Fictional Objects

In the previous post, I discussed the analytic-synthetic distinction with a view of approaching, to my mind, one of the most fundamental problems in philosophy, the question of how we can make true or at least meaningful statements about fictional objects. The previous post wasn't perfect. For instance I suggested that the proposition "Pluto is a planet" is a false analytic proposition although it would in fact be better described as a false synthetic proposition. Aside from this slight unclarity, the previous post provides some good groundwork for today's post. Readers may well ask – what does the analytic-synthetic distinction have to do with fictional objects? In today's post, I wish to move a little closer to an answer.

I am interested in fictional objects because my training is in the study and interpretation of literature. Suppose we wish to write an essay about the Sherlock Holmes stories. We want to do two things – we want to paraphrase some of the stories and move towards deeper meanings. We want the statement, "Sherlock Holmes is the world's greatest detective and lives at Baker Street," to be true, at least provisionally; we want moreover to say that the stories say something about rationality and the powers of human deduction to solve problems. We want literary interpretation to be something capable of making true claims about what stories say, and we wish stories themselves to at least sometimes assert true claims. However, there is a thread that has run through philosophy, from Hume to Frege to today, that says that a proposition like "Sherlock Holmes is a detective" is either false or meaningless by reason of the apparent fact that it is making a claim about a nonexistent entity. Because literary interpretation is, in a sense, parasitic on literature, to accept this view would also be to accept that all literary criticism is either false or meaningless. Obviously, this is a conclusion I cannot countenance.

The best way, in fact the essential, only, way, to approach the issue of fictional objects is through a kind of story, a true story. Bear with me.

I wake up in the morning from a nightmare about vampires. In the dream, I believed the vampires to be real but when I wake up I ascribe to them a different ontological status. The vampires were unreal. I wouldn't apply the adjective fictional to the vampires because they seemed real at the time and we almost always know that fictional beings are fictional, but I might use the words illusory, imaginary or dreamlike to describe them after I have fully woken and shrugged off the fog of sleep. I climb out of bed and make myself breakfast. The bed, the bread, the butter, the utensils, and all the other objects in my apartment are real, existent. I pick up the newspaper and read articles about Brazilian president Jair Bolsonaro and convicted murderer Mark Lundy. I have never met either man and have never had any personal direct acquaintance with either – rather, my knowledge about them is based on written accounts or testimonies by journalists and the occasional photo. Nevertheless, I believe both Bolsonaro and Lundy to be real, existent; they are real objects existing in my cognitive environment. I know them to be real because I put my trust in the newspaper journalists to tell the truth about events that have happened or are happening out there in the real world. I pick up The Hitchhiker's Guide to the Galaxy and read a little about Arthur Dent. Arthur Dent is also an object in my world and, like Bosonaro and Lundy, I know about him only through a written account. However, I do not regard Dent as real, rather I regard him as fictional. I know that he is fictional because he is the protagonist of a novel – there are multiple signs, both internal to the novel and external to it, that point to the fact of him being fictional. I then read a little from a recent translation of Cicero. Cicero is neither fictional nor existent; rather he is someone who died (and passed out of existence) more than two-thousand years ago. Nevertheless he is an object in my world because I can talk meaningfully about him with others. Having read a couple of passages from Cicero, I walk to the library. On the way I pass by a number of other human beings – I know nothing about them accept how they present themselves physically to my gaze but I regard them as real, existent. The library has two floors, the ground floor being devoted to fiction and the first floor to non-fiction. I assume that all the books held on the ground floor concern fictional objects while all the books on the second floor (with the possible exceptions of literary criticism and philosophy) concern objects that exist now, have existed in the past or are likely to exist in the future. I walk over to the graphic novels section and grab a collection of Sandman stories. Most of the characters in this comic book series, such as Morpheus, Death, and Delirium, are also objects in my world, objects in fact as important to me as Arthur Dent, but, like Dent, they are all also fictional.

This little story presents a kind of quasi-phenomenological perspective on how we engage with the world. It is not orthodox phenomenology. Orthodox phenomenology, as pioneered by Husserl, seeks out the universal characteristics of how we experience the world, brackets the individual out, but my little story is intentionally quite personal, autobiographical. Moreover, orthodox phenomenology brackets out the issue of existence (this is known as epoche) and focusses purely on experience. My intention is quite different. I am trying to show the variety of ontological categories into which we place the various objects that occupy our cognitive environments. Some objects in our world we know to be real, existent, because we have first hand acquaintance with them. Examples in my own experience include my bed, fridge, bread, butter, and my mother when I go to visit her. Other objects, such as Bolsonaro, Lundy, and Cicero (and for that matter China) I know about only through written and spoken reports, through hearsay – yet I know these objects to be existent because I place my faith in the honesty and integrity of the people who tell me about them. I choose not to believe the National Enquirer when it reports on sightings of Bigfoot. And then there are objects which I always know to be fictional, such as Arthur Dent, Sherlock Holmes, Garfield, Hamlet, and Scooby Doo. Even though these objects are fictional, and I know them to be fictional, they form part of my cognitive environment and I can talk meaningfully about them with others.

This perspective is at once obvious and surprising. What is surprising is the ease with which we categorise different types of object into different ontological compartments When we look at the world in this way, as a world containing both factual and fictional objects, it can seem mysterious the process by which we effortlessly divide the world up, into things that once existed, exist now, might exist, and have never and will never exist (and are thus either fictional like Dent or fallacious like Bigfoot). Yet children seem capable of making this distinction almost from the time they learn to talk. A four-year old child who is read Dr Seuss in bed by his parents almost certainly knows that the Lorax and the Grinch are imaginary, although he might not yet know this word. As soon as children begin to play, they play at make-believe, pretending to be superheroes or cops or Red Indians. The ability to play, to discriminate between fictional objects and non-fictional objects, comes naturally to children. I want to digress a little. Phenomenologists like Husserl thought there was something mysterious about why we believe other people have minds. Husserl's explanation is that a person compares the behaviour and physical form of the other to his or her own behaviour and physical form, and infers that the other has a mind as result. In reality, a child learns that other people around him have minds before the child realises that she has a mind herself. A four year old child never doubts that his or her parents have minds. If anything, children go too far in the opposite direction, ascribing conscious volition to natural processes. Similarly the distinction between fact and fiction comes naturally to children.

There is an important point I need to make before we move on. Although fictions primarily concern fictional objects, they contain existent objects as well. Sherlock Holmes lives in London, and London was and is a real place. Arthur Dent is obsessed with tea, and tea is an existent beverage. The Sandman sometimes features people who once lived, such as Augustus and Shakespeare. Moreover, fictions like Romeo and Juliet and Othello can tell us true things about love and jealousy. This fascinating topic, the way fictions mix fact and fantasy, is important, but is not something I wish to explore in this post.

The world-view I am presenting is one in which we live and move within a cognitive environment that contains fictional as well as existent objects. This leads to the conclusion that existence is a property that objects either have or don't have. The name for this position is Meinongianism, an ontological and logical system invented by Alexius Meinong and criticised by Bertrand Russell among others. Many logicians before, during and after the time of Meinong found the idea that existence might be a property or predicate intellectually or logically unpalatable (I think the first to say that existence isn't a property was Kant). In my opinion, however, I don't think Meinong went too far but rather that he didn't go far enough. Meinong argued, I understand, at least early in his career, that there is an object for every property or set of properties. I believe, by contrast, that all objects exist in the mind or world of a conscious knower. Meinong's position tends a little toward the Platonic world-view I criticised in the post immediately preceding this one– it suggests a realm of Platonic Ideas divorced from reality. I believe, by contrast, that objects are created in a conscious mind as a result of perceptions of the world and of verbal communications with others. If we adopt the position I am advocating, a number of paradoxes associated with conventional Meinongianism, paradoxes first proposed by Russell, evaporate.

For instance, Meinong argues that there are objects that each only have one property. This leads to the following paradox. "1. There is an object, x, which has the sole property of being blue. 2. The object x has the additional property of only having one property. 3. Therefore the object x has only one property and has at least two properties." This paradox can be avoided if we suppose that Meinong was wrong in postulating that there can be objects that only have one property, if we suppose that objects are created in the mind of a knower. Suppose a friend tells me, "There is an object that is blue." I have no reason not to believe her and consequently constitute such a blue object in my mind. Suppose my friend says, "There is an object that is blue and this is its sole property." I can choose not to believe her on the grounds that it is logically impossible for something to have both only one and two properties. If my friend says, "There is an object that has only one property," I can again choose not to believe her on the grounds that it is logically incoherent to presume the existence of objects that only have one property each. If we accept the position that objects exist in the minds of conscious knowers, all objects, existent and nonexistent alike, have a potentially infinite number of properties, as many as I or others wish to grant them.

According to Meinong, there is a sharp distinction between existent and non-existent, Meinongian, objects. Existent objects have infinitely many properties while Meinongian objects only have a finite number of properties. I think Meinong was wrong to propose this sharp distinction. I know that Arthur Dent is fictional because he is talked about by novelist Douglas Adams in a novel; I know that Bolsonaro exists because he is talked about by journalists in the newspaper. We base our determination of whether an object is existent or nonexistent on circumstantial evidence. There is no sharp distinction between the two types of object. Suppose a friend says to me, "I know a chap called Kevin who went to prison for tax evasion." In my mind, at first, there are only a few properties I can assign to Kevin: that a person called Kevin exists, that he went to prison, and that the reason he went to prison was for tax evasion. However, I can and do infer additional properties. I can infer that Kevin is a man, that he has two arms and two legs, that he has a heart, a brain, a liver and kidneys. I can presume that my friend knows Kevin and has a first-hand acquaintance with him. And so on. There are also things I don't know about Kevin. I don't know where he lives, if he is married, what his job was, how long he was imprisoned, if he is currently in prison, and so on. My understanding of Kevin will improve if my friend tells me more about him. I can add properties to the mental object "Kevin who went to prison for tax evasion" as a result of what my friend tell me about him. Now suppose that my friend says, "In the novel The Hitchhiker's Guide to the Galaxy, Arthur Dent escapes Earth just before it is destroyed by the Vogons to make way for an interstellar bypass". At first, there are only a few properties I can assign to Dent: that he is a fictional character, that he appears in a novel called The Hitchhiker's Guide to the Galaxy, that in the novel Earth is destroyed by entities called Vogons and that Dent escapes this fate. However, I can and do infer additional properties. I can infer that (in the novel) Dent is a man, has two arms and two legs, a heart, brain, liver, and kidneys. As my friend tells me more about Dent I can ascribe more properties to the Arthur-Dent-object in my mind, such as the property that his travelling companion is an alien called Ford Prefect, and the property that Dent likes tea. Obviously, this drastic revision of Meiongianism requires further work. But the fundamental idea I am proposing, here, is that existent and nonexistent objects alike both have a potentially infinite number of properties, as many as we assign to them.

Russell argued that Meinongianism leads to a second paradox. It seems that, if existence is a property, we can say "There is an existent golden mountain" or "Pegasus is an existent flying horse." In both cases, we are asserting that an object exists that we empirically know does not, yet it seems to have the property "existence". If however we assume that both "the golden mountain" and "Pegasus" are objects created in the mind of a conscious knower, we can confidently expect most people to deny that either proposition is true. Ordinary people, if you ask them, will tend to subscribe to the proposition "Sherlock Holmes lives on Baker Street" and deny the proposition "Sherlock Holmes exists". For most people, a property of "Sherlock Holmes" is that he is fictional and thus nonexistent. Now consider Bigfoot instead. In my mind, Bigfoot does not have the property of existence. I might have a friend however who believes in Bigfoot. In her mind, Bigfoot does have the property of existence. Which one of us is right? We might say, with Russell, that it is an empirical fact that Bigfoot does not exist but, in fact, neither I nor my friend can be absolutely sure. We can only say that the proposition "Bigfoot exists" is false for me and true for her. It might seem that I am now advocating an extreme form of relativism, the notion that all truth is subjective, but in fact my position is more subtle and complex. I think truth is arrived at dialectically, discursively, through conversation and negotiation with others and with the world. It is difficult for me to plainly and distinctly delineate this position and so I point to the work of Richard Rorty instead for a better description of it.

The reader may wonder what all this has to do with the analytic-synthetic distinction I talked about in the previous post. What I am saying is that objects are created in the minds of conscious beings, either through observations of the world or through verbal reports by other people. A verbal report can be considered a kind of definition, and a definition is a kind of speech-act. A work of fiction is an extended speech-act. When we first pick up The Hobbit, we know nothing about it except that it concerns a creature called a hobbit. As we read the book, Tolkien defines the particular hobbit who is the hero of the story, fleshes out this abstract object with more and more properties – the hobbit in question is called Bilbo Baggins, lives in a hole in a ground, a very comfortable hole, hates the idea of adventures, embarks on an adventure anyway with thirteen dwarves, finds a magic ring deep under the Misty Mountains, converses with a dragon... and so on and so forth. Tolkien creates the object Bilbo Baggins in the mind of the reader. It is important to recognise that although the mind of every conscious being is different, we tend to agree about Bilbo Baggins's properties even though there is no existent object in the world denoted by the term "Bilbo Baggins". This is a reason why I reject subjectivism.

There is much more I could say on this topic but I will end this post with just some final thoughts. The approach to fictional objects I am advocating is known as creationism – the term creationism has a different sense in the rarefied world of philosophy than it does in the popular imagination where it denotes the fundamentalist religious view that the Earth was created by God some six thousand years ago. The creationism I am talking about is something quite different. For an interesting essay about nonexistent objects, see the article on them in the Stanford Encyclopaedia of Philosophy (https://plato.stanford.edu/entries/nonexistent-objects/).