perception of a sensible object, the transcendental ego is such that it has access to all possibilities of an object’s being perceived. It is true that if an object x might be perceived under aspect y, then, if I perceive object x, I actually perceive under aspect y - provided that the "I" is here understood in the sense of the transcendental ego .
3. The homogeneity and ordering of mental faculties
This last section of my essay will have a symmetrical structure with the former two and will analyze another assumption of Berkeley. It is a very speculative section in that I may claim almost no textual evidence for what I ascribe to Berkeley, so what I will be doing is speaking about a "Berkeleyan line of thought" rather than about Berkeley's views as such. I claim to diagnose in Berkeley an assumption that resembles the assumption of the uniformity of the human and divine perceptions. The assumption that I want to analyze is that the aspects of one and the same object, as they are accessible to different mental faculties, are similar, type-identical or – at least – being guaranteed to apply to one and the same object.
This assumption will be diagnosed if we read carefully Philonous' strategy of defeating Hylas in the passage that I quoted in the beginning of my essay, we may notice that he draws two separate conclusions. First, that it is impossible to see a thing (a house or a tree) that is at the same time unseen; second, that it is impossible to conceive a thing (a house or a tree) that is at the same time unconceived. It is not fortuitous that Berkeley gives the same examples in both conclusions: the reason why it is not fortuituous is that Berkeley tacitly assumes that one and the same thing may be accessible to several different mental faculties. Houses and trees may be both seen and conceived. He seems to have uncritically accepted this assumption. But one might ask
What reason do we have to believe that the house (or the tree) that is seen is the same as the house (or the tree) that is conceived?
This question makes a lot of sense given Berkeley’s rejection of a substratum of properties: indeed, for a substantialist the answer to the above question is easy. The substantialist (the materialist, in Berkeley’s terminology) will say:
“No matter if I conceive the house, or see it, or touch it, or refer to it in language, I have access to the house as it is, in itself, as a propertiless substance, as the substratum of the various properties of the house. This substratum could be thought of as a structure of the properties of the house, as that thing (or structure/principle/whatever) that "holds together" the various properties of the house, turning them into the properties of the house".
For Berkeley however such an answer is not available for the obvious reason that he rejects the existence of the substratum. (He provides in The Principles of Human Knowledge a rather naive, Humean avant-la-lettre, answer to the question that I asked above, namely that it is constant associations of some properties that authorize the transition from mere properties to objects. This answer has the obvious shortcoming that it does not make room for the possibility of an object’s drastically changing its properties, or for two objects sharing most of their properties). In the absence of a substratum, there is only one answer that comes to my mind to the question that I asked above. This answer will say something like “all the properties of an object are such that they contain in themselves their being predicated to one and the same object” (There are not properties, but tropes: individually-designed properties such as the red of this carpet, the cold of this piece of ice and these tropes are such that their recognition implies the recognition of the object that they characterize). There is, however, no textual evidence that Berkeley would have endorsed such an ontology of tropes. What I think is that Berkeley would have adopted a different strategy to answer to the question I raised above. His answer is that any mental faculty is more or less transparent to all other mental faculties. That is: whenever I access an object by mental faculty S, then – if S' is another mental faculty – I can grasp by S' the result of my grasping by S. This is what ensures that any act of grasping the visual appearance of an object is accessible to , e.g., the sense of hearing. This thesis of mine (that Berkeley would have accepted such a conclusion) is rather speculative, in that the only textual evidence that I can claim for it is that I have all reasons to ascribe to him the inference from I conceive that I see an object to I conceive an object, i.e. the sense of seeing is transparent to the faculty of conceiving. (If he hadn't made such an inference, then he would have had no reason to speak of the same house as being both seen and conceived).
The problem "what accounts for the unity or co-belongingness of several properties into one and the same object" is particularly acute not only because Berkeley is committed to reject the existence of a substratum, but also because it is intuitive that there are sensible objects that are not accessible to all the senses. Songs cannot be seen, pictures and perfumes cannot be heard and so on.
There is another, more serious, question that Berkeley leaves open. This is related to the intuition that most, perhaps all, mental faculties are sometimes propositional (I imagine that …., I see that…., I hear that…) and sometimes object-directed (I see the tree., I hear the sound…). Related to this, the question may be raised whether, and how, the transition from the propositional to the objectual use of psychological verbs is allowed. It is not obvious that from the premise that I imagine that I see the tree it follows that I imagine the tree. But Berkeley draws this inference, since the premise from which he starts is that I imagine that a tree is seen and the conclusion he draws is that I imagine a tree. I have no textual evidence that should support this thesis, but I claim that Berkeley is committed to it since otherwise he could not support his claim that when I conceive an object and when I see it I have access to one and the same object. (There is, indeed, a subtle difference between imagining that a tree is seen and imagining a seen tree: the first does not commit us to the thesis that imagining is non-propositional, while the second does).
As a matter of historical accuracy, I don’t know if Berkeley explicitly endorsed the inference from I imagine that I see the tree to I imagine the tree. I am inclined to believe he does, since if this inference would be valid then it would be more intelligible how an object (a sensible object) could be accessible to several different mental faculties. This suggests the following principle (again, I am not certain that Berkeley endorsed it, but I suppose that he would have endorsed it):
Where F and G are verbs denoting mental operations, For any F and G and for any x, the following combinations make sense:
I F that I G the x
I G that I F the x
and the following principle holds:
(for any F and G)(I F that I G the x I F the x).
Call this the assumption of psychological homogeneity. It is what ensures that an object that is accessible to several mental faculties (senses) is one and the same object: the way it ensures this is that propositional constructions in which the verb denoting one mental faculty governs a that-clause containing the verb denoting other mental faculty are admissible.
The following lines of this paper are an elaboration and criticism of the assumption of psychological homogeneity. The reason why I criticize this assumption is that the meaning of psychological verbs is such that not all combinations of the form
I F that I G x
where F and G are verbs denoting mental faculties
are admissible. Indeed – no matter what we know or ignore, from a psychological point of view, about seeing or hearing or thinking, it is obvious that it is impossible (since it makes no sense to say) that I touch that I think something, or that I hear that I imagine something. No matter what we know or ignore about our different senses, at least we know that noone can hear that one imagines something. This is not to say that "inferior" senses such as smelling or touching are not propositional (cannot be denoted by the main sentence in a relative construction): perhaps it is possible to smell that so-and-so. This is not the issue. The issue is that relative constructions involving sense-verbs are in general assymetrical in such a way that when F and G are two sense-verbs and it makes sense to say that:
"I F that a is G-ed"
it doesn't always make sense to say that
"I G that a is F-ed".
This suggests a hierarchy (ordering) of senses (and, in general, mental faculties) into “inferior” and “superior” according to the following criterion: Mental faculty F is superior to mental faculty G iff it is possible that I F that I G something, but it is not possible that I G that I F something.
(REVISED PRINCIPLE OF PSYCHOLOGICAL HOMOGENEITY:)
When x is an object accessible to a mental faculty; and F and G are verbs denoting mental faculties; such that F is a faculty superior to G (in the above defined sense of 'superior'), then it is the case that
For any x)(any F)(any G)[(I F that I G x) --> (I F x))]. (1)
This schema could be generalized to:
(for any x)(any F)(any G)(I F that x is G-ed) I F x ) (2)
As I tried to prove in the second section, the IMPORTATION SCHEMA ensures that
(I F that x is G-ed) ( I F that I G x)
and, by transitivity, that
I F the x
The Revised Principle of Psychological Homogeneity is a principle of relative accessibility of the deliverances of a sense (a mental faculty) to other mental faculties. Insofar as conceiving is superior to smelling, this principle ensures that everything that is smelled may also be conceived but not vice versa- the reason being that I can conceive that I smell something but I cannot smell that I conceive something. Inferior faculties are transparent to superior faculties, but not vice versa. The difference between the Principle of Psychological Homogeneity and the Revised Principle of Psychological Homogeneity is that the latter postulates a hierarchy of mental faculties, such that inferior faculties are accessible to superior ones but not vice versa, whereas the former does not.
I don't know whether the ordering of mental faculties postulated by the Revised Principle is strict or not (i.e., whether there are examples of F and G such that it is possible that I F that I G x and that I G that I F x) The only example, about which I discussed above, is that of I see that I hear something and I hear that I see something – which is however debatable since it could be held that, after all, seeing that something was heard and hearing that something was seen are possible only indirectly, i.e. if one sees the signs that something was heard, or hears the signs that something was seen; in both cases, what is seen/heard are not merely the objects of some sense, but some propositional contents.
Correlated with the Revised Principle of Metapsychological Homogeneity, I postulate another principle:
When F and G are mental faculties such that F is superior to G, then if I G x, then it is possible that
(a) I F that I don’t G the x; that
(b) I F that I might have G-ed the x; that
(c) I F that I might not have G-ed the x
and
each of (a),(b) and (c) entails that I F the x.
Call this principle the Modal Version of the Revised Principle of Psychological Homogeneity. Indeed, it makes sense to say that I conceive that I don’t smell a flower, that I conceive that I might not have smelled the flower I actually smell and that I conceive that I might have smelled the flower that I don’t actually smell (but that would still be the same flower, which is presented to me through a different sense, or to a different mental faculty, than to smell). This is, of course a deviation from Berkeley’s original intention: according to Berkeley, as I held in the third section of this essay, I might have smelt the flower is equivalent to I smell how the flower might have been. (The reason is that, according to Berkeley, the flower and a fortiori the flower’s smell could not have been other than how I smell it).
Another principle that I postulate is (and that I would call the Principle of Modal Inclusion) is:
If F is superior to G, and I F x, and x could be G-ed, then I F how I would G x.
This principle seems intuitive. If I conceive a visual object, then I moreover conceive how it would look like if I saw it. If I refer in language to an auditive object, then I refer to how it would be heard if I listened to it. It is the principle of subordination of mental faculties that ensures the dependence of all objects on the supreme mental faculty. It is in this sense that the supreme mental faculty – be it the faculty of referring in language to , or of imagining, provides access to the objects themselves, that is, to more than the aspect under which the object momentarily happens to appear to my empirical mind.
Psychological homogeneity plays a crucial role in Berkeley's argument, as I understand it. Not only that it ensures that one and the same object may be accessible to several different mental faculties. By ensuring the validity of the inference from I imagine a seen tree to I imagine a tree, it ensures the mind-dependency of the objects that are not actually perceived, but are only thought or conceived to be perceived. However, Berkeley makes no mention of any hierarchical ordering of several mental faculties: he accepts the Psychological Homogeneity rather than the Revised Principle of Psychological Homogeneity.
But if the hierarchical account of mental faculties is correct, we have another corollary. A mental faculty that is superior in the sense that it can be denoted by the main verb of a relative construction which involves another psychological verb is a faculty that provides access to the object as it is independently on its appearing under a certain guise: this is the immediate corrolary of The Modal Version of the Revised Principle of Psychological Homogeneity and of the Principle of Modal Inclusion. Indeed, if I imagine that I smell a flower, then I also imagine that I could have not smelt the flower, or how I might have smelt the flower and so on. Moreover, if I imagine a flower, I imagine how it would have smelt if I had smelt it. On the other hand, an inferior mental faculty (such that smelling or touching) can only provide access to a sensible aspect of an object: to its olphactory aspect or to its tactile aspect. Thus, a superior mental faculty is closer to providing access to the object as it appears under all its aspects. At the limit: if a faculty were such that it could govern that-clauses that denote any other psychological verb, then it would provide access to the object as it appears to any mental faculty. (This is not yet the same as providing access to the substratum of the objects but it is as close as it may be: indeed, the substratum of objects is, on a minimal understanding, nothing but the principle in virtue of which all the properties of an object belong to that object). Call such a faculty, that is superior to any other mental faculty, the Supreme Mental Faculty.
The suggestion towards which I want to push this essay is that there is an obvious parallel between the transcendental subject and the supreme mental faculty (in the sense defined above). The supreme mental faculty is that faculty which may be denoted by the main phrase of that-clauses containing any psychological verb. As an empirical matter, I don't know which of the mental faculties deserves the title of supreme mental faculty: it may be the faculty of conceiving, or of referring in language to. As a matter of fact, in the examples I gave imagining appeared to be closer to the Supreme Mental Faculty than touching or hearing, or even seeing. The reason why this happened is, in my opinion, merely that it makes sense to say touching
(i) I imagine that I (don't) touch/see/hear the tree).
but it doesn't make sense to say:
(i) I touch that I (don't) imagine the tree
Whether imagining or conceiving, as Berkeley puts it, is indeed as close as it can be to the Supreme Mental Faculty is an empirical matter, that needs not concern us here. (Perhaps there are other creatures, call them Martians, who have a faculty of ENVISAGING such that it makes sense – for them – to say "I ENVISAGE that I conceive x" but it makes no sense to say "I conceive that I ENVISAGE x"); however, the hierarchy of faculties that is assumed in our natural language is not an empirical matter, since I do not think that it is possible to discover any tribe whose language allows constructions of the type "I smell that I would have thought that…" or "I touch that I don't imagine…" It should be noted that the postulation of The Supreme Mental Faculty – be it what it may – is a betrayal of Berkeley's ontology since this faculty provides access to the substratum of objects, i.e. to the faces of an object that are accessible only to a certain sense and not to other senses.
5. Conclusion
The main point of this essay was to argue that Berkeley’s sensualist version of idealism is wrong insofar as it is committed to a trivial version of solipsism. I analyzed the sophisms that Berkeley uses in order to make his sensualist idealism look more appealing and I argued that these sophisms may be transformed into valid arguments once the empirical subject is replaced by the transcendental subject. While arguing against Berkeley's sensualist idealism and in favour of transcendental idealism, I simultaneously argued for a propertiless substratum of objects, that is accessible to the transcendental subject though not necessarily to the empirical one.
In a first step, the transcendental ego that I invoked is the “limit of the world”, what conditions the possibility of objects' existing and being referred to. In a second step, it appeared as what is common to all empirical human minds and, moreover, to all possible human minds. The transition from the first to the second step is, in my opinion, a rather simple one: the possibility is indeed what conditions actuality. In order for an act of seeing to be possible (and for a visible object to be seen) , the possibility it needs to be secured. As I understand the relation between the transcendental subject and the empirical one(s), it is a relation between possibility and actuality: in order for my present act of seeing this visible-object-here to be actual, it needs to be a possible act of seeing, to belong to the same species as other acts of seeing – of mine or of other empirical subjects. More generally: in order for my present act of seeing this object to be actual, it needs to be a possible act of having access to this object. Thus, the more we generalize my current act of seeing this object, the closer we get to the substratum of this object. (This parallel generalization is obvious: the less specific we are about the act of referring to this object, the less specific we are about the object itself).
There is a beautiful symmetry between the transcendental subject and the substratum of objects. The transcendental subject is what provides access to the substratum of the objects to the extent at which such substratum is itself prior to the possibility of objects' appearing under this or that aspect to this or that human mind.
It goes without saying that Berkeley would have abhorred this discourse about possibilities-as-preceding actualities and of transcendental subject, seeing in it an instance of the much detested doctrine of abstract ideas. The price he pays for avoiding the transcendental subject, the substratum and the possibilities-as-preceding-actualities is his assumption of the uniformity of the human (and divine) perceptions. However, as I hope to have proved, even so he does not convincingly avoid present-moment solipsism, nor does he succeed in avoiding conflict with commonsense intuitions, such that the intuition that objects may exist even in the absence of my (empirical mind’s) perceiving them; the intuition that sensible objects might be (and appear) otherwise than they do appear to me.
Another problem should be addressed in this context. As it is known, God is an actor of Berkeley’s ontology. God is the mind that perceives objects in the absence of any human mind’s perceiving them. The question might be raised: is it not Berkeley’s God exactly the transcendental subject that I argued for? Indeed, Berkeley’s God is such that if any object is presented to a mental faculty, then it is presented to God; and is such that if any object is presented under a certain aspect, it necessarily has that aspect. (God cannot be mistaken in perceiving an object as cold without that object being cold). This suggests that Berkeley’s God could indeed play the role that the transcendental ego which I invoked plays in an idealist ontology. But then Berkeley could no longer claim that God has the same perceptions as humans have.
As I see the things, the most shocking aspect of Berkeley's theological and empiricist version of idealism is his claim that God’s mind is an exact replica of human minds, (I am not particularly knowledgeable in Judeo-Christian theology, but I believe that the traditional picture was this: the human mind has some "inferior" levels, such as sensibility, that it shares with the beasts and a "superior" level – the reason or the imagination – that it shares with God. At least this is the picture of the human mind according to Maimonides). Berkeley is however not sensitive to such distinction: for him God is “one of us” in that he perceives and has sensations exactly like humans do. Berkeley’s God is antropomorphised to the extent at which its mental states are indiscrimimable from those of humans .
It is this kind of antropomorphic God that cannot be equated with the transcendental ego. For the transcendental ego, as I understand it, is precisely a precondition of the world’s existence and of my referring to any object at all. The transcendental ego is thus indispensable from any act of referring to an object – by senses, imagination or language And the need for the transcendental ego is felt precisely because any empirical ego; any empirical act of referring to an object is dispensable. (The object would remain the same even if my access to it were different, or absent). But nothing in Berkeley’s God corresponds to these requirements. Berkeley’s God is a replica of an ordinary human mind: He is neither more nor less dispensable than they are. Berkeley’s God cannot be the transcendental ego because the transcendental ego would make useless the assumption of the uniformity of human perceptions – in that once the transcendental ego is recognized, it is no longer necessary that my perceiving this tree and your perceiving this tree should be identical mental states. (The reason why this is no longer necessary is that there is something common in both – namely the condition of possibility of perceiving this tree).