When trying to elucidate is the nature of truth; it is best to start with our intuitive conception of the concept, and see whether it stands up to critical scrutiny. Most people uncontaminated by philosophy would cash out truth as in terms of correspondence. The correspondence they would try to elucidate would be between either a thought or a statement and a state of affairs in the world. On this picture something is true if and only if the statement/thought corresponds with a state of affairs in the world.

 

            When trying to elucidate this correspondence theory of truth it is important to do so in reference to a series of examples. Some examples seem to perfectly fit the correspondence theory of truth, while other examples are more problematic. A nice clear example which seems to fit the correspondence theory of truth is the statement (1) Grass is Green. Statement (1) is true if grass is green and false if it is not. The true statement results from a correspondence between a fact and the state of affairs in the world that the statement represents. On the face of it (1) is a perfect example of a correspondence between a statement and a state of affairs in the world; hence it is a nice illustration of a particular instance of the correspondence theory of truth that we can use to see if it generalises to all supposed true statements.

 

            Obviously, however, before we proceed to test whether we can generalize from example (1) to other examples of true statements we need to first analyse the nature of the purported correspondence occurring with example (1). Now if we are arguing that the statement corresponds with a state of affairs in the world we are committed to the view that words in the sentence correspond to objects or properties in the world and that the structural features of the sentence are isomorphic with the structural features of the world.

 

            If we begin with first with structural features of our statement (1) Grass is green. On the face of it (1) involves picking out an object (grass) and predicating to the object the property of greenness. If the statement is true the then the object picked out by the statement (grass), will in fact be green hence the statement will correspond with a state of affairs in the world. Other structural variants of the sentence would be nonsensical as they wouldn’t pick out any sensible structural features of the world. Thus take (2) Is grass green. There are various different ways of interpreting (2) the most obvious one being that by moving ‘is’ to the front of the sentence (2) unlike (1) is no longer a statement but a question. On this plausible interpretation; a simple structural change to the sentence stops it from being a statement. Another possible structural change is illustrated by (3) Green is grass. Now (3) from a structural point of view is a statement. However, on the face of it, the statement seems to be senseless. There are various different ways of parsing the statement one could read it as treating ‘Green’ as an object and predicating ‘grassness’ as a property of the ‘green’. But the preceding interpretation seems barely intelligible. There are other ways of treating (3) a more plausible way of interpreting it would involve pragmatic features that would mean that green is understood in terms of a particular paradigm; ‘grass’. However, while a theory of linguistic usage of this form could be sketched it would involve considerations that would go well beyond any simple sketch of the truth of the statement being simply a correspondence with non-linguistic facts.

 

            So it could be argued that (1) is the only structural combination of the three words ‘grass’, ‘is’, and ‘green’ that makes sense as a statement; and this is because the structural features of (1) directly mirror the structural features of the world. So in the simple example of (1) we need to posit a kind of structural isomorphism between the sentence and the state of affairs in the world it picks out.

 

            This account of a mirroring relation between sentences and the states of affairs in the world they pick out works well for a variety of different sentences. A sentence such as (5) The cat is on the mat. Is a perfect exemplar of a statement having structural features in common with states of affairs in the world. The examples I have given are extremely simple ones, and obviously any attempt to generalize the structural isomorphism between syntax and structural features of reality in true sentences will raise a lot of technical difficulties when discussing more complex syntactic structures. However, for now, let us stick with our simple example (1) and discuss it in relation to relation to the semantic features of the sentence; what sense we can make of explicating the truth of the sentence in terms of correspondence.

 

            As I said above (1) would be a clear example of a true sentence for a person uncontaminated by philosophy. It is a reasonable conjecture that most people would think that (1) is true because it corresponds with a particular state of affairs in the world. Of course a moment of reflection will show that things are not as simple as they seem. Firstly, statement (1) ‘Grass is green.’ is a general statement, so its truth value cannot be cashed out by a single state of affairs. In (1) the word ‘Grass’ isn’t being used to pick out a particular blade of grass. Nor is ‘Grass’ being used to pick out a particular patch of grass. In (1) ‘Grass’ is being used to pick out a particular category of things in the world that we use the label ‘Grass’ to describe. When it comes to ‘Green’ it isn’t referring to a particular green thing in the world; rather it is describing a property which different things in the world have or don’t have. The word ‘is’ should be treated as a relational predicate that links the property with the object in the sentence.

 

            With the above little bit of complexity acknowledged we can no longer simply assert that (1) is true because it corresponds with a particular state of affairs in the world. If we consider ‘Grass’ as a general term that picks out a type of entity in the world, and not a particular entity, then our analysis becomes more complex. When we are speaking of what a word refers to we run the risk of presenting a misleading picture of what reference entails. As has long been noted in the philosophical literature (see Strawson 1971), words don’t refer; rather words are used by people to refer to things. So when we are discussing a reference relation between a word and an object in the world we need to keep in mind the context the word is used in.

 

            In trying to understand the reference of the word grass ‘Grass’, we will obviously need to have a theory as to the nature of reference. When thinking about reference it is best to begin with concrete acts of reference in a shared world of experience; and building up a theory of how we refer to abstract entities such as numbers, sets, etc. after we have accounted for simpler forms of reference.

 

            Reference is typically a social skill and involves communication between at least two parties about a shared world of experience[1]. Reference is possible without linguistic communication. A non-linguistic example of reference is pointing. The ability to interpret pointing is a human universal[2], but it is not shared by most other animals. Non-human primates, don’t use pointing as a tool to refer to the non-human world. In fact few animals are proficient at interpreting pointing. An exception to the rule that most non-human animals have trouble interpreting pointing; is domestic dogs. Wolves who domestic dogs evolved from cannot typically interpret pointing; but domestic dogs have somehow evolved the skill. Domestic dogs were primarily selected for by humans for their social abilities. It is probably no coincidence that domestic dogs who were selected for empathy with the humans they live with, have developed the capacity to interpret human gaze direction and pointing. However, an analysis of the evolution of domestic dog’s capacity to interpret gaze direction and pointing is beyond the scope of this paper.

 

            Normally developing humans have the capacity to interpret the eye gaze of their fellow humans as directed towards objects in their shared environment, and to interpret our fellow humans pointing as referring to objects in our shared environments. In order to interpret eye gaze or pointing as indicating an object in a shared world of experience a creature needs what anthropologist Tomasello calls shared intentionality. At a minimum the humans need to view each other as agents with a point of view about the world and as agents who may want to communicate useful information about the world to another agent.

 

            Animals who have the capacity for shared intentionality and the capacity to interpret eye gaze, and pointing; have the rudiments of reference in place. They can refer to objects in the mind independent world and can judge whether the person they are communicating is interpreting the reference correctly or not. So we can now see that with triangulation on a shared object of experience, comes the capacity for reference, and for correct or incorrect interpretations (Davidson 2001). Furthermore, the ability to use words, though more complex than bare pointing, still relies heavily on the foundation of shared intentionality and pointing. 

 

            So with two primitive humans who are interacting with each other in a particular environment; if one human points to grass in the environment, the other human can judge, either correctly or incorrectly, that that is what his partner was referring to. Of course without the linguistic abilities these creatures will remain at a primitive stage of communication and they won’t have the capacity to make statements about the aspect of reality they are referring to.

 

            In the philosophy of language there is a debate in the theory of meaning that is divided into two main camps: the descriptive theory of meaning (Frege, Russell, the Later Wittgenstein), and the direct reference theory of meaning (Kripke, Putnam). Roughly speaking the descriptive theory of meaning argues that our words manage to refer by picking out an object in the world via a description (or a cluster of descriptions). While the direct reference theorist argues that we pick out objects directly and our words keep their meaning by tracking the same object throughout time.

 

            The debate between descriptive theorists and direct reference theorists is at an impasse and at the moment amounts to nothing more than an appeal to competing intuitions. However there has been some excellent work in perceptual psychology and philosophy (see Fodor and Pylyshyn 2013) has provided a good perceptual grounding for direct reference theorists[3]. In this piece I am going to work within a direct reference theory of meaning and causal theory of meaning. Though given that the debate between descriptive theories of meaning and direct reference theorists hasn’t been decided any conclusions I make will be vulnerable to the outcome of this debate on the theory of meaning. Such is life.

 

            So in our (very) conjectural story truth enters the scene with triangulation on shared objects of experience between agents communicating with each other and judging whether the other has interpreted the others pointing correctly or not[4]. Our similar perceptual apparatus, and similar embodied nature, as well as our emphatic understanding of each other[5] makes this triangulation mostly successful.

 

            In going from triangulation using pointing and eye gaze to the use of words and the combining of these words with syntax we move into deep waters. In studies of the evolution of language there is lots of data and competing theories but no theory which stands out as the obviously correct one. Tomasello argues that language evolved from gestures and that spoken words came later. Everett argues that speech and gesture evolved together. While Chomsky almost entirely ignores gesture and traces the origins of language to a mutation which gave us syntactic structure. The dates of when language evolved are again radically different depending on the theorist; Everett places the evolution of language at about 1.5 million years ago, Tomasello argues that it evolved 200,000 years ago, while Chomsky argues that language evolved 50,000 years ago because of a random mutation.

 

            Obviously any discussion of how language evolved cannot be undertaken in this paper, as to discuss the issue in the detail, would require at least a book length treatment. And as our topic here is on the nature of truth and not the evolution of language we need to keep our focus on the salient issues and avoid getting bogged down into a morass of irrelevant detail.

 

            For our purposes we need only note that the next step that our primitive human must make beyond pointing; is to be able to pick out an object, and say something about it. This would involve our primitive man developing subject predicate structure. So for simplicity sake let us pretend that our primitive man is speaking proto-English. He can point to objects, can name the objects and say things about them. At this point our primitive man will have the capacity to say ‘Grass is green’ and the members of his tribe will be able to judge whether what he has said was correct or incorrect.

 

            Based on our causal theory of meaning we will assume that grass was originally picked out because of its bare perceptual features and given a name which was passed on to other members of the community. In this sense grass will have an extension that is recognised by most members of the community; based on the perceptual features of the environment. Presumably one of the key perceptual features of grass as-well as its shape will be its colour. So it will be partly built into the concept of grass that it is green. So most people who understand the concept of grass will agree that it is green; it will be part of the concept of grass that it is green.

 

            However, despite the fact that greenness will be one of the perceptual features used to pick out grass, it would be a mistake to assume that the greenness of grass is a conceptual truth. It is an obvious fact of experience that grass isn’t always green. During hot weather grass can turn brown and yellow. The greenness of grass may be a typical perceptual feature of grass but something can be grass and not be green.

 

            So when the general statement (1) “Grass is green” is asserted; the response won’t be true of false but rather a sensible person will reply that it is typically green though not always. The preceding response won’t rely on a simple correspondence between a statement and a state of affairs but will rather involve an averaging many perceptual experiences.

 

            Above we were speaking about a person who was answering the question based on their own casual experiences interacting with the world on a daily basis. However it is important to note that we have more precise ways of classifying the world and the entities it contains; than casual induction. Scientific analyses often categorise things in more precise ways than our ordinary perceptual judgements do. So, for example, while people may classify Whales as Fish, or the Moon as a Planet, scientific analysis divided up the world in a more precise manner using an intricate network of theory. Hillary Putnam calls this the division of linguistic labour where we use our concepts based on loose perceptual features but rely on scientific analysis to give the extensions of our concepts in a more finely grained manner. However, when it comes to a scientific analysis of something like grass, simple correspondence between statement and fact becomes less plausible. A scientific theory is a interconnected network of theoretical and empirical facts which are used to predict and control the data of experience. A scientific analysis of something like grass will not involve a simple statement that corresponds with non-linguistic facts; rather it will be a network of statements that are connected to experience only at the periphery (Quine 1951).

 

            We saw above that our ordinary language interpretation of (1) “Grass is green”, doesn’t yield to an unproblematic correspondence between statement and fact. Neither do we get an unproblematic correspondence between statement and fact when we resort to a scientific interpretation of (1). However it could be argued that (1) would yield to an unproblematic correspondence if we add the demonstrative ‘this’ to (1) to yield (6) This grass is green.

 

            The use of the demonstrative ‘this’ in (6) seems to give us a simple correspondence between statement and mind independent fact. We no longer have to worry about yellow grass and we can simply assert correctly that the grass in front of us is green. However, even here we run into some problems when we try to cash out the statement in terms of correspondence with mind independent fact. As it is unclear that green is in fact a property that exists in the mind independent world.

 

            As every school child knows, the standard scientific theory of colours are that they are a secondary quality that are created as a result of light reflecting off objects and hitting our retinas, resulting in information being translated along our neurons until the hit the occipital lobe in our brains, which results (nobody knows how) in our conscious experience of colour. So it is unclear whether we can cash out the greenness of grass in terms of simple correspondence with mind independent facts[6]. So even with the simple sentences like “This grass is green” cannot be explicated in terms of simple correspondence between statement and fact.

 

            It could be argued that I am by focusing on a secondary quality like colour I am making things too hard for those who are pushing for a correspondence theory of truth. However, this interpretation is incorrect. As Berkeley showed three hundred years ago even primary qualities don’t yield unproblematic access to a mind independent world.

 

            In this piece I tried to begin with a simple statement and explicate it in terms of a correspondence theory of truth. My plan was to begin with a simple statement demonstrate the correspondence there before moving on to more complex statements. However, I discovered that even with the simple statements an analysis in terms of correspondence proved impossible. There no generalisation from simple statements to more complex statements was possible. In the next blog-post I will discuss the coherence theory of truth and see if it fares any better as a theory than the correspondence theory did.