What is Real? (New Paper)

Published yesterday in Organon F: International Journal of Analytic Philosophy (30.2: 182−220).


Two of the most fundamental distinctions in metaphysics are (1) that between reality (or things in themselves) and appearances, the R/A distinction, and (2) that between entities that are fundamental (or real, etcetera) and entities that are ontologically or existentially dependent, the F/D distinction. While these appear to be two very different distinctions, in Buddhist metaphysics they are combined, raising questions about how they are related. In this paper I argue that plausible versions of the R/A distinction are essentially a special kind of F/D distinction, and conversely, that many F/D distinctions imply an R/A distinction. Nevertheless, while this does suggest that the F/D distinction is more basic than the R/A distinction, it does not favor a particular understanding of the F/D distinction. There are many kinds of existential or ontological dependence that cannot be meaningfully combined into a single notion, and reality does not force us to accept any specific kind of dependence as more fundamental. Consequently, what we consider to be ‘real’, ‘fundamental’, or ‘really existing’ is not entirely given by reality, but partially up to us.

(The remainder of this post consists of some edited extracts from the paper. The full paper can be downloaded here.)

Two of the most fundamental distinctions in metaphysics are that between things in themselves and phenomenal appearances, and that between entities that are fundamental, real, or independent (in some relevant sense) and entities that are not (or less so). According to the first distinction — which I shall call the reality/​appearances or R/A distinction hereafter — there is at least a possibility that things as we experience them (or as they appear to us) are different from how they really are, independently from us. There is considerable variation in the terms used to make this distinction. The world as it appears to us (or the world of appearances) is sometimes called ‘phenomenal reality’ or ‘conventional reality’, for example, leading to an apparent distinction between two different kinds or levels of reality or two realities. Alternatively, the distinction may be conceptualized as involving two perspectives on, or aspects of reality, or in similar terms. Kant’s distinction between things in themselves and phenomenal appearances is, more or less, the paradigmatic R/A theory, but Thomas Kuhn’s (1962) famous claim that “after a revolution scientists are responding to a different world” (111) also presupposes a distinction between some kind of independent reality and a world of experience (i.e., the world scientists respond to), and further variants of the distinction can be found throughout the history of philosophy.

According to the second distinction — which I will call the fundamental/​dependent or F/D distinction hereafter — not all things that can be said to exist have the same ontological status: some entities are substances, while others are ontologically dependent, or some entities are more fundamental than others, or more real (in some ontologically loaded sense of ‘real’), and so forth. An event of alpha decay, for example, is ontologically dependent on the atom that emits the alpha particle, and a water molecule is ontologically dependent on the oxygen atom and two hydrogen atoms that constitute it.

On the face of it, these appear to be two very different distinctions. Although R/A theories generally (implicitly) assume that phenomenal appearances depend for their existence on the independently real things that cause or ground them, they rarely appeal to an obvious or explicit F/D distinction to explain the relation between phenomena and things in themselves. F/D theories, on the other hand, typically assume a single reality without ‘levels’ or ‘aspects’, and thus appear to deny the R/A distinction. The water molecule and its constituent atoms in the last example do not exist in different kinds or levels of reality (or in different perspectives on reality, or different realities, etcetera). Rather, in the F/D perspective there is just one reality, but some things in that one reality are more fundamental or more real than others. However, despite this apparent incompatibility, in Buddhist metaphysics the distinction between ultimate reality (paramārthasat) and conventional/​phenomenal reality (saṃvṛtisat) is both an R/A distinction and an F/D distinction, and this raises the question of how different these two ontological distinctions really are.

In this paper I argue that plausible versions of the R/A distinction are essentially a special kind of F/D distinction, and conversely, that many F/D distinctions imply an R/A distinction; or in other words, that the two distinctions are not as fundamentally different as they may appear to be. R/A theories hold that phenomenal appearances depend (among others) on their independently/​externally real grounds or causes. This is an existential dependence relation in which appearances are the dependent and the things in themselves that ground or cause them are the independent (or more fundamental or more ‘real’). Hence, this is an F/D distinction. The other way around, many F/D distinctions involve some kind of conceptual dependence. In case of the dependence of wholes on their parts, for example, we probably would not even recognize the whole as an individual entity without a concept naming or describing it. In other words, we have a phenomenal appearance of that whole as something, which depends (among others) on a concept and which is not (necessarily) given (as such) by the independently real thing(s) that ground that appearance. This is an R/A distinction.


Varieties of ontological dependence form a subset of varieties of existential dependence, which is loosely defined by means of a counterfactual conditional: x existentially depends on y if and only if, if y would not exist, then x would not exist, and not just because x exists necessarily. Neither existential dependence, nor ontological dependence is more than a collection of varieties, however. There is no such thing as ‘ontological dependence’ or ‘existential dependence’. Rather, the many different kinds or varieties of existential dependence relations have different (formal and other) properties, and combining them into a single category is more likely to be misleading than helpful. At best, such a general/​aggregate notion of existential or ontological dependence is redundant because it does not explain anything.

One specific kind of existential dependence is the dependence of phenomenal appearances on conceptual construction, which grounds the distinction between appearance and reality. Consequently, the R/A distinction is a special kind of F/D distinction. Furthermore, many other kinds of existential dependence imply or involve some kind of conceptual dependence, and therefore, F/D distinctions often come with (implicit) R/A distinctions.

Different F/D distinctions are different ways of thinking about what is real and what is not (or what really exists and what does not), but a conception of ‘real’ based on some kind or kinds of existential independence — and this includes the R/A distinction — inherits the latter’s problems. If a general/​aggregate notion of existential or ontological dependence does not explain anything, then neither does a conception of ‘real’ built upon such a notion. Hence, a notion of ‘real’ based on such generalized dependence would be explanatorily redundant.

If the generalizing approach does not work, the most obvious alternative is to select or prioritize one or a few specific kinds of (in-) dependence. As mentioned, F/D distinctions often come with (implicit) R/A distinctions, and intuitively, F/D distinctions that involve R/A distinctions seem to be more fundamental (at least to me) than those that do not, just because there are more kinds of dependence involved. For example, whole/​parts dependence or events/​participants dependence both involve conceptual dependence (and thus an R/A distinction), because without a concept naming/​describing the whole or event, we would not (normally) recognize or experience it as such (i.e., as an individual thing, in the broadest possible sense of ‘thing’). If this intuition is right, then parts are more ‘real’ than the wholes they constitute, and endurants are more ‘real’ than the events they participate in. Causal dependence, on the other hand, does not necessarily involve conceptual dependence, as both cause and effect can be parts of independent reality and phenomenal appearance plays no role in their causal relation (in that case!). For this reason, whole/​parts and events/​participants dependence seem more fundamental kinds of existential dependence than causal dependence.

F/D distinctions in which the dependent and the (relatively) independent belong to different ontological categories also seem intuitively more fundamental (again, to me) than those that do not. For example, in events/​participants dependence, the event and the participants belong to different ontological categories (i.e., events or occurrents and endurants, respectively), while this is usually not the case for causes and their effects in causal dependence. This suggests again, that endurants are more ‘real’ than the events they participate in, and that effects are just as ‘real’ as their causes (or in other words, that causal dependence does not make something less real).

These are mere intuitions, however, and I have no good argument for either intuition. The problem is that, besides intuition, there does not seem much to go on. (And I do not trust intuition — yours even less than I trust my own.) Nothing in reality forces us to conceive of ‘real’ or ‘exists’ in a particular way, or to choose between varieties of existential dependence. What we consider to be ‘real’ or ‘really existing’ is not given by reality, but decided by us. By implication, ‘real’ is a relative term — it is relative to a conventional metaphysical distinction.

Nevertheless, this does not mean that we have complete freedom to decide what is real and what is not. In Realism with a Human Face, Hilary Putnam considers a ‘World 1’ consisting of three objects x1, x2, and x3, and a ‘World 2’ consisting of those same three objects plus their mereological sums (i.e., three combinations of two, and one combination of three) making seven objects in total (p. 97). I would not call these two different cases ‘worlds’, but two different descriptions of the same world, and the same world could also be described as consisting of only one object, namely, the mereological sum of x1, x2, and x3. (I suppose that this description could then be called ‘World 3’.) However, our choice in deciding which description is the ‘right’ one and which of these (three, seven, or one) objects ‘really’ exist is limited to those three options. Saying that there really are 42 objects would be plain false. Something similar applies to our choice in deciding what is ‘real’ in the world we live in. We can choose to say that chairs are real or that only the elementary particles they ultimately consist of are real, for example (and nothing important might depend on that choice), but we cannot decide that unicorns are real.

Nevertheless, while independent, external reality sets limits to our metaphysical description(s) of the world (at least, in as far as we want those to make sense), the description we choose within those limits is largely conventional. Again, the world does not force us to conceive of ‘real’ or ‘exists’ in a particular way. The qualification ‘real’ is not given by reality, but relative to a convention, and lacking objective criteria to transcend that convention (i.e., to objectively decide what ‘real’ really means), any use of the term ‘real’ (or ‘exists’ or any other variant) that does not (explicitly or implicitly) acknowledge this relativity is empty rhetoric. It is like claiming that the sky and lapis lazuli are really the same color, namely, blue, without recognizing that what is called ‘blue’ in English is at least partially conventional and that other languages (such as Russian or Japanese) have different conventions in this respect, and would, therefore, describe the colors of a cloudless sky and a piece of lapis lazuli with very different words. (Russians and Japanese would use the words голубой goluboy and 水色 mizuiro to describe the color of the sky, and синий siniy and 青 ao for the piece of lapis lazuli, respectively.)

The answer to the question ‘What is real?’ then, is ‘It depends.’ It depends on one’s conception of ‘real’, and there are multiple equally truthful conceptions of ‘real’ and no objective criterion to choose and elevate one of them as the one and only ultimate standard of reality. Nāgārjuna famously held that emptiness (i.e., existential dependence) is itself empty (i.e., merely conventional). We have reached a similar conclusion here — what we consider to be real (i.e., not existentially dependent) is itself dependent on convention (and thus empty, in Nāgārjuna’s terms) — but the argument that led to this conclusion is rather different from Nāgārjuna’s. (Not in the least because the paper actually rejects Nāgārjuna’s use of causal dependence as a criterion of reality.)

If you found this article and/or other articles in this blog useful or valuable, please consider making a small financial contribution to support this blog, 𝐹=𝑚𝑎, and its author. You can find 𝐹=𝑚𝑎’s Patreon page here.

Leave a Reply

Your email address will not be published. Required fields are marked *