One of the oldest and seemingly most impossible programs of classical philosophy is that of a description of what there is. Ontology, the description of reality as such, seems to have taken too big a field, and too open a goal, to ever be able to finish, and even worse (as other fields are also, in this sense, open), to even have a measure of how unfinished it is at every step of the way. This becomes even more obvious, when looking at the vast difference between what is generally considered to be "reality", and the portion thus studied. In what way then can we understand ontology? Can we study existence, or what can that even mean?
We must first understand that there is a way, in which it _seems_ as if ontology could operate, but which it can't, and that is the method of categorical reductionism. Reductionism in general is a form of splitting a difficult problem into simpler ones, and then asserting, that with these problems the whole would already be solved, that there is nothing beyond this to do; it's a rhetorical gimmick that closes discussion after, seemingly, having exhausted all possible alternatives or parts, when really these "alternatives" are parts of the method, not of the reality. Specifically, there are two obvious variants: Mereological reductionism treats reality as made up of stuff, and acts as if, when looking at what something is made from, the thing is also understood. There is a trivial way in which this is true, when we namely simply partition the thing into everything that belongs to it; but that in fact is not what is typically done. Instead, parts that can be seperated or seperatable from it are understood to be the only ones, and therefore relations, as non-removable parts of a whole, ignored. That an object consists not just of matter, but also of form, means, that material parts are simply not enough to compose the things. And even if all such parts where there - they would _make up_ the thing, but not _be it_, as a thing is not a collection of parts and properties, but a connection point of them to even become such a thing as one thing. That much to the absurdity of mereological reductionism.
Categorical reductionism is a lot more difficult to dismiss, and harder to avoid. It does not reduce a whole to parts, but parts to a whole, or more specifically: If I describe a group of things, and categorize them as A, B, C etc., then this can be useful to describe these things. That is one possibility. However, there can be other descriptions of these things, and it is clear that these categories are said _about_ these things. Let's imagine that any object in that group falls under exactly one category, has exactly one of those properties. Now, if we were to take this group of things apart, and group them by these properties, then they would be indeed divided by this property, they would simply be either A, or B, or C etc. But currently, they are not that. The "group of As" is only a concept created from these objects, is only something existing as a predicate of them, nothing existing in itself yet. It can be created, the objects can be seperated. But if they happen to be not seperated to begin with, then that is an extra step, and not part of description, but action. These qualities therefore are not real divisions, in the sense that there is no super-object that would contain all As etc. However, very commonly, a second attitude, a kind of reversion of mereological reductionism arises, that precisely assumes this: it assumes, that if anything is either A or B or C etc., than there is nothing but A and B and C etc. The things are reduced to exemplars of the categories that happen to be true of them. This is done by a way of _linearization_ or _substantiation_ of species: that all things are sub-categorized, as uniquely belonging into a specific nieche in a Porphyrian tree, that should also fit everything. It is not that there are no substantial qualities, but that they are substantial _relative to a criterion of relevancy_. Two different criteria of relevancy can exist that have different concerns, and will not make it possible to create a single tree of substances, even if both have one (which already is a bold assumption). This means, that we cannot treat categories as stand-ins for groups of objects which happen to fulfill them; and that means, the reduction can't work. And even if the tree existed, the category is different from that under it. At that point, it is only a technical point, but it is a technically significant point that becomes ever more important that its real, material condition of fulfillment, of these abstract substances being unique, is not true in the first place.
Therefore, we cannot say: everything is either matter or mind; the same way we can't say: everything is either a blue T-shirt or a non-(blue-T-shirt). In some way obviously, anything is either A or non-A, but it is not that, and a total, exhausting, unique and non-overlapping sequence of such categories, without any non-As in there at some point, is impossible, unless we were to list anything and everything, at which point there would be no reduction to do, anything would simply exist. We either are missing something, or are creating an aritifical object like the "non-(blue-T-shirt)", which is not a category that something substantially is. A house is not an non-T-shirt; in exactly this way it is also wrong to call anything not mental a physical or material thing.
These problems however are different: on the one side, that this tree of substances cannot be complete, that we can't have a full description; on the other, that the partial descriptions we have are not essential and substantial only to a certain criterion of relevancy. (For example, that the categorization of animals as feline, under which house cats and tigers fall equally, is appropriate in biology, but the differentiation into pets and non-pets into another; the cat however is never essentialy a pet or a feline, but a living being here and only for the purposes of categorization an instance of an abstract property that is true of it; that is, these are properties, not things-to-be). These two are however connected: in that the idea of the ultimate, complete tree of substances cannot safe us from the empirically observable imperfection of categorical schemas.
What was the point of talking about this? Well, to avoid the basic problem of ontology: of confusing the study of reality, of existence, with the study of substance and substantiality. Both are important, but mean a different thing. And their connection is in fact not in the substance or the reality, but within substantiality, in its criterion of relevancy and its desire. This needs to be further explained:
Any way of studying reality starts with a particular desire of knowledge, or of a way of looking at the world. This desire of knowledge creates its own form of relevancy, as that what is seen as important in that study, as opposed to that, what is rejected, because it was not the knowledge that was desired. Think about how we categorize things differently when we look at them in physics, art or politics. There are certain things important for this understanding - that what seems _substantial_ - and things irrelevant, that are mere _accidentals_ for this particular desire of knowledge. It can then have its fulfillment in the creation of its own (possibly self-overlapping) Porphyrian tree, starring its own categories of describing the world, specific to its own criterion of substantiality and that what is real (for example: material subsistence vs intention in the physical sciences, meaning/style vs material in interpretation of art, public vs private life in politics etc.). None of these categories are _more substantial_; rather, each of them come from the substantiality that is part of a different desire to know reality, whether it is to know beauty, justice or the laws of nature. However, it is important now to differentiate substantiality of this form against two very different things, that are often confused with it, and those are the structures or systems of reality itself on the one side, and truth and the structures of discourse, as social phenomena and their subjective representation, on the other.
Reality is that what knowledge can be about, but cannot have. I can know a tree, understand it biologically, write a poem about it, even possess the land it grows on; but none of these knowledges actally have the tree. What "has" the tree is the land its on, the people that care for it etc, so not forms of knowledge, but other realities. The relation to the tree is "intentional", that is, it refers to the tree, includes it, and "has" it in the sense of having the thought of it. But reality is always external to knowledge. This is one way of differentiation "knowledge" as one form of theoretical action from others. Erotic sentiment is not knowledge, and neither is a legal review, since they do in fact try to "have" their object, to create a situation. Political philosophy is a form of knowledge, laws aren't; theories about desires are knowledge, erotic poems aren't. This is not because they are less than knowledge, but because they are not "about" something; they are enacting something in a more direct way. And this might be related to how they are more closely related to language; how they try to enact or create realities in language, in a way in which knowledge or thinking about reality is stuck in a way to be in between them, and never to quite make them match.
Because while on one hand restrained by external reality, thinking and knowledge specifically is also limited by language as a way of expression. This is also not surprising, and indeed a stereotype of the description of knowledges is that they are seen as discourses, as bound in language. There is a lot of truth in that description, but also a strong limitation. And that limitation is the way in which it sees knowledge and thought itself as a languistic effect, and not as something that precedes it; that it equates ideas with expressions or descriptions of ideas.
The problem of this is not merely that ideas are not solely shaped by language (although one could hypothesize that most are to some extent created thus), but that any thought can be expressed any number of ways, and that any expression can be interpreted to have many different thoughts attached to it. The "arbitrariness", which is supposed about language, is not true about thinking, unless thinking is seen as a mindless processing of data, as is clearly absurd in my view of it. This is then often taken to refute this idea of language; to posit some "deep meaning" or such about language that would create a non-arbitrary way of expression. But I don't think that's necessary at all.
My main position here is that thinking is non-discursive. While language is important for expressing and describing thoughts, thinking itself is simply not a linguistic or discursive process. When I think about something, I try to connect ideas and images, general concepts and objects I know, to form some sort of meaning; I can express that to myself linguistically, and do need to do that to "have" that connection, since, as I stated, I don't have the objects, and even less abstractions, really only leaving me with words; but that doesn't mean that the thought itself is these words!
Thinking means to differentiate. Thoughts are differences or comparisons of ideas of things. Therefore we have three stages: First, the things themselves, reality; then, us seeing or understanding them, getting the idea of them; and thirdly, us thinking them together to a thought, by comparing and contrasting different ideas. We then associate words to ideas and sentences to thoughts; and we can even, when we have words already connected to ideas, construct arbitrary sentences and then try to think the thought associated to that. However, while we can say all thoughts (or basically all thoughts; we may forget the occasional word, or have to invent it), there are many sentences we cannot think, such as "this prime number is a yellow apple"; it tries to connect two ideas we can't even really compare. In this sense, thinking seems to me to be primary to language, and language, while important and great at expressing thoughts, not so much to be its medium as its imprint, as a necessity for memory and discussion, but as too broad to be a medium in which everything expressible in imaginable. (That is in opposition to other uses of language, such as poetry or song, where really all expressions are relevant to the art form, and the limits are that of grammar, not imagination, if we ignore the limits of the poem or song of making sense as an idea, as is often done within these forms already; the limitation to that, as far as it happens, is more due to the affinity of language as such to be expression of thought, attempted or failed, rather than an affinity of poetry or song specifically to be close to that form of use of language in particular.)
If our understanding of categories as thoughts comes from an criterion of relevancy, than this criterion is precisely a form of understanding which comparisons or contrastions make sense, and to interpret the ideas accordingly. They therefore are related to different "languages", namely to the subset of sentences that can express/represent the allowed or accepted connections as thoughts. Therefore, if we want to stay loyal to the comittment of that form of substantiality, we need to talk in a way both reflective of connections reflecting the corresponding interpretations of ideas of reality, and to the limitation of linguistic form to not express connections outside of those in a way expanding our interest beyond what is consistent of it. The virtue to do that - to stay in the bounds of the language of a particular form of relevancy or substance - is what we call truth. Truth is, in this precise way, the virtue of rhetoric: that it is a the virtue of a text to express thoughts that correspond to a particular desire to know or represent reality as ideas.
To say the truth therefore means to speak in such a way, that the corresponding thought is in the bounds of a particular relevancy, _and_ corresponds to the way, the thought is about, is represented, relative to that relevany, as ideas or representations of these objects. Truth is opposed both to methodological falseness - as meaninglessness - and to material falseness - as a factual mistake. Therefore, a sentence cannot be said to be true or false solely in relation to reality, but only in relation to a particular substantiality applied to reality. A sentence is not true or false; but a person can be in truth or in falsehood using a sentence for a particular goal of understanding on a particular thing.
However, it is not quite so easy with language. Language is mostly not as transparent as described here, and indeed it is almost a cliche to point to its opaqueness. What are we to make of that? Is this whole project in some way doomed - is there not internal representation? As I said, I think this is nonsense. There are thoughts, they are connecting ideas in the way I described, and truth really is the virtue of a certain kind of connection. However, truthfullness of this kind needs to be supplemented with the virtue of rhetorical honesty of making the connections between ideas and words consistently, of trying to limit misunderstandings. That is the point where an explificly plural understanding can become interesting: how can I be sure to have made these connections as I do now; how do I guarantee the consistency of language use? etc; and from that we can develop a whole dialectical theory regarding the various contradictions and internal confusions of the use of language vs. the confusions of thoughts in their possible connections of reality as represented by the variously motivated kinds of ideas and representations of it.
But all of that changes nothing of the fact that thoughts exist. The imperfect and shifting nature of language changes nothing about the fact that language can represent thoughts, and if it does, and if someones does represent (in a direct or more complicated linguistic structure) these thoughts accurately, that this person then is honest, and if that thought also reflects accurately the connection or disconnection of ideal representations of reality in relation to that same theoretical desire, that this person then is truthfull or in truth. It just means that it is more difficult a thing to accomplish.
What can we then however say about ontological truth? What can we truthfully say about what reality as such is?
Reality is not as much as throught related to language; it however is also not the anti-lingual, as that would be a categorical reductionist perspective of the only relative difference of reality and discourse (that are opposite as categories, not things; discourse is real, and reality can therefore also be discursive, if it is the reality of an already discursive thing etc.).
But there is a way in which, through the layers of language and thought, we can approach reality: that reality is that what is present in ideas, as the basis of phenomena which we try to understand; that reality is that, what appears. Reality is not appearance - reality is something, that appears. Its appearing becomes our ideas and perceptions, the basis of thought. But that, what appears, lies beyond that. It is, what we can approach as the basis of the ideas that are of it.
Again, that does not mean that reality is anti-appearance, as appearances are also real. But it means that for reality to become an object of thought, it must appear to the subjet that's trying to think it. And we have also the problem, that this is different to reflection, that is to the thinking of these appearances; for that is not based on the appearance of reality, but on the appearance of the appearance of another reality, which is only a specific type of reality (the reality of an appearance), and can therefore not be used to model or understand appearances of realities in general, as not all appearances might be itself such a thing, that appears.
That reality is, what appears, does not mean that all of reality appears. There might be elements of reality forever hidden, by whatever partition, and others only apparent at some time and to some people. What-appears is a portion of reality, not all reality; but it describes precisely that portion which could even become an object to any theoretical investigation. What-appears is, in this sense, the domain of ontology.
The topic of ontology is then to differentiate about this reality its appearing, as an act to appear, from its appearance in the subject. This difference can be more abstractly described as the difference between the being-there (Sein) and the what-is-there (Seiendes), as "being" is just a different way of (spatial-temporal or abstract) appearance. In this sense, the now classical distinction between the ontological and the ontical is precisely that between reality and its appearance. Insofar appearances are formed, as ideas and perceptions, by certain ideas of substantiality and criteria of relevancy, this means this is also describing the difference between reality and substance. "Substance", as described by a particular theoretical scheme, is a description of or about reality, for and by that particular goal or desire for theory; "reality" is that of what-appears that is not captured by the (theoretical) description of it as substantially this-or-that.
The only real method of ontology seems to me to constantly look at the various different appearances of the same thing, of the different substantializations under different ideas of substance in science, art, politics, erotics, religion etc, and to see that it still is _the same thing_, that the difference is one of substance, not of reality. This then still gives us no description of what this reality is, beyond all substantializations, as any description would again substantialize it, would choose certain qualities as "essential" over "accidentals". But what it does give us, is this sense of connection of even very different kinds of substantialities: that there really is a reality, and that as much as it is not truly "external" to thought or can only be captured by thought, it is a constraint on the bounds of truthfull expression of thinking, to understand, that truth exists, and that truth consists in a desire to understand reality, as that what is common and can only be understood through the common and connecting element of all the various ways, in which reality appears to be substantiality itself, to be there, to be something that can appear. If that desire is satisfied, or to what extent reality lies outside of this boundary, or remains dark, no ontology can answer; but it can give us a boundary, as to what at least should be seen to be in the light of truth.