"The Objects of Thought" Review

So I have just finished reading the very interesting work by Tim Crane entitled "The Objects of Thought." For those who are familiar with Crane's work, he is an analytic philosopher operating within the Brentanian tradition. Being a Meinongian myself, this means that he is operating from a very similar framework to the one from which I operate. In this post, I will elaborate my thoughts on the book, along with some initial critiques of the theory advanced.

To begin, Crane is specifically concerned with the problem of non-existence, which he understands as arising from a seeming inconsistency between two allegedly obvious truths. Firstly, there is the notion that many true statements can be made regarding nonexistent objects. Secondly, there is the notion that all truth must be grounded in existence. Now of course, the meinongian would wholeheartedly agree with the first proposition, but he would staunchly disagree with the second, since he allows for many truths to be grounded in what we may call 'unreality'. However, due to his Brentanian proclivities, Crane is forced to ground all truth in existence, leading quite directly to the seeming inconsistency.

The question now is: how to reconcile the seemingly contradictory statements? In a decidedly clever move, Crane advances what we might call a 'Neo-Brentanian' solution to the problem. Quite simply, he manages to achieve this by explaining all the truths regarding nonexistent objects in terms of truths about existent objects. Indeed, as will become evident in due course, Crane's account is a resolutely Noneist account, revealing at once that Noneism can include more than just meinongianism. With all that being said, let us delve deeper into Crane's proposed solution.

As one would expect from a Brentanian, intentionality plays a key role in Crane's theory. And like Brentano, Crane interprets intentionality (also aboutness) in a non-relational way. For rather than being a relation of mind to object, Crane holds that intentionality is a universally immanent property of all mental operations. One primary reason why Crane rejects a relational account of intentionality (such as the Meinongian account) is because he thinks that relations can only ever hold between existent objects. Thus, if intentionality were a relation, then our mental states could only ever be about existent objects, a conclusion which Crane voids at once, since he does not see it as doing justice to the phenomenological data. In contrast to this rather distasteful conclusion, Crane holds it as axiomatic that many of our mental states can be about nonexistent objects.

But quite unlike Brentano, Crane also holds that we can quantify over nonexistent objects. As anyone who is familiar with Brentano's logical theory is surely aware, he interprets his particular quantifiers existentially, whereas Crane interprets them neutrally. Crane also diverges from his master by holding that nonexistent objects can possess a limited range of properties. These two facts, coupled with the further fact of his equating being with existence, is sufficient to make Crane a Noneist.

But whereas most Noneists equate the phrase 'there is' with the phrase 'there exists', Crane holds that 'there is' is not ontologically committing. So Crane is quite happy to say such things as "There are nonexistent objects" while this phrase hurts the ears of most Noneists. Now since Crane rejects the Quinean view of ontological commitment, one would expect him to hold that existence can be formally captured with an existence predicate, and this is precisely what he says. So just like in Meinongian Logics, Crane can define the existentially-loaded quantifiers through the use of both his neutral quantifiers and the existence predicate.

But when are we committed to the existence of a particular item O? Well for Crane, we are only ever ontologically committed to O if we both believe that O exists and if we have a conception of O.

We have already mentioned how Crane allows for nonexistent objects to possess properties in his theory, but he allows this only for a quite limited range of properties, since he takes the vast majority of properties to be existence-entailing. This places his theory much closer to Priest's Noneism, rather than Routley's. But there are 2 radical differences with Priestly Noneism, and that is Crane's rejection of both the Characterization Postulate and worlds-machinery. The Characterization Postulate (CP) is briefly put a principle that allows us to gain epistemological access to nonexistent objects. By appealing to the CP, we can determine just which properties any particular nonexistent object has. (If you would like a more in-depth explanation of the CP, either stay tuned for a future post on the matter, or ask me about it in the comments below).

Now for those who are familiar with Priest's version of the CP, it stipulates that if any existence-entailing properties appear within the characterization of a nonexistent object, then we can infer that that object has those properties not at this world, but at all the worlds that realize the characterization. In this case in particular, that would be all the worlds in which it exists. Of course, since characterizations can be incomplete or inconsistent in various ways, we must allow for these objects to have their properties at impossible worlds of various kinds.

However, Crane rejects all of this theoretical machinery. Again, he still asserts that many properties are existence-entailing (indeed, he thinks that the number of these is far greater number than what Priest thinks), what he does not allow is for nonexistent objects to possess these properties in any sense whatsoever. So if a particular nonexistent object O were to be characterized in such a way that some of the properties in it's characterization were existence-entailing, then Crane thinks that we can infer that O does not possess these properties at all.

Now with this being said, what properties can nonexistent objects possess? For Crane, nonexistent objects can only possess what he calls 'pleonastic properties.' He claims that pleonastic properties are insubstantial and that they are merely properties that follow from truths about an object. So, if we were to say that 'Item a has the property f,' we can infer from this that 'There is a property f that a has.' If there is nothing more to this property than this second-order particular generalization, then we say that f is pleonastic. In addition to this fact, Crane also holds that almost all of the properties that nonexistent objects possess are representation-dependent.

This is a quite striking difference from Meinongianism. For as we know, Meinong held that nonexistent objects possess properties in a quite substantial way, in fact in exactly the same way in which existent objects possess them.

We can see at once how the account Crane provides in this book is a reductionist account, for he thinks that nearly all the truths of nonexistent objects are grounded in certain human representations. Note also that this makes his account an anti-realist one (so long as we are interpreting 'realism' as denoting mind-independence). Thus, he quite naturally construes nonexistent objects as being intentional objects. Crane understands  intentional objects to be merely those things which are represented by the human mind. As expected, such objects lack any intrinsic nature.

As a corollary of this fact, Crane also points out that nonexistent objects can never be identical or distinct on his theory. Rather, he claims that nonexistent objects can only ever be more or less similar. If a high enough degree of similarity is reached, then we are justified in saying that they are 'the same', but only as a matter of speaking.


So concludes our presentation of the theory. As we can see, it is a very interesting theory. Indeed, I would be willing to say that it is in many ways much superior to the mainstream theories on offer. But I have some serious misgivings regarding it, which in my opinion make it fall short of Meinongianism. They are as follows:

1. The pleonastic/substantial property distinction seems quite ad-hoc to me. Crane seems to take the distinction as being obvious, but to my eyes, it seems rather hastily put together for the express purpose of stripping away all substantiality from nonexistent objects. We often hear it said by critics that the nuclear/extranuclear extinction is ad-hoc, but I can perfectly well understand that distinction, and I can very easily discern whether any particular property is nuclear or extranuclear. I am not so confident that I could do the same with the pleonastic/substantial distinction.

2. The CP is a powerful epistemological tool that the Meinongian has at his disposal. So as I see it, Crane's refusal to make use of the CP only spells bad news. Instead of having a consistent and rational a priori method for discovering truths about nonexistent objects, Crane is forced to use a posteriori methods of varying reliability in order to discover such truths. Some people might actually see this as a strength, but certainly not me.

3. The theory of identity is sorely lacking. Recall that for Crane, two nonexistent objects can never be said to be identical. This leads to two massively untoward consequences. Firstly, it directly implies that we cannot count all of the nonexistent objects. Thus, we are forced into the strange position of affirming a multiplicity of nonexistent objects, while also recognizing that we cannot ever say how many of them we are giving status to in our theory. Secondly, suppose that we have two nonexistent objects, a and b, which each have exactly the same set of pleonastic properties. Since we do not have worlds-machinery at our disposal, this should be sufficient ground for affirming that a and b are the same object. But not so for Crane, for all we can really say is that a and b are sufficiently similar such that we can take them to be 'the same', in some sort of amorphous way. In my mind, this is a major strike against the theory.

4. Related to the third objection, it also follows from Crane's theory that a single nonexistent object cannot appear in multiple representations. Since we cannot ever make sense of saying that two nonexistent objects are identical (on Crane's theory), we cannot say that, for example, the Cthulhu that appears in the Lovecraft stories is the same being that appears in the August Derleth stories. But of course, we need to be able to say this, in order to make sense of the stories as given.

4. Finally, the logic that Crane founds his theory upon is Positive Free Logic. While this is no doubt an improvement upon Classical Logic, it is wholly inadequate as a tool for reasoning about nonsensical, incomplete, and inconsistent objects (which are legion in fiction).


So concludes my review. All in all, I would highly recommend this work to anybody interesting in our thought about the nonexistent. Unlike many analytical philosophers, Crane takes this challenge head-on and develops an original theory that goes a long way towards answering the central questions. I do not think the theory works in the end, but it will need to be grappled with by anyone who is interested in these issues.

Reflections on Sylvan's Last Man Argument

I would now like to offer my reflections on a classic essay, namely, "Is There a Need for a New, an Environmental Ethic?" by Richard Sylvan. This essay is very short, but it deals with a very deep issue in a profound way. Sylvan begins the essay by pointing out that an increasing number of philosophers have expressed the need for an evaluation of the ethical status of the environment, and particular of humanity's relation to it. This is where the title of the essay comes in, for it is only natural to ask ourselves the following question: Does this new evaluation require merely an extension of existing ethical systems, or does it require a new ethic altogether?

This question actually has a simple enough answer, for if we examine the dominant Western ethical systems, they all place humanity as a despot or tyrant over the rest of the natural world. Thus, if humanity is a despot over the rest of nature, then, a fortiori, humanity does not have any ethical obligations towards the rest of nature. If humanity does not have any obligations toward the rest of nature, the question of an environmental ethic is null and void right from the start. Therefore, it would seem that a true environmental ethic must of necessity be a radically new ethic.

Now that we have that out of the way, what we must consider next is the foundation of the dominant ethic. Sylvan points out that the dominant Western ethical systems have a number of guiding principles (he mentions the Golden Rule as one such example), but one plausible candidate for a 'core' principle of the Western ethic is the Liberal principal that one is free to act as one pleases, so long as one does not harm other human beings and that one is not likely to irreparably harm themselves. Of course, this principle is not as explicitly formulated as we might like, but it gives us a good place to begin our investigation. Sylvan refers to this principle as 'basic (human) chauvinism.' It is easy enough to see why this name fits, since the principle as such includes only human beings within its reach.

After having established basic (human) chauvinism as a plausible ground for the Western ethic, Sylvan continues by bringing forth some counterexamples against the Western ethic, and this is the most interesting part of the essay. He discusses four different counterexamples, but we will focus on the first one; the so-called 'last man' argument. The argument is very simple; we are tasked with imagining that there is only one human left on the planet. The human decides to kill every living plant and animal before he expires. The argument formally laid it is as follows:



PREMISE 1: If basic (human) chauvinism is correct, then what this man does is morally permissible.

PREMISE 2: But what he does is in fact morally impermissible.

CONCLUSION 1: Thus, basic (human) chauvinism is incorrect.



This is a quite simple argument. Indeed, it is just an instance of Modus Tollens and is thus valid Relevantly, Classically, and in many other systems of Logic. In addition, the argument does not crucially depend on the Reference Theory. So we need to see whether it is sound.

The first premise is quite clearly true, since it follows from the guiding principle of basic (human) chauvinism. I believe that it is the second premise that is the crux of the argument. Is it in fact not morally permitted for someone to act in the way the man has in our scenario?

Now, since I am a Vegan, I obviously think it would be impermissible for the man to kill all of the various animals on the Earth. But I believe that the most pressing question concerns whether it is impermissible for the man to kill all of the various plants on the Earth.

If you will recall my post on the variation of the Argument from Marginal Cases, my intuitive sense leads me to think that it would be impermissible for the man to do this. Indeed, the plants in the last man scenario function almost identically to the irreversibly comatose person I mentioned in my post.

Indeed, imagine that we simply changed the scenario to where there is only one last sentient man and a billion irreversibly comatose people on an automated 10 year supply of life support. Now if the 'last man' in this scenario decides to spend the rest of his life killing off all of these comatose humans would he then be acting impermissibly? I still think that he would.

But why is it impermissible, on my view? As I see it, these acts are impermissible precisely both because life has some sort of value on its own, and because the last man lacks a good reason for committing these acts.


Now, it would be unbecoming on me not to discuss the paragraph in which Sylvan suggests that the new environmental ethic can be consistent with mainstream views that couple having rights with having responsibilities. Of course, as Sylvan points out, this would imply that most sentient items would not possess rights, even if right-holding items do have responsibilities towards them. Now I reject wholeheartedly the notion that having rights entails having responsibilities; for some humans are not equipped to have responsibilities (certain mentally challenged humans, or infants, or children). Thus on this view, all of these humans would not be rights-holders. But as I see things, they are rights-holders, so the view that having rights entails having responsibilities must be false.