Skip to main content
Article

How Berkeley Can Go Quasi and Lose His Mind-Dependence

Author
  • Bart Streumer

Abstract

Berkeley defends idealism partly by taking ordinary objects such as tables and stones to consist of ideas. Materialists reject this account of ordinary objects. I argue that Berkeley can undermine this rejection by making quasi-realist moves of the kind that many expressivists about normativity make. This ultimately results in a version of idealism that we cannot believe, since trying to believe it will make us believe materialism. But I argue that this version of idealism may nevertheless be true.

Keywords: Berkeley, Idealism, Quasi-Realism, Normativity, Mind-Dependence

How to Cite:

Streumer, B., (2026) “How Berkeley Can Go Quasi and Lose His Mind-Dependence”, Philosophers' Imprint 26: 27, 1–17. doi: https://doi.org/10.3998/phimp.7876

446 Views

77 Downloads

Published on
2026-10-05

Peer Reviewed

Berkeley defends idealism partly by taking ordinary objects such as tables and stones to consist of ideas. Materialists reject this account of ordinary objects. I will argue that Berkeley can undermine this rejection by making quasi-realist moves of the kind that many expressivists about normativity make. This results in a version of idealism that we cannot believe but that may nevertheless be true: a version that fits with Hume’s claim that Berkeley’s arguments “admit of no answer and produce no conviction”.1

In Section 1 I will outline Berkeley’s account of ordinary objects. In Section 2 I will argue that he should be an expressivist about the normativity that plays a key role in his account. In Section 3 I will argue that Berkeley can undermine materialists’ rejection of his account by copying expressivists’ quasi-realist moves. In Section 4 I will show that this ultimately results in a version of idealism that may be true but that we cannot believe, since trying to endorse this version makes us believe materialism. In Section 5 I will ask what defenders of this version should believe instead.

1. Berkeley’s account of ordinary objects

George Pitcher writes that Berkeley’s view “rises in the garden of British thought like some fantastic plant – beautiful and extravagant”.2 For there is a striking difference between this plant and the rest of the garden: whereas his predecessors and contemporaries are materialists who think that matter exists, Berkeley is an idealist who thinks that only minds and ideas exist.

Materialists take ordinary objects such as tables and stones to consist of matter. Berkeley’s view may therefore seem to entail that such objects do not exist. But he prevents his view from entailing this by taking such objects to consist of ideas. This account of ordinary objects is a crucial part of Berkeley’s defence of idealism.3 Without it, his view would arguably have been refuted by Johnson when he famously “answered, striking his foot with mighty force against a large stone, till he rebounded from it, ‘I refute it thus’”.4

But how exactly does Berkeley take ordinary objects to consist of ideas? There are three main interpretations. According to the first, he thinks that

  • (O1)

    Ordinary objects are collections of ideas.5

This interpretation is suggested by many passages in his work, such as:

Take away the sensations of softness, moisture, redness, tartness, and you take away the cherry. Since it is not a being distinct from sensations, a cherry, I say, is nothing but a congeries of sensible impressions or ideas perceived by various senses.6

Consider the sentence ‘This table is rectangular’. Materialists take ‘this table’ to refer to an object that consists of matter and take ‘is rectangular’ to ascribe the property of being rectangular to this object. By contrast, if (O1) is true, ‘this table’ refers to a collection of ideas and ‘is rectangular’ says that certain visual and tactile ideas are members of this collection.7

Does this entail that ordinary objects are mind-dependent? Not in the sense that

  • (M1)

    Our minds bring ordinary objects into existence.

For Berkeley thinks that when we perceive ordinary objects the relevant ideas are put into our minds by God.8 That is how he distinguishes perceiving a table from merely imagining a table. But (O1) does entail that these objects are mind-dependent in the sense that

  • (M2)

    Ordinary objects cease to exist when we do not perceive them.

You may think that Berkeley can prevent (O1) from entailing this by taking ‘this table’ to refer to a collection of ideas in the mind of God.9 But since God has all possible ideas, the vast majority of collections of ideas in God’s mind constitute merely possible objects. To distinguish real objects from merely possible ones, Berkeley has to say that an object exists only when God has put the relevant ideas into our minds, which makes (O1) entail (M2) again.10

Since Berkeley can be read as being committed to (M2), this is not a fatal objection to this interpretation.11 But materialists will be unhappy about it: they take ordinary objects to consist of matter partly because they take them to be mind-independent.12 And the other two interpretations avoid (M2). According to the second, Berkeley thinks that

  • (O2)

    Sentences about ordinary objects are made true by conjunctions of subjunctive conditionals about ideas we would have in certain circumstances.

This interpretation is defended by Kenneth Winkler.13 It is suggested by the following passage:

The table I write on, I say, exists, that is, I see and feel it; and if I were out of my study I should say it existed, meaning thereby that if I was in my study I might perceive it, or that some other spirit actually does perceive it.14

Consider again the sentence ‘This table is rectangular’. If (O2) is true, this sentence is made true by a conjunction of conditionals of the following form:

If we put ourselves in such-and-such circumstances, we would have such-and-such ideas.

Since the truth of such conditionals does not depend on whether we actually have these ideas, (O2) does not entail (M2). But it does make ‘This table is rectangular’ equivalent to an indefinitely long conjunction that contains many far-fetched conditionals, such as:

If I put on sunglasses, rolled over twice underneath this table, tried to get up and then hit my head against the table top, I would have such-and-such visual and tactile ideas.

As Kenneth Pearce points out, this makes it easy for a Cartesian demon to deceive us about ordinary objects: for example, to deceive us about the table’s currently being rectangular, the demon merely needs to make one of these far-fetched conditionals false, which he can do by forming a single aberrant counterfactual intention.15 That fits badly with Berkeley’s opposition to scepticism about ordinary objects.16

Since (O2) leaves room for what Winkler calls “conjectures” that “deserve to be counted as knowledge”, this is not a fatal objection to this interpretation.17 But materialists will again be unhappy about it: they oppose scepticism about ordinary objects just as much as Berkeley does. And the third interpretation avoids this problem. According to this interpretation, Berkeley thinks that

  • (O3)

    Sentences about ordinary objects are made true by ideas in combination with linguistic rules that take these ideas as inputs and yield these sentences as outputs.

This interpretation is defended by Pearce.18 It is suggested by the following passage:

And as several [ideas] are observed to accompany each other, they come to be marked by one name, and so to be reputed as one thing. Thus, for example, a certain colour, taste, smell, figure and consistence having been observed to go together, are accounted one distinct thing, signified by the name ‘apple’.19

Pearce takes this passage to suggest (O3) because of the intermediary role it gives to language: it starts with ideas occurring together, which are then “marked” by a single name, which brings into existence a distinct thing that is “signified” by the name but that does not coincide with the ideas.20

Consider again the sentence ‘This table is rectangular’. If (O3) is true, this sentence is made true by ideas in combination with linguistic rules that take us from these ideas to the sentence ‘This table is rectangular’. This reverses the materialist order of explanation. If materialism is true, ‘This table is rectangular’ is true in virtue of the existence of a table that has the property of being rectangular. By contrast, if (O3) is true, the table exists and has the property of being rectangular in virtue of the truth of the sentence ‘This table is rectangular’. This turns the table and the property of being rectangular into what we can call semantic shadows of ‘this table’ and ‘is rectangular’.21

Pearce takes this to entail that “the introduction of the conventions for the word [‘table’] creates [tables] more generally”.22 But that is not quite right. For (O3) takes ‘This table is rectangular’ to be made true by linguistic rules in combination with ideas that these rules take as inputs. To ensure that (O3) does not entail (M2), Berkeley should say that these ideas need not only be ideas we currently have, but can also be ideas we would have in alternative circumstances. This incorporates the core insight of (O2) into (O3). But it does so without making it easy for a Cartesian demon to deceive us about ordinary objects, since whether the table is currently rectangular depends only on our current ideas in combination with relevant linguistic rules.

You may think that (O3) is overly linguistic. But ‘linguistic rule’ is just a label for a function that takes us from ideas to true sentences, and the role sentences play in (O3) could also be played by propositions. We could therefore instead take (O3) to say that

  • (O3′)

    Propositions about ordinary objects are made true by ideas in combination with functions that take these ideas as inputs and yield these propositions as outputs,

as long as we do not take these propositions and functions to be abstract objects, since that would conflict with Berkeley’s rejection of abstraction.23

Which interpretation is best? That depends on what we are trying to do. The interpretation that stays closest to what Berkeley most often says is clearly (O1).24 But I think that the interpretation that makes his view most defensible is (O3). I will therefore use (O3) as my starting point. My arguments can also be combined with other interpretations that give a central role to linguistic rules, however, including promising versions of (O1) and (O2).25 Moreover, my arguments matter even if you prefer a different interpretation. For instead of taking them to be about what Berkeley can say, you can also take them to be about what endorsing (O3) enables idealists to say.

2. The role of normativity

If (O3) is true, whether there is a stone in front of us depends on whether the sentence ‘There is a stone in front of us’ is true, which in turn depends on whether this sentence is yielded as an output by linguistic rules that take our ideas as inputs. But which rules are these?

They may simply be the linguistic rules we happen to follow.26 Suppose that we follow rules that take some of our current ideas as inputs and yield the sentence ‘There is a stone in front of us’ as output. If (O3) is about the rules we happen to follow, there is then a stone in front of us. But (O3) does not stop us from following different rules that take the same ideas as inputs but yield the sentence ‘There is no stone in front of us’ as output. If (O3) is about the rules we happen to follow, there is then no stone in front of us. This therefore makes Berkeley’s view entail that

  • (M3)

    If we had followed different linguistic rules, different ordinary objects would have existed.

That may seem innocuous: (M3) may merely seem to say that if we had followed different rules, we would have carved up the same environment in a different way. But this cannot be what (M3) says if (O3) is true. For (O3) treats objects and properties as semantic shadows of sentences yielded by rules that take our ideas as inputs, which means that there is no environment that exists independently of these sentences: our environment is then just as much a semantic shadow of these sentences as the objects and properties it consists of. What (M3) says is therefore that if we had followed different linguistic rules, we would have had a different environment.27 This makes ordinary objects mind-dependent again in a sense that materialists will be unhappy about.28

Berkeley can avoid (M3) by taking (O3) to be a normative claim: by taking these rules to be the correct linguistic rules. But what is it for a linguistic rule to be correct? You may think Berkeley can say that this is for it to track regularities in our flow of ideas. That may be what he has in mind when he writes that

men combine together several ideas, apprehended by divers senses, or by the same sense at different times or in different circumstances, but observed however to have some connection in nature, either with respect to co-existence or succession; all which they refer to one name and consider as one thing.29

But this is less straightforward than it may seem. Berkeley takes both minds and ideas to be particulars, which are non-repeatable entities, and denies that there are universals, which are repeatable entities that can be instantiated by different particulars.30 And he often suggests that similarities between ideas obtain in virtue of our subjective propensities to group ideas together: for example, he writes that

this naming and combining together of ideas is perfectly arbitrary, and done by the mind in such sort as experience shows it to be most convenient: without which our ideas had never been collected into such sundry distinct combinations as they now are.31

But objective regularities require objective similarities. If Berkeley takes similarities between ideas to be subjective, he must therefore take regularities in our flow of ideas to be subjective in the same way: he must take them to obtain in virtue of our subjective propensities to group ideas together. If he says that what it is for a linguistic rule to be correct is that it tracks regularities in our flow of ideas, his view therefore seems to entail that

  • (M4)

    If we had had different subjective propensities to group ideas together, different ordinary objects would have existed.

This once again makes such objects mind-dependent in a sense that materialists will be unhappy about.32

You may think Berkeley can avoid (M4) by reconsidering his rejection of universals. But this would get him into trouble elsewhere: his rejection of universals plays an essential role in his rejection of abstraction, which in turn plays an essential role in his rejection of materialism.33 Alternatively, you may think he can avoid (M4) by endorsing an account of objective similarity that does not appeal to universals, such as trope theory or resemblance nominalism.34 But these accounts are almost as hard to square with Berkeley’s rejection of abstraction as the view that there are universals: both trope theory and resemblance nominalism appeal to an objective notion of resemblance that may not be a universal but that certainly seems abstract.35

You may also think that Berkeley can say that (M4) is unproblematic because God has given us similar subjective propensities to group ideas together.36 But if there are no objective similarities, God cannot make our subjective propensities to group ideas together objectively similar. He can then at most make these subjective propensities subjectively similar to Him: similar in virtue of God’s subjective propensity to group them together. This not only fails to ensure that our subjective propensities to group ideas together are objectively similar, but even fails to ensure that these subjective propensities are subjectively similar to each of us.

These problems may not be fatal. Perhaps Berkeley can give an account of objective similarity that does not clash with his rejection of abstraction, and perhaps there is a way to show that (M4) is unproblematic by appealing to God. But it is certainly not obvious how to do this. It is therefore worth investigating whether Berkeley can avoid (M4) in a different way. And I think he can.

If Berkeley takes the correctness of linguistic rules to consist in tracking regularities in our flow of ideas, he in effect endorses reductive realism about this correctness: he takes it to be identical to the descriptive property of tracking such regularities. I think he should instead endorse a different view about this correctness: expressivism, according to which the sentence ‘Linguistic rule R is correct’ expresses a non-cognitive attitude, such as approval of rule R, and the sentence ‘Linguistic rule R is incorrect’ expresses an incompatible non-cognitive attitude, such as disapproval of rule R. This may seem an odd view to hold about the correctness of linguistic rules. But as I have said, ‘linguistic rule’ is just a label for a function that takes us from ideas to true sentences. There is no need to think of these functions as ‘linguistic’ in the sense that they are studied by linguistics. Moreover, Berkeley does not have to take sentences about the correctness of these rules to express attitudes of approval or disapproval, but can also take them to express non-representational states of some other kind.37 And though the historical Berkeley was not an expressivist, the view is not completely alien to him: he himself says that

the communicating of ideas marked by words is not the chief and only end of language, as is commonly supposed. There are other ends, as the raising of some passion, the exciting to, or deterring from an action, the putting the mind in some particular disposition.38

More broadly, expressivism fits with Berkeley’s pragmatist tendencies.39

Does expressivism about the correctness of linguistic rules enable Berkeley to deny that ordinary objects are mind-dependent? You may think it does not, since you may think it makes his view entail that

  • (M5)

    If we had approved of different linguistic rules, different ordinary objects would have existed.

But expressivists have a long-standing way to deal with such worries about mind-dependence. Their way to deal with these worries is reminiscent of Berkeley’s response to materialists.40 And I think it enables Berkeley to take his response much further.41

3. Quasi-materialism

Suppose that the sentence ‘Action A is right’ expresses approval of action A and that the sentence ‘Action A is wrong’ expresses disapproval of action A. This may seem incompatible with the claim that

  • (1)

    If we had not thought that action A is wrong, action A would still have been wrong.

Our inclination to endorse claims like (1) may therefore seem to show that expressivism is false. But expressivists argue that their view is compatible with such claims by making two moves. The first is to interpret (1) as an expression of a non-cognitive attitude: for example, they may take (1) to express

  • (1*)

    Disapproval of performing-action-A-in-circumstances-in-which-we-do-not-disapprove-of-performing-action-A.

If (1) expresses this attitude, expressivism is compatible with this claim. And expressivists can repeat this move for other claims that may seem incompatible with their view. For example, they may take the claim that

  • (2)

    It is possible to be mistaken about whether an action is wrong

to express

  • (2*)

    Approval of being willing to revise attitudes of disapproval in response to new information,

and they may take the claim that

  • (3)

    When one person thinks that an action is wrong and another person thinks that this action is right, at least one of them is mistaken

to express

  • (3*)

    Disapproval of one person approving and another person disapproving of a single action in response to the same facts.42

If (2) and (3) express these attitudes, expressivism is compatible with these claims as well.

We may think that (1) to (3) instead express true beliefs. But expressivists’ second move is to endorse minimalism: they take the sentence ‘“p” is true’ to express the same mental state as the sentence ‘p’, they take a sentence to express a belief if and only if it can be true, and they take an object to have a property if and only if this property can be ascribed to it with a true sentence.43 If minimalism is true, the sentence ‘Action A is wrong’ expresses a belief, expressivists can take this belief to be true, and they can take action A to have the property of being wrong. And they can then also take (1) to (3) to express true beliefs.44

These moves result in a sophisticated version of expressivism that is known as quasi-realism. There are different versions of this view: instead of taking (1) to (3) to express higher-order attitudes of approval or disapproval such as (1*) to (3*), quasi-realists can also take (1) to (3) to express different non-cognitive attitudes, and instead of endorsing minimalism, they can also endorse a different expressivism-friendly account of truth, beliefs and properties.45 Moreover, it is controversial whether these quasi-realist moves work. But suppose they do. In that case, I think Berkeley can avoid (M5) by copying these moves.

Just as quasi-realists do, Berkeley can take the claim that

  • (4)

    If we had thought that different linguistic rules are correct, the same ordinary objects would have existed

to express a non-cognitive attitude: for example, he may take it to express

  • (4*)

    Disapproval of thinking-that-different-ordinary-objects-exist-in-circumstances-in-which-we-approve-of-different-linguistic-rules.

If (4) expresses this attitude, Berkeley’s view is compatible with this claim. And given expressivism, (4) entails the negation of (M5): it entails that if we had approved of different linguistic rules, the same ordinary objects would have existed. This therefore enables Berkeley to avoid (M5).

Berkeley can also copy what quasi-realists say about (3) and (4): he can take the claim that

  • (5)

    It is possible to be mistaken about whether a linguistic rule is correct

to express

  • (5*)

    Approval of being willing to revise attitudes of approval of linguistic rules in response to new information,

and he can take the claim that

  • (6)

    When one person thinks that an linguistic rule is correct and another person thinks that this linguistic rule is incorrect, at least one of them is mistaken

to express

  • (6*)

    Disapproval of one person approving and another person disapproving of a single linguistic rule.

If (5) and (6) express these attitudes, Berkeley’s view is compatible with these claims as well.46

As before, we may think that (4) to (6) instead express true beliefs. But Berkeley can also copy quasi-realists’ endorsement of minimalism. If minimalism is true, the sentence ‘Linguistic rule R is correct’ expresses a belief, Berkeley can take this belief to be true, and he can take rule R to have the property of being correct. And he can then also take (4) to (6) to express true beliefs.

If these moves enable quasi-realists to deny that rightness and wrongness are mind-dependent, I think they similarly enable Berkeley to deny that the correctness of linguistic rules is mind-dependent. And this in turn enables him to deny that ordinary objects are mind-dependent. For if (O3) is true, (4) entails that

  • (4†)

    If we had thought that different ordinary objects existed, the same ordinary objects would in fact have existed,

(5) entails that

  • (5†)

    It is possible to be mistaken about whether an ordinary object exists,

and (6) entails that

  • (6†)

    When one person thinks that an ordinary object exists and another person thinks that this ordinary object does not exist, at least one of them is mistaken.

Becoming an expressivist about the correctness of linguistic rules and copying these quasi-realist moves therefore enables Berkeley to reject not only (M5), but also other claims about mind-dependence, including (M1) to (M4). And it thereby enables him to deny that ordinary objects are mind-dependent in any sense at all.

This results in a version of idealism that we can call quasi-materialism. This view clearly goes significantly beyond what the historical Berkeley says. But its building blocks are not completely alien to him: as we have seen, some of his claims are reminiscent of expressivism, this view fits with his pragmatist tendencies, and (O3)‘s treatment of objects and properties as semantic shadows is reminiscent of minimalism. More generally, Berkeley often responds to opponents by interpreting what they say in a way that makes it consistent with idealism, which is reminiscent of the quasi-realist moves I have outlined.47 The main difference between quasi-materialism and Berkeley’s view is that quasi-materialism extends such moves further than he does and thereby undermines materialists’ rejection of his account of ordinary objects to a greater extent.

Moreover, I think quasi-materialism is just as well-motivated as quasi-realism. For most of us are more certain that ordinary objects are mind-independent than that rightness and wrongness are mind-independent. Berkeley’s need to deny that ordinary objects are mind-dependent is therefore at least as great as expressivists’ need to deny that rightness and wrongness are mind-dependent. And this is what quasi-materialism enables him to do.48

4. Super-quasi-materialism

But materialists will still think that something important is missing from quasi-materialism. Suppose that you and another person have the same relevant ideas but follow different linguistic rules: you follow rule X, which takes these ideas as inputs and yields the sentence ‘There is a table in the room’ as output, but the other person follows rule Y, which takes these ideas as inputs but yields the sentence ‘There is no table in the room’ as output. Moreover, suppose you have the following attitudes:

  • Approval of rule X.

  • Disapproval of thinking-that-different-ordinary-objects-exist-in-circumstances-in-which-you-approve-of-rule-Y.

  • Approval of being willing to revise your attitude of approval of rule X in response to new information.

  • Disapproval of one person approving and another person disapproving of rule X.

And suppose the other person has the following attitudes:

  • Approval of rule Y.

  • Disapproval of thinking-that-different-ordinary-objects-exist-in-circumstances-in-which-they-approve-of-rule-X.

  • Approval of being willing to revise their attitude of approval of rule Y in response to new information.

  • Disapproval of one person approving and another person disapproving of rule Y.

If quasi-materialism is true, you then believe that there is a table in the room, you take this belief to be true, and you endorse (4†) to (6†) while taking these claims to express true beliefs. But to materialists there will then nevertheless seem to be an alarming symmetry between you and this other person. What is needed to break this symmetry, materialists will think, is that reality determines which belief is true: your belief that there is a table in the room or the other person’s belief that there is no table in the room. And they will think that reality can only play this role if it contains matter that is distributed in such a way that there either is or is not a table in the room.

Can Berkeley also say that reality determines which belief is true? What gives rise to this symmetry is that you and the other person have the same relevant ideas but follow different linguistic rules.49 For Berkeley to be able to say that reality determines which belief is true, he therefore has to say that

  • (7)

    Reality determines which linguistic rules are correct.

And I think he can say this. For just as he can take (4) to (6) to express non-cognitive attitudes, he can take (7) to express another non-cognitive attitude: for example, he may take it to express

  • (7*)

    Approval of forming attitudes of approval or disapproval of linguistic rules in response to subjective regularities in our flow of ideas.

As before, if (7) expresses this attitude, quasi-materialism is compatible with this claim.

Materialists will no doubt deny that this breaks the symmetry between you and the other person: they will complain that it merely adds another non-cognitive attitude to your and the other person’s attitudes. And they will conclude from this that

  • (8)

    (7) does not express a non-cognitive attitude.

But Berkeley can then respond in the same way: he can take (8) to express another non-cognitive attitude, such as

  • (8*)

    Disapproval of taking (7) to express a non-cognitive attitude.

And he can keep making such moves in response to any further claim of this kind.

Materialists may also try to spell out what they take to be missing from quasi-materialism by making more general claims: for example, they may say that

  • (9)

    Claims about the correctness of linguistic rules do not express non-cognitive attitudes.

But Berkeley can then simply keep going: he can take such general claims to express non-cognitive attitudes as well, such as

  • (9*)

    Disapproval of taking claims about the correctness of linguistic rules to express non-cognitive attitudes.

If he keeps making such moves, he combines quasi-materialism with what I have elsewhere called super-quasi-realism: an extreme version of quasi-realism that treats any claim about what is missing from quasi-realism as an expression of another non-cognitive attitude.50

Materialists may also try to spell out what they take to be missing from quasi-materialism by insisting that what is needed to break the symmetry between you and the other person is that

  • (10)

    Ordinary objects consist of matter.

This clearly begs the question against Berkeley. But merely pointing out that your opponents beg the question is never a strong response, particularly if your view is as counterintuitive as Berkeley’s idealism. And I think he can do better: since we are already taking him to say that

  • (O3)

    Sentences about ordinary objects are made true by ideas in combination with linguistic rules that take these ideas as inputs and yield these sentences as outputs,

he can similarly say that

  • (11)

    Sentences about matter are made true by ideas in combination with linguistic rules that take these ideas as inputs and yield these sentences as outputs,

and he can combine (11) with super-quasi-realism about the correctness of these rules.51

Of course, Berkeley frequently denies that matter exists. But he does come close to (11) in the following passage:

Hylas: What think you therefore of retaining the name ‘matter’ and applying it to sensible things? …

Philonous: With all my heart: retain the word ‘matter’ and apply it to the objects of sense, if you please, provided you do not attribute to them any subsistence distinct from their being perceived.52

Though this is clearly not Berkeley’s considered view, it shows that even (11) is not completely alien to him.

Finally, materialists may insist that what is needed to break the symmetry is not just the truth of (10) but the falsity of claims like (O3) and (11): they may say that what is needed is that

  • (12)

    Ordinary objects and matter exist not merely as semantic shadows of sentences that are made true by ideas in combination with linguistic rules.

Like Mill, they can say that (12) is made true by the meaning of ‘object’ and ‘matter’ and that this claim therefore follows from the correct linguistic rules.53 But I think Berkeley can then again keep going: he can agree with materialists that

  • (13)

    (12) follows from the correct linguistic rules,

but he can take this claim to express

  • (13*)

    Approval of linguistic rules that take ideas as inputs and yield (12) as output,

and he can again combine (13) with super-quasi-realism about the correctness of these rules.

When quasi-realists take (1) to (3) to express non-cognitive attitudes, they normally make use of these interpretations to endorse these claims. Suppose that Berkeley does this as well when he makes the moves I have just described. This results in a version of idealism that we can call super-quasi-materialism. Whereas quasi-materialism starts from (O3) and goes quasi-realist about the correctness of the relevant linguistic rules, super-quasi-materialism starts from quasi-materialism and goes super-quasi-realist about the correctness of the relevant linguistic rules, and then repeats this for (11) and (13).

Moving towards super-quasi-materialism is a dynamic process. For Berkeley to coherently endorse the claim that

  • (8)

    (7) does not express a non-cognitive attitude,

he will have to stop taking the claim that

  • (7)

    Reality determines which linguistic rules are correct

to express a non-cognitive attitude. To coherently endorse the claim that

  • (9)

    Claims about the correctness of linguistic rules do not express non-cognitive attitudes,

he will have to stop taking any claim about the correctness of linguistic rules to express such an attitude, including (9) itself. He will then no longer take (9) to express

  • (9*)

    Disapproval of taking claims about the correctness of linguistic rules to express non-cognitive attitudes,

but this does not matter: what enables him to endorse (9) is that this claim in fact expresses this attitude, not that he himself takes it to express this attitude. To coherently endorse the claim that

  • (10)

    Ordinary objects consist of matter

in the way I have described, he will have to stop denying that matter exists. And to coherently endorse the claim that

  • (12)

    Ordinary objects and matter exist not merely as semantic shadows of sentences that are made true by ideas in combination with linguistic rules

in the way I have described, he will have to stop taking ordinary objects and matter to exist merely as semantic shadows of such sentences. At the end of this process Berkeley will therefore have endorsed every claim that materialists endorse and will have rejected every claim that materialists reject. In other words, at the end of this process he will have become a materialist.54

This may seem to show that super-quasi-materialism and materialism are the same view. But they are not. For super-quasi-materialism to be true, the world and our mental states must be such that

  • (14)

    (O3) and (11) are true

and that

  • (15)

    (4) to (9) and (13) express the attitudes described by (4*) to (9*) and (13*).

If those who try to endorse super-quasi-materialism thereby become materialists, they thereby come to endorse the negation of each conjunct of (14) and (15): in other words, they come to believe that

  • (~14)

    (O3) and (11) are false

and that

  • (~15)

    (4) to (9) and (13) do not express the attitudes described by (4*) to (9*) and (13*).

But this is compatible with the world and our mental states in fact being such that (14) and (15) are true. Super-quasi-materialism and materialism have different truth-conditions: whereas for super-quasi-materialism to be true (14) and (15) must be true, for materialism to be true (∼14) and (∼15) must be true. This means that super-quasi-materialism and materialism are different views. That trying to endorse super-quasi-materialism turns Berkeley into a materialist merely shows that we cannot believe super-quasi-materialism, since trying to endorse this view makes us believe a different view instead.55

Super-quasi-materialism clearly goes far beyond what the historical Berkeley says. And there are no contemporary idealists who defend this view. But super-quasi-materialism has several advantages over standard Berkeleyan idealism. It undermines materialists’ rejection of Berkeley’s account of ordinary objects to the greatest possible extent. It avoids Berkeley’s objections to materialism, since the existence of matter is not part of super-quasi-materialism’s truth-conditions. It does not need to appeal to God, which the historical Berkeley would not have liked but which many contemporary idealists will welcome.56 And it does all this simply by extending quasi-realist moves of the kind I described in Section 3 to anything else that materialists may say to spell out what they take to be missing from idealism.

Does super-quasi-materialism also avoid scepticism about ordinary objects, as Berkeley takes his idealism to do? Since super-quasi-materialism incorporates (O3), the view itself avoids such scepticism: if super-quasi-materialism is true, we can have knowledge about ordinary objects by having certain ideas in combination with an implicit grasp of relevant linguistic rules. But trying to endorse the view will make us believe materialism, which Berkeley argues leads to scepticism. Super-quasi-materialism therefore avoids scepticism about ordinary objects while at the same time explaining why it is easy for us to have sceptical thoughts. This arguably gets it exactly right.

5. What should defenders of super-quasi-materialism believe?

Super-quasi-materialism may also seem to have a disadvantage that outweighs all of its advantages: it is a view that we cannot believe.

But just as a view can be true if we do not believe it, a view can also be true if we cannot believe it. And just as a view can be true if we cannot believe it, a view can also be true if trying to endorse it makes us believe a different view that we take to be false. We therefore should not regard our inability to believe super-quasi-materialism as a disadvantage of this view, at least not in the sense that it makes the view less likely to be true. This applies to super-quasi-materialism just as much as it does to standard Berkeleyan idealism, which is also often regarded as hard or even impossible to believe.57

You may think that our inability to believe super-quasi-materialism makes it impossible for idealists to defend the view: you may think that if they take the arguments for super-quasi-materialism seriously, they will try to endorse the view, they will thereby become materialists, and they will then come to reject the arguments for the view. But I think idealists can instead defend super-quasi-materialism by being in a dynamic state of mind. They can start out as standard Berkeleyan idealists. They can then become quasi-materialists: they can endorse (4) to (6) while taking these claims to express (4*) to (6*). By taking seriously what materialists take to be missing from quasi-materialism, they can then move towards super-quasi-materialism: they can first endorse (7) while taking it to express (7*), they can then endorse (8) while taking it to express (8*), and so on. At the end of this process they will fleetingly endorse (10) and (12) and become materialists. But they can then immediately remember Berkeley’s objections to materialism, which will move them back towards standard Berkeleyan idealism. And from there they can start moving towards super-quasi-materialism again.58

Of course, idealists may not want to defend super-quasi-materialism. But I think my arguments can then nevertheless help them. For it is open to idealists to agree with super-quasi-materialism that (4) to (9) and (13) express non-cognitive attitudes. If so, the negations of these claims must also express such attitudes: for example, the claim that

  • (~4)

    If we had thought that different linguistic rules are correct, different ordinary objects would have existed

may then express

  • (~4*)

    Approval of thinking-that-different-ordinary-objects-exist-in-circumstances-in-which-we-approve-of-different-linguistic-rules.

and so on for (∼5) to (∼9) and (~13). But instead of endorsing (4) to (9) and (13), idealists can reject these claims and endorse (∼4) to (∼9) and (~13) instead. This enables them to say that the difference between them and materialists is merely a difference in non-cognitive attitudes: whereas idealists have attitudes such as (∼4*) to (∼9*) and (~13*), materialists have attitudes such as (4*) to (9*) and (13*). And idealists can then say that this difference is too shallow to support materialists’ rejection of Berkeley’s account of ordinary objects.

It is hard to wrap our minds around a philosophical view that we cannot believe. Fully understanding a view may require being able to see what the world looks like to those who hold the view: as Winkler says when discussing Berkeley, we would like to have “the understanding that comes when we comprehend a belief by seeing how we ourselves might be brought to hold it”, since “understanding a philosophical view is not so different from being convinced by it”.59 Our inability to believe super-quasi-materialism may therefore make the view hard to fully understand. But that would not show that the view is false. All it would show is that we should try to come close to understanding the view by being in the dynamic state of mind I have described.

6. Conclusion

I have argued that Berkeley can undermine materialists’ rejection of his account of ordinary objects by making quasi-realist moves. I have also argued that when these moves are taken far enough, they result in a version of idealism that we cannot believe, since trying to endorse this version will make us believe materialism. But I have argued that this version of idealism has several advantages over standard Berkeleyan idealism and may therefore be true. Just as Berkeley can be seen as a precursor to Wittgenstein, I think super-quasi-materialism can be seen as a descendant of Wittgenstein’s remark that “idealism leads to realism if it is strictly thought out”.60

Acknowledgements

For helpful comments on previous versions of this paper I am grateful to Han Thomas Adriaenssen, Laura Georgescu, Lodi Nauta, César Reigosa Soler, several anonymous referees, and audiences at Harvard University, the University of Nottingham, the University of Reading, and Trinity College Dublin. I am especially grateful to Kenny Pearce for his extensive and very helpful comments.

References

Atherton, Margaret (1995). Berkeley without God. In Robert G. Muehlmann (Ed.), Berkeley’s Metaphysics: Structural, Interpretive, and Critical Essays (231–248). Pennsylvania State University Press.

Barker, Stephen (2020). Global expressivism. In Ricki Bliss and J. T. M. Miller (Eds.), The Routledge Handbook of Metametaphysics (270–283). Routledge. doi:10.4324/9781315112596-21

Belfrage, Bertil (1986). Berkeley’s theory of emotive meaning (1708). History of European Ideas, 7(6), 643–649. doi:10.1016/0191-6599(86)90014-8

Bennett, Jonathan (1971). Locke, Berkeley, Hume: Central Themes. Oxford University Press.

Berkeley, George (1709). An Essay Towards a New Theory of Vision. In Desmond Clarke (Ed.), Berkeley: Philosophical Writings. Cambridge University Press, 2012. Cited as NTV with paragraph numbers.

Berkeley, George (1710). A Treatise Concerning the Principles of Human Knowledge. In Desmond Clarke (Ed.), Berkeley: Philosophical Writings. Cambridge University Press, 2012. Cited as PHK with paragraph numbers.

Berkeley, George (1713). Three Dialogues Between Hylas and Philonous. In Desmond Clarke (Ed.), Berkeley: Philosophical Writings. Cambridge University Press, 2012. Cited as DHP with marginal page numbers from the Luce and Jessop edition.

Berkeley, George (1987). George Berkeley’s Manuscript Introduction. Edited by Bertil Belfrage. Doxa. Cited as MI with paragraph numbers.

Blackburn, Simon (1984). Spreading the Word: Groundings in the Philosophy of Language. Oxford University Press.

Blackburn, Simon (1993). Essays in Quasi-Realism. Oxford University Press.

Blackburn, Simon (1998). Ruling Passions: A Theory of Practical Reasoning. Oxford University Press.

Boswell, James (1791). Boswell’s Life of Johnson, volume 1. Oxford University Press, 1904.

Bracken, Harry M. (1965). The Early Reception of Berkeley’s Immaterialism 1710–1733 (revised edition). Martinus Nijhoff.

Campbell, John, and Quassim Cassam (2014). Berkeley’s Puzzle: What Does Experience Teach Us? Oxford University Press.

Campbell, Keith (1990). Abstract Particulars. Blackwell.

Dancy, Jonathan (1987). Berkeley: An Introduction. Blackwell.

Davidson, Donald (1974). On the very idea of a conceptual scheme. In Donald Davidson, Inquiries into Truth and Interpretation. Oxford University Press, 2001.

Davidson, Donald (1988). The myth of the subjective. In Donald Davidson, Inquiries into Truth and Interpretation. Oxford University Press, 2001.

Dicker, Georges (2011). Berkeley’s Idealism: A Critical Examination. Oxford University Press.

Fields, Keota (2018). Berkeley’s semiotic idealism. In Stefan Storrie (Ed.), Berkeley’s Three Dialogues: New Essays (61–83). Oxford University Press.

Flew, Anthony (1974). Was Berkeley a precursor of Wittgenstein? In William B. Todd (Ed.), Hume and the Enlightenment: Essays Presented to Ernest Campbell Mossner (153–163). Edinburgh University Press.

Friedman, Lesley (2003). Pragmatism: the unformulated method of Bishop Berkeley. Journal of the History of Philosophy, 41(1), 81–96. doi:10.1353/hph.2002.0096

Gert, Joshua (2023). Neopragmatist semantics. Philosophy and Phenomenological Research, 106(1), 107–135. doi:10.1111/phpr.12842

Gibbard, Allan (2003). Thinking How to Live. Harvard University Press.

Hight, Marc A. (2007). Why my chair is not merely a congeries: Berkeley and the single-idea thesis. In Stephen H. Daniel (Ed.), Reexamining Berkeley’s Philosophy (82–107). University of Toronto Press. doi:10.3138/9781442684751

Horgan, Terry, and Mark Timmons (2006). Cognitivist expressivism. In Terry Horgan and Mark Timmons (Eds.), Metaethics After Moore (255–298). Oxford University Press.

Hume, David (1748). An Enquiry Concerning Human Understanding. Oxford University Press, 2000.

Kail, P. J. E. (2007). Berkeley, the ends of language, and the Principles of Human Knowledge. Proceedings of the Aristotelian Society, 107(1), 265–278. doi:10.1111/j.1467-9264.2007.00221.x

Lenman, James (2020). Ethics and perception: two kinds of quasi-realism. In Christopher Cowie and Rach Cosker-Rowland (Eds.), Companions in Guilt Arguments in Metaethics. Routledge. doi:10.4324/9780429454677

Mill, John Stuart (1871). Berkeley’s life and writings. In J. M. Robson (Ed.), Collected Works of John Stuart Mill, volume 11. Routledge & Kegan Paul, 1978.

Pappas, George S. (2000). Berkeley’s Thought. Cornell University Press.

Pearce, Kenneth L. (2017a). Language and the Structure of Berkeley’s World. Oxford University Press.

Pearce, Kenneth L. (2017b). Berkeley on unperceived objects and the publicity of language. History of Philosophy Quarterly, 34(3), 231–250.

Pitcher, George (1977). Berkeley. Routledge & Kegan Paul.

Price, Huw (2013). Expressivism, Pragmatism and Representationalism. Cambridge University Press.

Reid, Thomas (1764). An Inquiry into the Human Mind on the Principles of Common Sense. Edinburgh University Press, 1997.

Rickless, Samuel C. (2013). Berkeley’s Argument for Idealism. Oxford University Press.

Ridge, Michael (2014). Impassioned Belief. Oxford University Press.

Rodriguez-Pereyra, Gonzalo (2002). Resemblance Nominalism: A Solution to the Problem of Universals. Oxford University Press.

Schroeder, Mark (2010). How to be an expressivist about truth. In Cory D. Wright and Nikolaj Pedersen (Eds.), New Waves in Truth (282–298). Palgrave Macmillan.

Schroeder, Mark (2015). Higher-order attitudes, Frege’s abyss, and the truth in propositions. In Robert N. Johnson and Michael Smith (Eds.), Passions and Projections: Themes from the Philosophy of Simon Blackburn (228–245). Oxford University Press.

Sinclair, Neil (2021). Practical Expressivism. Oxford University Press.

Stoneham, Tom (2002). Berkeley’s World: An Examination of the Three Dialogues. Oxford University Press.

Streumer, Bart (2017). Unbelievable Errors: An Error Theory About All Normative Judgements. Oxford University Press.

Streumer, Bart (2024). Superspreading the word. Noûs, 58(4), 927–947. doi:10.1111/nous.12484

Streumer, Bart (2025). Quasi-realism for realists. Philosophers’ Imprint, 25(10), 1–17. doi:10.3998/phimp.4728

Streumer, Bart (2026). How to be a self-effacing consequentalizing quasi-realist. Ethics, 136(4), 843–873. doi:10.1086/741040

Szałek, Piotr K. (2019). Berkeley, expressivism, and pragmatism. Forum Philosophicum, 24(2), 435–456. doi:10.35765/forphil.2019.2402.18

Szałek, Piotr K. (2020). The early modern origins of pragmatism. Disputatio, 12(59), 433–456. doi:10.2478/disp-2020-0020

Tipton, I. C. (1974). Berkeley: The Philosophy of Immaterialism. Methuen.

Winkler, Kenneth P. (1989). Berkeley: An Interpretation. Clarendon Press.

Wittgenstein, Ludwig (1961). Notebooks 1914–1916. Edited by G. H. von Wright and G. E. M. Anscombe. Translated by G. E. M. Anscombe. Harper & Brothers.

Yetter-Chappell, Helen (2017). Idealism without God. In Kenneth L. Pearce and Tyron Goldschmidt (Eds.), Idealism: New Essays in Metaphysics (66–81). Oxford University Press.

Yetter-Chappell, Helen (2025). The View From Everywhere: Realist Idealism Without God. Oxford University Press.

Notes

  1. Hume 1748, §12.15 n. 32. Many others make similar claims: for example, Berkeley’s contemporary Father Tournemine writes in 1718 that idealists “will persuade on one, they will not even persuade themselves” (see Bracken 1965, p. 27 and pp. 109-10), and Reid 1764, p. 20, writes that Berkeley “hath proved by unanswerable arguments what no man in his senses can believe”. Present-day philosophers make such claims as well: for example, John Campbell writes that the “view that we can think only in terms of mind-dependent objects is simply impossible to believe” (Campbell and Cassam 2014, p. 29). They are all echoing Hylas’ remark to Philonous that “your arguments seem in themselves unanswerable, but they have not so great an effect on me as to produce that entire conviction, that hearty acquiescence which attends demonstration” (DHP, p. 223). ⮭
  2. Pitcher 1977, p. 4. ⮭
  3. Of course, Berkeley’s defence of idealism is much broader than just his account of ordinary objects. For discussion, see, among many others, Bennett 1971, Tipton 1974, Pitcher 1977, Dancy 1987, Winkler 1989, Pappas 2000, Stoneham 2002, Dicker 2011, Rickless 2013, and Pearce 2017a. I focus on Berkeley’s account of ordinary objects because I think this part of his defence can be strengthened with quasi-realist moves. ⮭
  4. Boswell 1791, p. 315. Johnson’s answer so obviously ignores what Berkeley says about ordinary objects that it is usually regarded as a joke. ⮭
  5. Those who interpret Berkeley in this way include Bennett 1971, Dicker 2011, Rickless 2013, and many others. There are also more complicated versions of (O1): for example, Hight 2007 argues that Berkeley takes ordinary objects to be ideas of the imagination that signify collections of sensory ideas, and Fields 2018 argues that Berkeley takes ordinary objects to be composites that have collections of divine ideas as essential parts and ideas of the imagination and sensory ideas as contingent parts. ⮭
  6. DHP, p. 249; see also p. 175, and PHK §1 and §37; and PHK §4, §33, §§38-9, where, as Winkler 1989, p. 194, suggests, Berkeley seems to treat a collection of ideas as a single idea. ⮭
  7. See PHK §49. I say ‘certain visual and tactile ideas’ because these ideas are particulars for which we do not have specific words. ⮭
  8. See PHK §§29-33, §§56-57, §90. ⮭
  9. See PHK §6, §48, §90. ⮭
  10. See Pitcher 1979, pp. 171-2, 179, Winkler 1989, pp. 216-20, and Stoneham 2002, pp. 168-72, 246-7. Defenders of (O1) could also appeal to what Berkeley says about regularities in our flow of ideas; I will return to this in Section 2. ⮭
  11. See, for example, PHK §3-4, §6, §45. Of course, Berkeley is usually read as trying to avoid (M2). He can also be read as not caring much about the issue: see Bennett 1971, p. 171, and Pitcher 1977, pp. 174-5. ⮭
  12. For further objections to (O1), see Bennett 1971, pp. 152-64, Dicker 2011, pp. 245-51; for further objections to versions of (O1) that appeal to divine ideas, see Foster 1982, p. 31, and Pearce 2017a, p. 106, and 2017b, p. 233-4, 240-5. ⮭
  13. See Winkler 1989, pp. 191-237. Winkler’s interpretation also involves two further claims: that these conditionals are made true by God’s intentions, and that in having these intentions God perceives these ideas, which makes (O2) incorporate a core element of (O1). Bennett 1971, pp. 135-64, puts forward (O2) as what Berkeley should say rather than as what he actually says, and Mill 1871, pp. 460-1, also suggests this; see also Foster 1982, pp. 22-6. ⮭
  14. PHK Section 3; see also PHK §58 and DHP p. 253. Of course, what follows after ‘or’ in this passage suggests (O1) rather than (O2). ⮭
  15. See Pearce 2017b, pp. 234-5, 239-40, and also 2017a, pp. 105-6. ⮭
  16. You may think that this Cartesian demon is irrelevant to Berkeley. But he explicitly says that it is “a jest … for a philosopher to question the existence of sensible things, till he has it proved to him from the veracity of God” (DHP, p. 230); as Pearce 2017b, p. 233, points out, this means that Berkeley has to “show that we can gain knowledge of the natural world without assuming the veracity of God – that is, he must show that the [Cartesian] meditator has a winning strategy”. See also Pearce 2017a, p. 105. ⮭
  17. Winkler 1989, p. 202. ⮭
  18. See Pearce 2017a, in particular pp. 102-4. ⮭
  19. PHK §1 (I have replaced ‘of these’ with ‘ideas’); see also DHP p. 245. ⮭
  20. As Pearce 2017a, p. 103, puts it, “the ‘colour, taste, smell, figure and consistence’ are ‘accounted one distinct thing’, and that thing is called ‘apple’”. PHK Section 1 continues that “[o]ther collections of ideas constitute a stone, a tree, a book, and the like sensible things”, which suggests (O1) rather than (O3). ⮭
  21. Pearce 2017a, p. 97, says that (O3) treats ordinary objects as quasi-entities, which are entities that “depend for their existence on our linguistic conventions”. I take this to be equivalent to treating them as semantic shadows. Tipton 1974, p. 210, somewhat similarly suggests that according to Berkeley ordinary objects “are fictions”, which are “constructed by us out of sense data”, and Pitcher 1977, pp. 160-2, suggests that Berkeley should regard ordinary objects as “convenient fictions” but that this is not what he actually does. ⮭
  22. Pearce 2017a, p. 104 (I have replaced ‘apple’ with ‘table’). See also Atherton 1995, p. 244. ⮭
  23. We can do this by taking these propositions to be what Gert 2023, p. 112, calls ‘lightweight propositions’, and by taking these functions to be conventions, as Berkeley takes linguistic rules to be (see DHP p. 216). ⮭
  24. As Winkler 1989, p. 195, says, (O1) “owes much of its appeal to the ease with which statements of [it] can be drawn from his writings”. ⮭
  25. One example is the version of (O1) defended by Fields 2018. Another example would be a version of (O2) that takes the relevant conditionals to hold in virtue of linguistic rules, which would make the truth of these conditionals independent of future circumstances (as Pearce 2017b, p. 240, suggests). ⮭
  26. Stoneham 2002 and Pearce 2017a and 2017b suggest something like this when they call these rules ‘conventions’, and Berkeley himself suggests it when he writes that “[c]ommon custom is the standard of propriety in language” (DHP p. 216). ⮭
  27. You may think that Berkeley can reject (M3) by rejecting what Davidson calls the ‘scheme-content dualism’: the assumption that there is an uninterpreted reality outside our conceptual scheme (see Davidson 1974 and 1988). But both Berkeley and his materialist opponents are empiricists who presuppose this dualism; the difference between them is that whereas materialists take this uninterpreted reality to consist of matter, Berkeley takes it to consist of ideas (see also Davidson 1974, pp. 191-2). Neither Berkeley nor his materialist opponents can therefore reject this dualism without fundamentally changing their view. For as Davidson notes, if we give up the scheme-content dualism “it is not clear that there is anything distinctive left to call empiricism” (1974, p. 189). ⮭
  28. You may think that Berkeley can instead take our environment to consist of minds and ideas. But then the ordinary objects that our environment consists of must be collections of ideas, which is not in keeping with the role that (O3) takes ideas to play: (O3) takes ideas to serve as inputs that together with linguistic rules make certain sentences true, and takes ordinary objects to be semantic shadows of these sentences. ⮭
  29. DHP p. 245; see also PHK §§30-32, §§43-44, §105, and NTV §17, §21, §26. Alternatively, you may think that Berkeley can avoid (M3) by saying that a linguistic rule is correct if and only if it yields true sentences. But since (O3) takes sentences to be made true by linguistic rules in combination with the ideas they take as inputs, (O3) leaves no room for linguistic rules to yield false sentences. This means that if we combine (O3) with the claim that a linguistic rule is correct if and only if this rule yields true sentences, we leave no room for linguistic rules to be incorrect. ⮭
  30. He writes that “it is a universally received maxim, that ‘everything which exists, is particular’” (DHP p. 192) and that the existence of universals “with reason is exploded as nonsensical and absurd” (MI §34). See also PHK introduction §10-16, §24. ⮭
  31. NTV §109; see also NTV §64, PHK §12, and DHP pp. 245-9. The role of our subjective propensities is rightly stressed by Stoneham 2002, pp. 89-90, 99-100, 240, 264, 269-70. ⮭
  32. As Stoneham 2002, p. 299, writes, this makes Berkeley’s idealism “less staunchly realistic about the empirical world of tables and trees than [he] would have liked”. ⮭
  33. See PHK introduction §§6-25, PHK §5, §§10-11, §17, §68, §74, §81, §§97-100, §143, and DHP pp. 193-4. ⮭
  34. For trope theory, see, for example, Campbell 1990; for resemblance nominalism, see Rodriguez-Pereyra 2002. ⮭
  35. Moreover, trope theorists usually also take tropes themselves to be abstract (Campbell even calls them ‘abstract particulars’) and take properties to be sets, which are clearly abstract. Perhaps Berkeley can be regarded as a resemblance nominalist, as Pearce 2017a, p. 32, suggests, but only if we take his notion of resemblance to be subjective: if we take it to obtain in virtue of our subjective propensities to group ideas together. See also Stoneham 2002, pp. 100, 270-1. ⮭
  36. Pearce 2017a, p. 185, suggests something along these lines when he says that “some groupings fit better than others, and this is due to the linguistic rules that God has implemented” and that we must “put them (back) together as part of our interpretive activity”; see also Pearce 2017b, pp. 237-40, and 2017a, pp. 139-204. ⮭
  37. For example, he can take them to be plans (see Gibbard 2003) or non-representational beliefs (see Horgan and Timmons 2006). ⮭
  38. PHK introduction §20. Sinclair 2021, p. 1, takes this passage to hint at expressivism; more extensive hints can be found in MI §§36-43. But the historical Berkeley clearly does not have a fully worked out view about normativity. For discussion, see Belfrage 1986, Kail 2007, Pearce 2017a, pp. 143-9, and Szałek 2019. ⮭
  39. For discussion of these pragmatist tendencies, see, among others, Friedman 2003 and Szałek 2019 and 2020. Expressivism is not the only way to develop these tendencies; I focus on expressivism because I think it enables Berkeley to avoid (M5), as I will explain below. ⮭
  40. As is noted by Kail 2007, pp. 274-7, and Lenman 2020, though they also note differences. ⮭
  41. This response could also be taken further by endorsing a global version of expressivism along the lines proposed by, among others, Price 2013, Barker 2020, and Gert 2023: for example, Berkeley could endorse Gert’s expressivist treatment of existentially quantified descriptive sentences like ‘There is a stone’ (see Gert 2023, pp. 131-2). But global expressivism is further removed from the historical Berkeley than the account I will propose. ⮭
  42. I say ‘may’ because expressivists do not have to take (1) to (3) to express higher-order attitudes like (1*) to (3*). A higher-order attitude account of (1) was first proposed by Blackburn 1984, who took (1) to express disapproval of alternative versions of ourselves who do not disapprove of this thing; see also Blackburn 1993 and 1998. Blackburn’s higher-order attitude account is developed in a promising way by Schroeder 2015, focusing on Blackburn’s early solution to the Frege-Geach problem. ⮭
  43. I take minimalism about truth to be the view that that nature of truth is exhausted by the claim that ‘p’ is true if and only if p, and I take this view to entail that the sentence ‘“p” is true’ expresses the same mental state as the sentence ‘p’. Minimalism about properties can also be taken to be the view that the nature of properties is exhausted by the claim that X has the property of being F if and only if X is F. ⮭
  44. See Blackburn 1993 and 1998 and Gibbard 2003. ⮭
  45. For example, they can endorse the expressivist account of truth proposed by Schroeder 2010 or the ecumenically expressivist account of truth proposed by Ridge 2014. ⮭
  46. You may think that (6) conflicts with the fact that we speak different languages and with the possibility of verbal disputes within a single language. But as I said above, ‘linguistic rule’ is merely a label for a function that takes us from ideas to true sentences. ⮭
  47. See Kail 2007, pp. 274-7, and Lenman 2020. ⮭
  48. Of course, expressivists have an even greater need to solve the Frege-Geach problem. If Berkeley endorses expressivism about the correctness of linguistic rules, the same applies to him. But Berkeley may have an escape route: he may be able to avoid this problem by endorsing relativism rather than expressivism about the correctness of linguistic rules. For I argue in Streumer 2025 that relativists can make the same quasi-realist moves as expressivists, including the moves I will discuss below. ⮭
  49. You may think that you and the other person only really disagree about whether there is a table in the room if you both follow the same relevant linguistic rules. But materialists will take there to be an alarming symmetry between you and the other person even if you cannot really be said to disagree. ⮭
  50. See Streumer 2024, 2025, and 2026. ⮭
  51. Materialists may also make a claim along the lines of (10) about the existence of universals. But Berkeley can then respond in the same way, by treating universals the way (O3) treats ordinary objects and the way (11) treats matter. ⮭
  52. DHP, p. 261. In PHK §37, he similarly writes that “if the word ‘substance’ be taken in the vulgar sense, for a combination of sensible qualities, … this we cannot be accused of taking away”. ⮭
  53. See Mill 1871, pp. 459-60, though Mill does not put the point in terms of linguistic rules. ⮭
  54. Trying to endorse super-quasi-realism will turn expressivists into robust realists via a similar dynamic process; see Streumer 2024, pp. 935-8. ⮭
  55. I here use the term ‘belief’ to mean full and explicit belief, which I take to be the kind of belief we would like to have as a result of philosophical theorizing, and I assume that we cannot have full and explicit beliefs that we fully and explicitly believe to be inconsistent with each other. I discuss this in more detail in Streumer 2017, pp. 129-38. There may be weaker senses of the term ‘belief’ in which we can believe super-quasi-materialism: for example, by having partial or implicit beliefs that we partly or implicitly believe to be inconsistent with each other, or by being in the dynamic state of mind that I will describe in Section 5. ⮭
  56. See Atherton 1995, Stoneham 2002, pp. 128-33, and Yetter-Chappell 2017 and 2025. ⮭
  57. See the references in note 1 above. ⮭
  58. I argue in Streumer 2024, p. 941, that expressivists can defend super-quasi-realism in a similar way. ⮭
  59. Winkler 1989, p. vii. ⮭
  60. Wittgenstein 1961, p. 85. For the idea that Berkeley is a precursor to Wittgenstein, see Flew 1974. ⮭