1. Introduction
Philosophers have failed to notice a rich class of objects. Phenomenal constructions are concrete, sensible entities whose reality consists in a certain sort of experiential coherence. There are many types of phenomenal construction. Some are public (i.e., presented to distinct subjects); others are private. Some exist only when perceived; others exist unperceived. Some have objective character; others do not.
In particular, I will defend the Leibnizian thesis that ordinary objects (e.g., trees) are themselves phenomenally constructed.1 This result has two important implications.
First, it vindicates the naive realist’s view that veridical perceptual experience involves direct acquaintance with ordinary objects. Against this view, it is often objected that the physical world is very different from how the world naively seems in experience. But this objection has no purchase if ordinary objects are not physical, but are instead phenomenally constructed. Second, and optionally, one might use this result as a step towards defending idealism: the thesis that all physical truths depend on phenomenal truths. I discuss these issues in section 5.
Section 2 presents the intuitive idea of phenomenal construction. Section 3 adapts arguments from metaontological deflationists to show that phenomenal constructions exist. (While I offer some arguments in favor of this metaontology, it is outside the scope of this paper to provide a full defense. Thus, one may view my conclusion as conditional: assuming deflationism, phenomenal constructions exist.)
2. An intuitive example
We refer to rainbows as real, public objects.2 We point to them and ascribe colors to them. Indeed, to some extent, we even allow that they exist unperceived. (“We didn’t even notice the rainbow near the waterfall!”)
One might say that such talk is not metaphysically serious (cf. Alspector-Kelly (2004, 335ff)). Rainbows are hallucinations, not real objects. In an airplane, one does not find “physical, multi-colored arches across the sky” (ibid., 335).
But why not say that rainbows exist? Rainbows are not material objects: unlike tables, they cannot be identified with arrangements of particles. But what of it? These are items from a different ontological category. To say that rainbows cannot exist without material composition is like saying that numbers cannot exist without material composition. But perhaps rainbows have different existence conditions. We might say, very roughly, that rainbows exist just in case subjects in certain conditions have or would have appropriately-cohering rainbow experiences. (If, when doing fundamental ontology, one wants to deny that rainbows really exist, one may do so. Still, there is a perfectly familiar sense in which rainbows exist.)
One might object: “Rainbows are not really public: they have different locations depending on the observer. So, if anything, there are different rainbows for each subject” (ibid., fn. 19). But why not say that rainbows are public? They are not public as material objects are: they lack unique physical locations. But perhaps rainbows are public in a different way. Very roughly, their publicness consists in different subjects having appropriately-cohering rainbow experiences.
One might object: “Rainbows do not really exist unperceived.” But, again, why not say that rainbows exist unperceived in a different way? Rainbows exist unperceived just in case they are the objects of appropriately-cohering counterfactual experiences but not the objects of any actual experience.
One might critique the above analysis. No matter—my purpose is not to defend rainbows. In fact, there are differences between rainbows and the constructions discussed below.3 I offer this example simply to make phenomenal construction more intuitive.
2.1 A modest approach
As an idealist, Leibniz thought that the whole world was phenomenally constructed. My conclusion is more modest: phenomenal constructions exist, even if other things do too. In particular, my arguments are compatible with the existence of an independent physical reality.
Indeed, I will assume that there is such a physical reality. This assumption allows me to set aside certain traditional objections to phenomenal construction. For example, one might worry that, with phenomenal construction, we cannot then explain the coherence of phenomenal experience. But I will suppose that experiences cohere because they are supported by an underlying physical reality.
A caveat: if ordinary objects are themselves phenomenally constructed (see section 4), then this underlying reality will not include ordinary objects. This suggests a kind of Kantian distinction: the phenomenal constructions presented in experience vs. the reality grounding that experience. We might view the latter as the reality described by fundamental physics.4
2.2 Counterfactual experiences
Phenomenal constructions require cohering counterfactual experiences. That is, they require conditional truths of the form ‘If subject were to experience , then would experience ’. As in 2.1, I will suppose that these conditionals are ultimately grounded by underlying physical truths (plus psychophysical laws).
This grounding is causal, not metaphysical. Compare: dualists allow that physical states cause phenomenal states, but deny that we can analyze phenomenal states in physical terms.5 Similarly, should not be analyzed in physical terms; otherwise, it would be unsuitable for phenomenal construction.
Here, we might say that physical truths cause to have the potential to experience on the condition of experiencing . We can then analyze as: experiences in those possible worlds in which experiences that are most similar to ours with respect to phenomenal potentials.6
3. The defense
To test whether phenomenal constructions exist, suppose that a demon causes our experiences. Most philosophers deny that concrete, sensible objects exist in this scenario. But I will argue that there are many such entities. Because truths about these phenomenal constructions supervene on phenomenal truths, these items also exist in the actual world. This section establishes the following claims:
In a demon scenario, subjects can successfully attend to, refer to, and entertain true thoughts about various items apparently presented in experience.
These items exist.
They are concrete and have sensuous character.
-
They have a range of existence and identity conditions:
(a) Some are public; others are private.
(b) Some exist only when perceived; others exist unperceived.
(c) Some are 3D; others are 2D.
(d) Some have objective character (i.e., character that persists despite changing appearances); others do not.
These items also exist in the actual world where our experiences, under present assumptions, have physical causes.
At present, I set aside ordinary nonveridical experiences: illusions that even unknowing demon victims could potentially recognize (see 4.5 for discussion).
3.1 Claim (I): Attention, reference, truth
Right now, I seem to be directly presented with a blue book. But this, we are supposing, is a demon scenario. On the standard view, this implies that I am not presented with a book. Then what is the metaphysical status of the presented item?
At the very least, I can visually attend to, refer to, and entertain true thoughts about it. This is shown by demonstration. Right now, I attend to the presented item. I now say “That,” thereby referring to it. I now think: “That is presented to me,” thereby entertaining a true thought.
Of course, I am not really in a demon scenario. So perhaps I have merely shown that I can attend to (refer to, etc.) a book. But let’s temporarily suppose that, in fact, I am currently hallucinating. I claim that, in this situation, I can still attend to the presented item. If this is possible during an actual hallucination, then it is also possible in the demon scenario where I perform the same mental operations. (I consider objections below).
3.2 Claim (II): Existence
From the truth in 3.1, it is trivial to infer that the presented item exists:
That is presented to me. (true thought from 3.1)
There exists something—namely, the presented item—that is presented to me. (existential generalization)
This conclusion also follows from the fact of successful reference:
The demonstrative ‘that (presented item)’ successfully refers. (see 3.1)
If ‘that X’ successfully refers, then the X exists. (a priori truth)
Therefore: the presented item exists.
These arguments do not show that a book exists; they only show that the presented item exists, whatever it turns out to be. But most philosophers deny that any presented item exists, book or otherwise. Such philosophers will reject claim (I) or else reject the inference from (I) to (II). I consider these objections in turn.
3.2.1 Objection 1: the alleged priority of existence
It may seem that I have put things backwards. To attend to X, X must exist. Thus, to establish (I), I must first establish (II). And, regardless, (I) begs the question against those who deny that the presented item exists, and who thereby deny that we can attend to (refer to, etc.) this item.
In response, how can I doubt that I am, right now, successfully thinking about the presented item? I think that it looks darker than its surrounding background. Will the objector deny that this is a genuine thought, or deny that I am making genuine cognitive contact with something? Will they deny that this thought is true?
Of course, many thoughts fail due to empty names. Let ‘Vulcan’ refer to the planet nearer to the Sun than Mercury. Then the thought “Vulcan is hot” fails. Still, there are clear differences between the cases. First, the Vulcan thought seems defective to any suitably-informed subject in a way that the thought “That is presented to me” does not. Second, I am not acquainted with Vulcan as I am acquainted with the presented item. Accordingly, the possibility of presupposition failure does not extend to the latter case.
Might there still be presupposition failure? One might cite hallucinations: experiences that merely seem to present ordinary objects. But, at present, I do not presuppose that the presented item is an ordinary object.7
There is a third contrast. The thought “Vulcan exists” carries a certain risk: we can imagine evidence (e.g., new astronomical observations) that would lead subjects to abandon this belief. In contrast, the thought “That is presented to me” is “conditionally” risk-free. By this I mean: so long as all actual/counterfactual experiences (for both myself and other subjects) are as we suppose them to be, there is no ordinary8 empirical evidence that would lead me to reject this thought.9
In summary: in many cases, we must antecedently determine whether X exists in order to know whether claims invoking X are true. But this is not required in the present case, where we believe that the item exists because we are acquainted with it and because we manifestly succeed in making cognitive contact with it.10
3.2.2 Objection 2: truth/reference without existence
I now consider philosophers who accept (I) but nonetheless maintain that the presented item does not exist.
According to representationalists, perceptual states are representational states. Representational states can be about nonexistent objects. Accordingly, many representationalists (e.g., Loar (2003, 84)) claim that hallucinated items are “merely intentional”: they do not exist but are merely represented to exist.
There are three stances representationalists might take towards (I) and (II). First, they could deny (I). For example, some representationalists claim that hallucinations involve relations to general contents (e.g., There is a sphere) or gappy contents (e.g., ___ is blue).11 Presumably, such philosophers will deny that we can entertain true singular thoughts about concrete objects in the demon scenario. (I) confutes such views.
Second, representationalists might accept both (I) and (II). According to some philosophers, “merely intentional” entities must exist if they are to perform explanatory work. These philosophers identify such items with uninstantiated property complexes or other abstracta.12 These “ontologically serious” representationalists can accept the discussion thus far.
Finally, representationalists might accept (I) but deny (II). On this view, intentional entities with no ontological status can still fill the cognitive roles mentioned in (I). This may seem plausible: can I not refer to Sherlock Holmes? But the analogy fails. Contrast the first argument from 3.2 with the following:
1*. Holmes is clever.
2*. There exists someone—namely, Holmes—who is clever.
Unlike the earlier argument, the Holmes argument is unpersuasive. (1*) is only true when assuming a fictive operator. But then (1*) does not entail (2*), which is understood to involve no such operator. This suggests that merely intentional objects, like Holmes, are poor models for the current case.
3.2.3 Secondary argument: deflationary metaontology
Suppose that there are particles “arranged chairwise.” Is there also a further entity—a chair—composed of them? According to deflationists, we can resolve this question through conceptual reflection. The same goes for many ontological debates. Here, I adapt deflationary arguments to further defend (II). While it is outside the scope of this paper to defend deflationism, I aim to show how one familiar metaontology supports the inference from (I) to (II).
According to many deflationists13, we can establish X’s existence via pleonastic inferences from uncontroversial premises with no apparent commitment to Xs. For example, let “There are particles arranged chairwise” be shorthand for whatever uncontroversial empirical facts obtain when we typically judge that a chair exists. Then the existence of chairs follows trivially:
There are particles arranged chairwise.
There exists a chair. (pleonastic inference)
How is this inference trivial when it introduces new ontological commitments? According to Thomasson (2007), it is constitutive of competence with the terms ‘chair’ and ‘exist’ that one recognize the correctness of this inference.14 Accordingly, anyone denying this inference makes a conceptual mistake.
I claim that, analogously, pleonastic inference establishes (II). Here, (1′) is shorthand for whatever phenomenal truths obtain in virtue of my current experience (with properly-focused attention):
1′. I have an experience as of a book.
2′. That exists. (pleonastic inference)
In the chair case, we know that the inference is trivial because it is undeniably treated as such in everyday speech. But it is difficult to isolate if, and when, speakers use the demonstrative in (2′). This difficulty is aggravated by my refusal, thus far, to identify what I am demonstrating.
To finesse these difficulties, I stipulate that the inference from (1′) to (2′) is meaning-constitutive for the relevant demonstrative. Accordingly, anyone denying this inference makes a conceptual mistake, just like the chair case.
Or is my stipulation defective? One might complain that my demonstrative lacks determinate rules of use. I discuss this issue in 3.4. For now, suffice to say that the rules are at least sufficiently settled to secure the inference from (1′) to (2′).
3.3 Claim (III): sensuous character
The presented item exists. But what is it? Here are some possibilities.
An item filling a certain structural role: According to functionalists (e.g., Chalmers (2012, E15)), we should identify macroscopic objects with whatever items have the role of normally causing experiences of those objects. In a demon scenario, these might be the demon’s ideas of those objects. On this view, my demonstrative refers to the demon’s book-idea. I grant that it is possible to use a demonstrative thus. Still, this book-idea is not what I referred to above. Instead, I intended to refer to something directly presented in experience.
A spatial region: Alston (1999) identifies hallucinated items with spatial regions instantiating qualitative properties. Similarly, Peacocke (2008, 12) claims that hallucinations involve sensational properties instantiated in regions near the eyes. One might try to extend these ideas to the demon scenario. But this proposal fails because, ex hypothesi, there is no physical space in the present scenario.
An uninstantiated sensible profile: One might also try to extend Johnston’s (2004) proposal that hallucinations are complex qualitative universals that could be, but are not, instantiated. However, Johnston notes that uninstantiated sensible profiles cannot be an original source of de re thought about particular things (129). So (I) confutes this view.
What, then, is the presented item? We can at least say this: it is a concrete item with sensuous character. I know this because I am directly presented with it, and all along, I intended to refer to the item directly presented to me. One might naturally describe this item as blue, book-shaped, etc. But perhaps only physical entities have colors and shapes. To remain neutral, I will say that the presented item is blue* and book-shaped* (see 3.7-3.8 for discussion).
Many questions remain. Is the presented item public or private? Does it exist unperceived? To address these questions, I use a metaontological framework from Thomasson (2007).
3.4 Application and coapplication conditions
I point to a kangaroo and say “That.” Do I thereby refer to a kangaroo, or to an instantaneous time-slice, or to the kangaroo’s surface, or …? (For I could use demonstratives in all of these ways.) Causal relations do not decide between these options. Then what determines the referent?
Thomasson (2007) argues that the meaning of any nominative expression is partially constituted by its application conditions: conditions under which the expression successfully refers. For example, ‘that iceberg’, unlike ‘that kangaroo’, will successfully refer in certain arctic environments.
While application conditions determine whether a term refers, they do not establish which item the term refers to. This is because terms with different referents—e.g., ‘that kangaroo’ vs. ‘that instantaneous kangaroo time-slice’—can share application conditions. Thus, for determinate reference, terms must also have coapplication conditions: rules specifying under which conditions a second use of the term applies to the same entity (40). For example: ‘that kangaroo’, unlike ‘that instantaneous time slice’, applies to the same entity after five minutes. Competent speakers need not be able to articulate these conditions, but they manifest their understanding of these conditions in their ability to evaluate how a term correctly applies across possible situations.15
As linguistic rules, coapplication conditions (CCs) are expressed metalinguistically. But we can read off the identity conditions of their referents via disquotation into the object language. Here, Let ‘S1’ and ‘S2’ be two distinct uses of the sortal ‘S’:
| The CCs for ‘’ are met | iff | ‘That ’ and ‘That ’ co-refer |
| (def. of CCs) | ||
| iff | ‘That = that ’ is true | |
| (a priori truth) | ||
| iff | that = that | |
| (disquotation) |
One might imagine leaving it entirely open what kind of entity, if any, a demonstrative refers to. But the link between coapplication conditions and identity conditions circumscribes the possibilities for discovering the latter. Suppose that I point at a kangaroo and say “Let ‘Klut’ refer to that, whatever it turns out to be.” Then I cannot empirically16 discover whether ‘Klut’ refers to the kangaroo vs. the mereological sum of cells composing it. This is because the term ‘Klut’ would successfully refer on either such possibility. Indeed, it would have the same application conditions either way. Thus, for ‘Klut’ to determinately refer, its coapplication conditions must distinguish these options. But then one can read off Klut’s identity conditions as described above.
These remarks allow for certain empirical discoveries about identity conditions. For example, I might use ‘Klut’ to refer to that thing on the hillside, not knowing whether it is a kangaroo, or a statue, or … . I could then discover which of these Klut is. In this case, my categorial intentions are conditional on what the world is like. I use ‘Klut’ such that: ‘Klut’ refers to an animal when ‘animal’ conditions are met, an artifact when ‘artifact’ conditions are met, etc. Because these application conditions are distinct, I can discover Klut’s category by discovering which conditions the world turns out to meet. But empirical investigation could not similarly decide between categories that share application conditions.17
Coapplication conditions do not fix identity conditions in a metaphysical sense. We cannot change kangaroos’ identity conditions by changing coapplication conditions.. Rather, coapplication conditions semantically fix the fact that ‘kangaroo’ refers to kangaroos (with certain identity conditions) rather than sums of cells (with other identity conditions).
3.5 Claim (IVa): publicness
I now apply Thomasson’s framework to the original demonstrative. Is the presented item private or are multiple subjects acquainted with it? Suppose that the demon causes experiences (that are appropriately related) as of a certain book in ten subjects. Let and be token uses of demonstratives by subjects and in this set, respectively. I hereby stipulate that, for any such subject , has an experience as of a book is a coapplication condition for the demonstrative. Then . Thus, the demonstrative refers to a public object.
“But no public objects exist in the demon scenario.” In response: I already established that the presented item exists in 3.1-3.2.3. As for whether distinct subjects perceive the same presented item: this is read off of the demonstrative’s coapplication conditions, just like other questions about identity conditions. Compare: whether Klut is a kangaroo or a sum of cells is read off the coapplication conditions governing ‘Klut’ (see 3.4). I just stipulated coapplication conditions on which the object is public (though I could have done otherwise).
“But the presented item is not really public.” If tempted by this objection, one probably rejects the metaontology from 3.4 on which we can read off identity conditions from coapplication conditions. For now, I ask readers to accept this metaontology as a working hypothesis (or, at least, to accept that the presented item is public “in a deflationary sense”). I discuss this issue further in 4.3.
3.6 Claim (IVb): existence unperceived
As in 3.1, I attend to the presented item and refer to it with token demonstrative . Then I “close my eyes.” We are supposing that, in the demon scenario, all counterfactual experiences are just as we normally suppose them to be. Then the following counterfactual obtains: if I were to have the experience as of opening my eyes, then I would have an experience as of a book. Let be the counterfactual token demonstrative I might issue were I having this experience. I stipulate that the italicized counterfactual is a coapplication condition for . Thus, : I refer to something that exists even when unperceived. If I made the opposite stipulation, then I would refer to something for which esse est percipi.
3.7 Claim (IVc): spatial* identity conditions
Again, I refer to the presented item with token demonstrative . Right now, the apparent book is “closed.” But the following counterfactual obtains: if I were to have the experience as of opening the apparent book (in such-and-such way), then I would have an experience as of such-and-such opened page . Let be the counterfactual token demonstrative a subject might issue were they having this experience. I stipulate that, for all , the italicized counterfactual is a coapplication condition for . Thus, for all , .18 Accordingly, the demonstrated item is three-dimensional rather than a mere surface.
3.8 Claim (IVd): objective character and appearances
A “shadow” falls on the presented item. Accordingly, there is a sense in which it appears to have two distinct colors* (see 3.3) and a sense in which it seems uniformly colored*. Which description is correct? In other words, do color* demonstratives directed at the shadowed and nonshadowed parts refer to the same color*?
This is like asking: “Does ‘Klut’ refer to a kangaroo or a mereological sum of cells?” This depends on our choice of coapplication conditions. Similarly, the coapplication conditions for color* demonstratives settle whether the colors* are identical.
Let “I experience (non)shadowed blue*” be shorthand for the relevant phenomenal truths obtaining when attending to the relevant parts of the presented item. Let and be corresponding token color* demonstratives. I stipulate that I experience shadowed blue* and I experience nonshadowed blue* are coapplication conditions for the demonstrative. Thus : the object has a uniform color*.
On the opposite stipulation, . This does not mean that my linguistic choice changes the presented property. Rather, my choice determines which kind of color* property I refer to with the demonstrative. We might call the former an objective color* and the latter a color* appearance. Experience does not present one or the other; it presents both. Acquaintance with color* appearances no more screens off acquaintance with objective colors* than acquaintance with time-slices screens off acquaintance with kangaroos.
3.9 Claim (V): actual existence
In the demon scenario, experience presents many objects and properties. Some have two dimensions; others have three. There are private items that only exist when perceived; there are public items that exist when unperceived. We can refer to any such item by stipulating appropriate coapplication conditions.
These items are phenomenally constructed in the following, precise sense: the correctness conditions—i.e., application and coapplication conditions—for the perceptual demonstratives denoting them invoke only phenomenal truths. By contrast, these rules do not mention the reality causing our experience (i.e., the demon).19 Of course, our actual experiences are (presumably) caused by a physical reality. But because the phenomenal truths are the same, the same phenomenal constructions exist in either case.
3.10 How permissive?
As emphasized above, phenomenal constructions’ identity conditions do not metaphysically depend on our choice of coapplication conditions; rather, this choice settles the type of construction to which we refer. How permissive is this ontology? For example, if I see a bird and then a book, is there a phenomenal construction that is book-shaped* before time and bird-shaped* after?
This issue confronts deflationists in many domains. There are at least two possible responses.
First, one might adopt a permissive ontology on which, for any coherent assignment of coapplication conditions, there is a corresponding entity.20 This approach avoids the arbitrariness of any proposed restriction to a deflationist’s ontology. However, one might worry that this approach conflicts with ordinary thought and speech. For example, we would ordinarily deny the existence of the “book-birds” described above.
Second, one might adopt a Carnapian view on which different conceptual schemes carve up the world in different ways.21 For example, we do not ordinarily judge that two cups on a table compose a third object. But imagine that aliens talk about “cup-cups.” For Carnapians, this is no real dispute; the aliens simply use a different conceptual scheme. Nor is there anything problematic about our ontology excluding cup-cups (or book-birds): it is simply more useful, at least for humans, to adopt one conceptual scheme rather than the other.22
My discussion is neutral between these alternatives.23
4. Ordinary objects
On the standard view, the term ‘stone’ correctly applies when certain physical conditions obtain. For example, Thomasson (2007) discusses application conditions of the following form:
(*) The term ‘stone’ refers (and, therefore, a stone exists) iff there are particles arranged stonewise.24
By contrast, if stones are phenomenally constructed, then ‘stone’ correctly applies when certain phenomenal conditions obtain. I already demonstrated that phenomenal constructions can have many of the features of ordinary objects: objective character, mind-independence, etc. But the rules governing ordinary object terms are much more complex than the sketches from section 3. I will argue that these rules—or, at least, application conditions—have roughly the following form:
(†) The term ‘stone’ refers (and, therefore, a stone exists) iff actual and counterfactual phenomenal experiences cohere so as to collectively indicate a stone.25
For example, suppose that thirty people visit Stonehenge. Their experiences cohere so as to present various large, grey objects from various perspectives. The term ‘stone’ successfully refers because these experiences, and others, cohere in a stone-suggestive way. It follows, via disquotation, that these stones exist. Of course, many parts of the world are unperceived. This is why () also invokes counterfactual experiences.26
The next several subsections provide a more precise account.
4.1 The idealized subject
On present usage, acquaintance terms are demonstratives used to denote (or express) items directly presented in experience, such as ordinary objects, sensible properties and relations (e.g., colors and shapes), and locations and regions of space. Let be an acquaintance sentence: a sentence formed using only acquaintance terms, logical/mathematical vocabulary (including vocabulary for expressing counterfactuals), and the terms ‘cause’ and ‘disposition’. For example, acquaintance sentences include: ‘This (object) has that (color)’, ‘This (object) is thus (cube-shaped)’, ‘This (object) is identical to that (object)’, and ‘If this (object) were to move there, then it would next move there’.
Let’s say that actual phenomenal truths are truths of the form ‘Subject has phenomenal property ’. Let’s say that counterfactual phenomenal truths are truths of the form ‘If subject were to have phenomenal property , then would have phenomenal property ’.27 Let be the conjunction of all actual and counterfactual phenomenal truths. Then the following principle indirectly expresses the correctness conditions governing acquaintance terms28:
Phenomenal Construction (PC): is true iff an idealized subject judges that conditional on .29
To explain PC, I borrow an illustration from the literature on Humean laws. According to Humeans, laws of nature are merely patterns in the Humean mosaic. The Humean mosaic is the set of particular facts about the existence of physical objects, their nondispositional physical properties, and their spatiotemporal relations. It excludes facts involving laws, causation, counterfactuals, and dispositions.
To illustrate their view, some Humeans invoke an idealized physicist.30 The physicist has perfect memory, can entertain arbitrarily complex thoughts, and never commits logical/mathematical mistakes. Suppose that this physicist is magically afforded knowledge of the entire Humean base. Given this knowledge, certain regularities will most look like the natural laws to the physicist. According to certain Humeans, these regularities just are the laws.
In this illustration, the physicist is an idealized human subject, or at least someone who recognizes laws using the same implicit criteria that we do. (If they were instead some arbitrary intellectual being, then who knows what they would say about the laws?) The same assumption holds below.
Now suppose that the idealized subject is instead magically given the phenomenal base: all actual and counterfactual phenomenal truths for all subjects across history. (Here, we might imagine that the idealized subject possess a hypothetical virtual reality device that can induce any subject’s phenomenal state at any time—see Chalmers (2012, 115f).) Given this enormous body of knowledge—knowledge of —certain sentences will most look like the acquaintance truths. According to PC, these sentences just are the acquaintance truths.
For example, I see a blue book. When given , the idealized subject knows exactly what my experience is like. In addition, perfectly describes what I would experience were I to have the experience as of approaching the apparent book, as of pushing it, etc. On the basis of this and everything else in , the idealized subject would judge “That (book-shaped object) is thus (blue).” Thus, according to PC, that (book-shaped object) is thus (blue).31
Note that, because of the restriction to acquaintance sentences, PC says nothing about objects’ nonsensible properties (e.g., having charge c, being a statue). I discuss this technical issue in the appendix.
4.2 Interpreting PC
With the physicist illustration, Humeans do not claim that laws depend on anyone’s idealized judgment. This illustration is simply an intuitive, albeit indirect, way to specify which particular regularities are the laws.
This illustration is available to Humeans because they view laws as mere patterns in the mosaic. Thus, the judgments of an idealized physicist—someone who is given the mosaic and can examine its patterns—align with the actual laws. An advantage of this indirect approach is that Humeans can remain neutral on which specific patterns count as laws. These patterns just whatever patterns human physicists would, under idealization, implicitly recognize as laws.
Analogously, if objects are phenomenally constructed, then they exist when experiences cohere in a certain way. Thus, the judgments of an idealized subject—someone who is given the phenomenal truths and can examine their coherence—align with acquaintance truths about objects. This does not mean that objects are brought into existence by any ideal observer. As before, the idealized subject is just an illustration: an indirect way to specify how, on my account, truths about objects correlate with phenomenal truths. With this approach, I can remain neutral on how exactly experiences must cohere for there to be an object. It is whatever coherence that we would, under idealization, implicitly recognize as establishing the object’s existence.32
Despite this neutrality, PC is not devoid of content. The idealized subject judges acquaintance sentences solely using . This entails that the correctness conditions for acquaintance terms are purely phenomenal (and, therefore, that ordinary objects are phenomenally constructed).
4.3 The character of phenomenally-constructed ordinary objects
If PC turns out to be true, then what are ordinary objects like? This is best answered through demonstration. I see a book. This object, which I claim is phenomenally constructed, is just like this.
Of course, the mere act of saying “this” is not a full answer, for familiar reasons. First, for demonstratives to determinately refer, they require coapplication conditions establishing their referents’ ontological category (see 3.4). Second, my present experience does not present just a book; it also presents a facing surface, a private item, etc.
But while nothing in my present perspective reveals the difference between the book and its facing surface, these items are nonetheless distinct because certain other perspectives count as presentations of the one but not the other. For example, I flip the book. I now view the same book from a new perspective, but not the same facing surface. Similarly, my present experience does not distinguish the book from a private item (see 3.5). These are distinguished because my friend’s experience also presents the former but not the latter.
In general, any phenomenal construction is distinguished by the set of possible perspectives on it across the space of actual/counterfactual experiences. When we use the term ‘this (book)’, we carve the space of possible perspectives one way. When we use ‘this (surface)’ or ‘this (private item)’, we carve the space differently. In each case, we group together different sets of possible perspectives as those regarded as presenting the same item.
We group possible perspectives into these different sets by implicitly choosing different coapplication conditions for the relevant demonstratives. PC indirectly expresses these coapplication conditions (in the object language) when the idealized subject judges “Thisi is identical to thisj.”
Imagine, for example, that the idealized subject considers my present experience. They demonstrate the apparent book: “thisi.” Now they take their virtual reality device (see 4.1). They simulate the counterfactual experience as of shifting attention to another part of the book and then complete their judgment: “…is identical to thisj.” They simulate the experience as of rotating the book, as of pushing the book: “…and thism, and thisn.” They simulate experiences at other times and experiences of other subjects.
The idealized subject thus distinguishes the book from a facing surface, a private item, and so on. As for what the phenomenally-constructed book is like, there is nothing more to say: it is the item presented from all and only these perspectives.
This is how, on the present account, ordinary objects are constructed. They are not constituted by experience like statues from clay. Nor are they are composed of experiences like skeletons from bones. Nor are they “logical constructions,” where sentences invoking them are merely disguised phenomenal truths (cf. Russell (1918)). Instead, they are constructed in the sense that, by choosing correctness conditions, we group certain perspectives together as those regarded as presenting a single item. Our choice does not make the object. Rather, it settles which perspectives we regard as revealing the object: the object that we care to talk about.33
4.4 Objects as public
The idealized subject can link experiences from different subjects when proceeding through . One can visualize the set of my actual/counterfactual experiences as a intricately detailed Google Earth simulation. When this simulation loads, the idealized subject sees a narrow part of the world: the part presented by my actual experience. They have controls to “move forward,” “turn left,” etc. Accordingly, they can explore the simulation (i.e., the space of my counterfactual experiences). When finished, they have my map of the world. The simulation of your experience has a different starting point. The idealized subject explores this simulation too. Upon completion, they have your map.
Barring skeptical hypotheses, the above maps should look largely the same. The idealized subject points to the book in my map and says “Thisi.” They find the corresponding item in your map and complete the judgment “…is identical to thisj.” As discussed in 3.5, these coapplication conditions entail that the phenomenally-constructed book is public.
I now consider two objections.
“The idealized subject may say that the books are the same, but different subjects do not really perceive the same phenomenal construction.”
Compare: “Subjects at different times who ‘look at the same river’ do not really see the same entity; they merely see different water molecules related in a certain way.” This thought mistakenly assume that there is a privileged set of identity conditions—those met by molecules—that entities must satisfy if they are public. But, in fact, rivers are just a different kind of thing: what it is to see the same river at different times just is to see water molecules related in a certain way. Similarly, phenomenal constructions are just a different kind of thing: what it is for subjects to see the same phenomenal construction just is for them to have appropriately-cohering experiences. (For a more general metaontological diagnosis of this objection, see 3.4 and 3.2.3.)
“What if the Big Bang generated twin sub-universes and evolving in parallel? Any -subject’s map would be indistinguishable from a corresponding -subject’s map. Then what prevents the idealized subject from judging that -subjects and -subjects perceive the same objects?”34
To respond, I note that counterfactual phenomenal truths encompass dependence between experiences of distinct subjects (see fn. 27). For example, suppose that five -subjects look at a stone. If subject were to have the experience as of approaching the apparent stone, then the other -subjects would have different experiences. But, presumably, there is no such dependence for -subjects. Here, I adopt Lewis’ (1973) “standard interpretation” of counterfactuals, where the past—or, in the present context, ’s personal map of the past—is held fixed. Otherwise, the -subjects’ experiences might indeed depend on ’s experience.35
4.5 Defense 1: ordinary epistemology
The last several subsections have explained what it would mean for objects to be phenomenally constructed. Now, at last, I defend PC.
According to PC, acquaintance sentences have phenomenal correctness conditions. In one sense, this is obviously true. To say otherwise would be to deny that experience entitles us to judgments like “That (book-shaped object) is thus (blue).” But any nonskeptic accepts that we are so entitled. And if we are so entitled, then so is the idealized subject, who has all of our evidence and more besides.
Thus, to reject PC, an objector must claim that these rules—i.e., the rules about when experience warrants judgments like —are merely defeasible. Here, the objector might appeal to hallucinations. Suppose that I hallucinate the book. Then is false even though my experience indicates its truth. This may seem to show that ’s correctness is not, in the final analysis, settled by phenomenal truths.
However, notice that describes many experiences. Some of them—e.g., the hallucinatory visual experience—indicate ’s truth. But others—e.g., my friend’s visual experience—suggest ’s falsity. When the idealized subject considers all such experiences, she judges that is false. After all, this is what we actually judge when given similar, although more limited, evidence. So PC rightly predicts ’s falsity.
These remarks apply generally. We know how to identify illusions using experiential evidence. Even when we fail to detect illusions, we recognize how counterfactual experiences could reveal them. But if we can distinguish illusions using experiential evidence, so can the idealized subject. Accordingly, “ordinary” illusions do not falsify PC. They merely show that the rules governing acquaintance terms are intricate: they may invoke any of the phenomenal truths within . Judgments like are indeed defeasible, but only because we lack complete knowledge of .36
One might worry about indetectable illusions: illusions not revealed by any counterfactual experience. For example, most philosophers claim that, in a demon scenario, is true but no objects exist. This would indeed entail that the fundamental correctness conditions for acquaintance terms are not purely phenomenal.
But there are strong arguments against this skeptical interpretation of the demon scenario. In Smithson (2023), I imagine learning that, in fact, we are demon victims. I argue that this would not, in the long term, affect our judgments about objects. Because all experiences are just like before, we would revert to speaking as we had before. I argue that the best explanation of this behavior is that our object judgments are still true in this scenario. Chalmers (2012, 431-440) provides a similar anti-skeptical response.
In summary, acquaintance sentences at least have defeasible phenomenal correctness conditions, and the sources of their defeasibility do not support a contrary view.
4.6 Defense 2: the naive view
Set aside the question of whether PC expresses how we actually use acquaintance terms. Regardless, we could use demonstratives in this way. Let’s call the items denoted by these PC-demonstratives “objects*,” “colors*,” etc.
The application conditions expressed by PC are often met. Thus, it can be stated immediately that many PC-demonstratives refer and many objects* exist. (At present, I assume Thomasson’s framework as a working hypothesis—see 3.4.)
Whatever else we might say, objects* clearly have many features that we naively ascribe to ordinary objects. They have volume, are public, exist unperceived, have sensuous character, and are directly presented in experience. Objects* also align with objects in a fine-grained sense: whenever we (under idealization) judge that an object exists, or has certain sensible properties, there is a corresponding object* with matching starred properties. This is because the correctness conditions for object terms are built into PC via its appeal to an idealized subject employing those correctness conditions. Thus, objects* align perfectly with ordinary objects as naively conceived.
By contrast, there are reasons to think that physical objects are very different than how we naively conceive of ordinary objects. For example, color phenomenology better correlates with subjects’ neurocomputational states than with the physical states of perceived objects.37 Thus, it has been argued that objects are not colored in the way that we naively suppose.38 Similarly, we experience space and time as objectively independent, each with its own qualitative nature. However, we know from relativity that the division of physical spacetime into spatial and temporal dimensions is relative to inertial frame. Thus, it has been argued that physical objects lack the qualitative spatial character that they naively seem to have.39
Thus, while objects* perfectly align with objects as naively conceived, physical objects are, arguably, very different. But then why think that ordinary object terms refer to physical items rather than objects*?
One might reply that the naive conception is mistaken. But we are not required to say this. It is clear that, in daily life, we intend to refer to the things directly presented in experience. But then what stops us? Objects* exist. They are there, presented to us. So nothing stops us. When talking about ordinary objects, we talk about objects*, exactly as we intended.
“But don’t we also naively judge that ordinary objects are physical?” Here, we must distinguish theoretical judgments about objects from ordinary judgments about objects. We ought to uphold ordinary judgments because they express, in the object language, the linguistic rules governing our terms for objects. By contrast, theoretical judgments about linguistic rules are often mistaken. So, while PC is surprising as a theoretical claim, it nonetheless best explains our everyday judgments about objects.
5. The promise of phenomenal construction
Phenomenal constructions are not some esoteric philosopher’s invention. Nothing could be more familiar: they are the ordinary objects that make up the world that we think of ourselves as inhabiting in everyday life. To conclude, I mention two implications of this thesis.
Naive realism: First, it helps vindicate the naive view that perceptual experience involves direct acquaintance with ordinary objects. We can uphold this view of experience because phenomenal constructions, unlike physical entities, are exactly how objects naively seem in experience (see 4.6).
Why should we care to vindicate naive realism? I myself am motivated by the inescapability of the naive view itself.40 Our naive understanding of perception is unlike our naive understanding of nutrition or gardening: beliefs that we adjust or discard as we learn and grow. It is, rather, an understanding continuously pressed on us by conscious life. It manifests our basic grip on what ordinary objects are like. It is also resilient: even when entirely convinced by objections to naive realism, we inevitably lapse into the naive view in everyday life. Skepticism towards it “admits of no answer but produces no conviction” (Hume (2000), ch. 12).
But then I would suggest that any skepticism powerless to affect thought and action is empty: a skepticism founded on mistaken presuppositions about what we mean or about what matters to us. Upon acknowledging phenomenal constructions, we realize that these are the entities that we meant to talk about all along.
Phenomenal construction also deflects the objection that naive realism cannot adequately explain the phenomenal character of hallucinations. In future work, I will defend the view that hallucinations involve direct acquaintance with another species of phenomenal construction: the private objects of 3.5. Unlike traditional sense data, private objects do not screen off ordinary objects in normal perception.
Compare: “Experience cannot present both kangaroos and time-slices. The latter would screen off the former.” This objection fails to recognize that ‘kangaroo’ and ‘kangaroo time-slice’ share application conditions: the former’s application conditions subsume those of the latter. Accordingly, a kangaroo’s presence entails the time-slice’s presence (cf. 3.4). In precisely the same way, private objects do not screen off (phenomenally-constructed) ordinary objects: we are acquainted with both.
Idealism: In 2.1, I assumed an independent physical reality. However, in Smithson (2024), I argue that, if ordinary objects are phenomenally constructed, then fundamental physical entities can also be phenomenally constructed. Thus, the present thesis may serve as a step towards a more ambitious idealist view on which all physical entities are phenomenally constructed.
Of course, anyone taking this road must address major objections to idealism (e.g., that it leaves the coherence of experience unexplained). By assuming an independent physical reality in 2.1, I could set these thorny issues aside.41
This potential connection to idealism should interest philosophers working on fundamental ontology. However, I have argued that even non-dealists should acknowledge phenomenal constructions. Their existence follows straightforwardly from certain familiar, though contested, assumptions of deflationary metaontology
Appendix A: Other properties
We need not insist that objects’ nonsensible properties are phenomenally constructed (i.e., that predicates expressing these properties have purely phenomenal correctness conditions). Still, if ordinary objects are phenomenally constructed, it must at least make sense to ascribe these properties to phenomenal constructions.
First, consider theoretical properties: the properties assigned to ordinary objects by fundamental physics. I will argue that truths about these properties are fixed by phenomenal truths plus microphysical truths (i.e., truths about fundamental particles and natural laws). In semantic terms, the correctness conditions for theoretical predicates (as applied to macroscopic objects) invoke both phenomenal and microphysical truths.
There is one potential obstacle. Ordinary objects inhabit a phenomenally-constructed space (see 3.7). But, under present assumptions (see 2.1), fundamental particles do not.42 Even so, there is a natural functional mapping here: regions and events in physical space typically cause our experiences of regions and events in phenomenally-constructed space.43 Given this mapping, tables and electrons are spatiotemporally related in a functional sense:
A region in phenomenally-constructed space corresponds with a region in physical space when is the normal cause of our experience of .44
An object in region in phenomenally-constructed space is co-located with a set of fundamental particles iff those particles occupy the corresponding region in physical space.
One might object that, on this account, tables and electrons do not really inhabit the same space. But I would insist that they inhabit a perfectly real space, albeit one that is functionally characterized. Compare: upon rejecting plenum physics, one might say that objects are not really solid. But it is better to say that objects are solid, albeit in a functional sense.
We can then adopt further principles to express, in the object language, the correctness conditions of other theoretical predicates applying to macroscopic objects. For example:
has mass iff is the sum of its particles’ masses.
has charge iff is the sum of its particles’ charges
…
Of course, these are rough illustrations.
As for other nonsensible properties (e.g., being a statue), let be the set of truths about ordinary objects involving these properties. Let be the union of microphysical and phenomenal truths. It is typically supposed that supervenes on . With PC, acquaintance truths are settled by alone (as opposed to ). Regardless, the same acquaintance sentences are true (see 4.6). So PC is compatible with standard accounts of higher-level truths.45
One case deserves special comment. We would normally say that ordinary objects, or events involving them, are among the causes of our phenomenal states. This would be strange if objects were constructed from experience like statues from clay. But to say that objects are phenomenally constructed is just to say that acquaintance terms for objects have purely phenomenal correctness conditions. Not only can such items cause experiences, but these causal relations are probably themselves phenomenally constructed. A proxy for whether a relation is phenomenally constructed is whether the idealized subject can judge whether obtains conditional on . Presumably, the idealized subject could judge that ordinary objects cause our phenomenal states; after all, we often make such judgments using much more limited evidence. Alternatively, one could say that such causal relations supervene on , as above.
One might worry that this contravenes my earlier assumption that fundamental physical entities are the ultimate causal ground of phenomenal truths (see 2.2). But, here, we can at least adopt a Humean account of the causal relations between ordinary objects and phenomenal states. These relations obtain just in case some appropriate regularity between macroscopic and phenomenal states obtains. The proponent of PC accepts such regularities.
Notes
- For discussion, see Look (2020, section 5.2) ⮭
- Leibniz used this example; for discussion, see Adams (1994, 247ff). ⮭
- For example, with rainbows, the analysis of “appropriate” coherence will mention the physical causes of rainbow experiences. But the existence/identity conditions for phenomenal constructions are purely phenomenal. ⮭
- Is this distinction tenable? Section 5 discusses whether fundamental physical entities might also be phenomenally constructed. But, short of monism, I describe how we might understand this distinction in the appendix. ⮭
- See Pelczar (2023, 5), who discusses the view that a physical reality causally grounds conditional facts about experiences. I am assuming such a view (see 2.1), though Pelczar rejects the need for any underlying physical reality. ⮭
- Alternatively, Pelczar (2023) analyzes these counterfactuals in terms of facts about experience-related conditional probabilities. One might also take these counterfactuals as primitive. ⮭
- I acknowledge that determinate reference requires some presuppositions about the referent (see 3.4). But these presuppositions are all met—see the italicized clause in the next paragraph. ⮭
- Here, I ignore, e.g., higher-order evidence that I am cognitively-impaired. ⮭
- N.b.: I do not say that this thought is risk-free simpliciter, as a foundationalist might. I grant that some temporal/counterfactual stability in experience is necessary to successfully refer to the presented item. But I am assuming that there is such stability. ⮭
- What exactly is the connection between the presented item and the subject that establishes this cognitive contact? (I thank an anonymous reviewer for this question.) I am inclined to view this connection as a primitive (i.e., unanalyzable) relation of conscious acquaintance. Many philosophers (e.g., Fumerton (1995), Chalmers (2003)) have invoked such a relation, despite disagreement on what exactly we are acquainted with. For detailed discussion, see Chalmers (2003). While Chalmers discusses acquaintance between subjects and phenomenal properties (rather than concrete, sensible items), much of his discussion adapts neatly to the current context. In particular, I agree with Chalmers that, with acquaintance, we can form especially direct concepts whose content is constituted by their referents. There may be other ways to understand this connection. For example, in the appendix, I argue that subjects are causally related to phenomenal constructions. Thus, someone skeptical of primitive acquaintance might instead adopt a causal analysis. Regardless, it remains a manifest fact that we successfully think about the presented item. This fact is all that the present argument requires. ⮭
- See Hill (2021) and Schellenberg (2010), respectively. ⮭
- See Tye (2014, 51). ⮭
- See, e.g., Schiffer (2003), Thomasson (2007). ⮭
- Of course, ‘particles arranged chairwise’ is a technical term. But this expression is just shorthand for the relevant uncontroversial empirical facts. ⮭
- This framework allows for error and ignorance about referents. For example, we may use the name ‘Gödel’ despite knowing nothing of Gödel’s life. Still, speakers must at least tacitly associate ‘Gödel’ with the category of ‘person’ for this name to refer to Gödel rather than the mereological sum of cells composing him. See Thomasson (2007, section 2.6). ⮭
- Some metaphysicians debate whether particles arranged kangaroo-wise compose some further entity: a kangaroo. These philosophers will reject Thomasson’s metaontology (which I am presently assuming), claiming that we can discover the referent (or lack thereof) of ‘Klut’, albeit through broadly a priori metaphysical argument. ⮭
- See Thomasson (2007, section 2.6). ⮭
- The subject could use itself to refer to different things (the single page, the surface, etc.). But, here, we imagine that the subject intends to refer to a three-dimensional entity. Accordingly, this is a case where two demonstratives refer to the same three-dimensional object via ostensions of its distinct parts. ⮭
- In 3.5, I assumed that subjects’ experiences were “appropriately related.” But this assumption did not figure in the correctness conditions themselves. See 4.4 for discussion of why this is unnecessary. ⮭
- See Thomasson (2015, section 6.1). ⮭
- See Chalmers (2009) for discussion and for the example below. ⮭
- This may be partly explained by the nature of human conscious experience. Arguably, our experience presents cups as unified in a way that cup-cups are not. However, I will remain neutral on the precise explanation for why we adopt certain coapplication conditions rather than others. ⮭
- Carnapians settle ontological debates through reflection on our conceptual scheme. But, it may objected, our conceptual scheme does not mention phenomenal constructions. So won’t the Carnapian exclude phenomenal constructions from our ontology? In fact, I argue below that our conceptual scheme does countenance phenomenal constructions. But, if I am mistaken, this paper can be reinterpreted as a pragmatic proposal to introduce terms for phenomenal constructions. See Thomasson (2007, 183-185) for discussion. ⮭
- (*) expresses semantic dependence between the term ‘stone’ and its application conditions, rather than metaphysical dependence between stones and particles. There are delicate questions regarding the connection between these types of dependence—cf. Chalmers (2012, E16). Here, I focus on the semantic mode. ⮭
- Compare to Foster (2008, vii): “the existence of the physical world is logically sustained by the world-suggestive way in which … things are disposed to appear at the human empirical viewpoint.” ⮭
- See 2.2 for discussion. ⮭
- Here, may or may not be identical to ; see 3.5 for discussion. ⮭
- By “indirect,” I mean: the application and coapplication conditions for acquaintance terms can be read off from truths of the form ‘Thisi exists’ and ‘Thisi is identical to thisj’, respectively. See 4.2 and 4.3 for discussion. ⮭
- Strictly speaking, should be replaced by &, where is a “that’s all” clause stating that there are no further phenomenal truths other than those that are implied by . See Chalmers (2010, ch. 6) for discussion. ⮭
- See Dorst (2019) for discussion. ⮭
- Strictly speaking, must be expressed using pure phenomenal concepts: concepts that express phenomenal properties in terms of what it is like for subjects having those properties. This is to avoid circularity: no acquaintance truths should be built into . But, for convenience, I will refer to, e.g., “the experience as of approaching the apparent book.” ⮭
- A reviewer asks: what if we only believe that such coherence establishes the existence of objects because we also believe that there is a non-phenomenal cause of such coherence? This stance is consistent with the assumptions of this paper, according to which phenomenal coherence is explained by an underlying physical reality (see 2.1). ⮭
- See 3.10 for discussion of the relevant metaontology. See also Price (1932, 286f). Note, however, that the current account does not invoke sense data. ⮭
- I thank an anonymous reviewer for this question. ⮭
- For a second approach to intersubjectivity, see the account of ideal spacetime in Pelczar (2023, ch. 4). As backup, one might distinguish -experiences from -experiences via their distinct physical causes (see 2.1). While this would no longer be true phenomenal construction, the resulting entities would still have the sensuous character described in 4.3. Accordingly, much of the discussion ahead could be adapted for these “mixed constructions.” ⮭
- We also make cognitive mistakes. But PC’s cognitive idealization precludes such errors. ⮭
- See Brouwer and Heeger (2013). ⮭
- See Pautz (2010). ⮭
- See Thompson (2010). ⮭
- For other motivations, see Martin (2002) and Campbell (2002, ch. 6). ⮭
- For discussion, see Pelczar (2023, ch. 3), Smithson (2024). ⮭
- For a contrasting view, see Foster (1982, 162-180), Pelczar (2023, 185-193), and Smithson (2024). ⮭
- See Chalmers (2012, section 7.5). ⮭
- Alternatively, we might define correspondence between manifolds of spatial properties, as does Chalmers (2012, section 7.5). ⮭
- Excluding obvious exceptions, e.g., ‘Objects are not phenomenally constructed’. ⮭
References
Adams, R. M. (1994). Leibniz: Determinist, Theist, Idealist. New York: Oxford University Press.
Alspector-Kelly, M. (2004). Seeing the unobservable: van fraassen and the limits of experience. Synthese 140(3), 331–353. doi: http://doi.org/10.1023/B:SYNT.0000031323.19904.45.
Alston, W. P. (1999). Back to the theory of appearing. Philosophical Perspectives 13, 181–203. doi: http://doi.org/10.1111/0029-4624.33.s13.9.
Brouwer, G. J. and D. J. Heeger (2013). Categorical clustering of the neural representation of color. The Journal of Neuroscience 33(39), 15454–15465. doi: http://doi.org/10.1523/JNEUROSCI.2472-13.2013.
Campbell, J. (2002). Reference and Consciousness. Oxford: Oxford University Press.
Chalmers, D. J. (2003). The content and epistemology of phenomenal belief. In Q. Smith and A. Jokić (Eds.), Consciousness: New Philosophical Perspectives, pp. 220–272. Oxford: Oxford University Press.
Chalmers, D. J. (2009). Ontological anti-realism. In D. J. Chalmers, D. Manley, and R. Wasserman (Eds.), Metametaphysics: New Essays on the Foundations of Ontology, pp. 77–129. Oxford: Oxford University Press.
Chalmers, D. J. (2010). The Character of Consciousness. New York: Oxford University Press.
Chalmers, D. J. (2012). Constructing the World. Oxford: Oxford University Press.
Dorst, C. (2019). Humean laws, explanatory circularity, and the aim of scientific explanation. Philosophical Studies 176(10), 2657–2679. doi: http://doi.org/10.1007/s11098-018-1145-0.
Foster, J. (1982). The Case for Idealism. London: Routledge & Kegan Paul.
Foster, J. (2008). A World For Us: The Case for Phenomenalistic Idealism. Oxford: Oxford University Press.
Fumerton, R. (1995). Metaepistemology and Skepticism. Lanham, MD: Rowman & Littlefield.
Hill, C. S. (2021). Perceptual existentialism sustain ed. Erkenntnis 86(6), 1391–1410. doi: http://doi.org/10.1007/s10670-019-00160-z.
Hume, D. (1748/2000). An Enquiry Concerning Human Understanding: A Critical Edition. New York: Oxford University Press.
Johnston, M. (2004). The obscure object of hallucination. Philosophical Studies 120(1–3), 113–183. doi: http://doi.org/10.1023/B:PHIL.0000033753.64202.21.
Lewis, D. K. (1973). Counterfactuals. Cambridge, MA: Harvard University Press.
Loar, B. (2003). Transparent experience and the availability of qualia. In A. Jokić and Q. Smith (Eds.), Consciousness: New Philosophical Perspectives, pp. 77–96. Oxford: Oxford University Press.
Look, B. C. (2020). Gottfried Wilhelm Leibniz. In E. N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy (Spring 2020 ed.). Metaphysics Research Lab, Stanford University.
Martin, M. G. F. (2002). The transparency of experience. Mind & Language 17(4-5), 376–425. doi: http://doi.org/10.1111/1468-0017.00205.
Pautz, A. (2010). Why explain visual experience in terms of content? In B. Nanay (Ed.), Perceiving the World, pp. 254–309. New York: Oxford University Press.
Peacocke, C. (2008). Sensational properties: Theses to accept and theses to reject. Revue Internationale de Philosophie 62, 7–24. doi: http://doi.org/10.3917/rip.243.0007.
Pelczar, M. (2023). Phenomenalism: A Metaphysics of Chance and Experience. Oxford: Oxford University Press.
Price, H. H. (1932). Perception. London: Methuen.
Russell, B. (1918). The philosophy of logical atomism. The Monist 28(4), 495–527. doi: http://doi.org/10.5840/monist19182843.
Schellenberg, S. (2010). The particularity and phenomenology of perceptual experience. Philosophical Studies 149(1), 19–48. doi: http://doi.org/10.1007/s11098-010-9540-1.
Schiffer, S. (2003). The Things We Mean. Oxford: Oxford University Press.
Smithson, R. (2023). Edenic idealism. Australasian Journal of Philosophy 101(1), 16–33. doi: http://doi.org/10.1080/00048402.2021.1981804.
Smithson, R. (2024). In defence of macroidealism. Australasian Journal of Philosophy. doi: http://doi.org/10.1080/00048402.2024.2416974.
Thomasson, A. L. (2007). Ordinary Objects. Oxford: Oxford University Press.
Thomasson, A. L. (2015). Ontology Made Easy. New York: Oxford University Press.
Thompson, B. (2010). The spatial content of experience. Philosophy and Phenomenological Research 81(1), 146–184. doi: http://doi.org/10.1111/j.1933-1592.2010.00359.x.
Tye, M. (2014). Transparency, qualia realism and representationalism. Philosophical Studies 170(1), 39–57. doi: http://doi.org/10.1007/s11098-013-0177-8.