How does language ‘hook’ onto the world? It seems there are broadly two views, both of which can be credited to Wittgenstein, more or less. The first is the ‘picturing’ view where a statement is true if it pictures the facts of the real world. Other statements can, of course, be made, but there are true or false by virtue of other statements, and so on back to some fact or facts about the real world.

This view is usually associated with the ‘early’ Wittgenstein, that of Wittgenstein, L., Tractatus Logico-Philosophicus (New York: Cosimo, 2007 [1921]). The ‘later’ view is that language is to be seen against the backdrop of human activity and communication. On this view, there is no simple reduction of a proposition or statement to a bedrock of reality.

The logical empiricists developed the earlier view, coming to the conclusion that there are elementary propositions given by the data of sensation, and that everything else is constructed upon them. Thus, for the logical empiricist, all data and all propositions or statements are, ultimately, confirmed by experience. In general, this is entirely true. We confirm what we hear by what we see: ‘It is raining’. As a statement, is confirmed or not by looking out of the window. The connection between the real world and the world constructed by language is immediate.

The later Wittgenstein observes that most language use does not conform to this fairly simple rule. The instruction ‘give me a slab of stone’ is not, in fact, something that we can confirm by looking at the state of the world. The world is dynamic, not a static picture that can be analysed for truth or not. In fact, by our actions, we confirm or dis-confirm the statement ‘give me a slab of stone’ by our own actions in doing so or not.

This second view of language, that of Wittgenstein, L., Philosophical Investigations (Oxford: Blackwell, 2001 [1953]), can lead to the view that the human world is totally constructed by language, and reality (whatever it may be) is shuffled off to play a bit part in our experience. What is in the foreground, in this view, is the state of the world as constructed by a human.

This second view also has something going for it. Language does construct our world. A marriage, for example, is created by the saying of certain words and the doing of certain activities. In the real world, nothing much has changed; there are still two people in a building, with a bunch of other and so on. There might be extra pieces of paper, but they only, in fact, have a meaning in the context of the human construct known as marriage. A marriage certificate only has a meaning in a context where marriage is known about and accepted in a society. We create and construct marriage in that society.

Which view is right then? Does the world determine our language or does our language determine the world? The answer, of course, is neither is totally true or false. The world has to determine what we can say about it, after all. The truth of the statement ‘The Sun is shining’ depends upon whether there are clouds in the sky, whether it is day and so on. On the other hand a statement such as ‘give me a slab of stone’ depends of a variety of factors, both within the world (do I have a slab of stone nearby?) and of human intention (would I pass you a stone if I had one?). The latter is a complex question – you might, in my view, be about to drop the slab from a bridge onto a car passing underneath, and I may not be minded to equip you to do so. In this case, the response to language use is contextual.

The danger is that we start to suspect that all language has been loosed from its bonds to the real world, that everything is merely a human construct. The extremes of postmodern thought suggest that this is possible: if we refuse to think about something it does not exist. Most people live in a slightly more sensible human construct where reality does intrude rather more. But language does affect what we can think about. In this sense, George Orwell was correct when, in 1984, he suggests that by stopping people using certain words we can stop certain thoughts. If we could remove the language of dissent from common usage, for example, no-one could dissent because they could not express what they were thinking.

Meaning and use are therefore interlinked. In human terms use defines the meaning of a word. If I use the word ‘oojimaflip’ you may well not know to what I am referring. On the other hand, if I point to a piece of complex machinery, you will probably deduce that I mean that object. Perhaps more famously, the expression ‘Fuck, the fucking fucker’s fucking fucked’ does convey some information, albeit by using the same basic word in different ways (and now the blog has a 15 certificate, as well). But meanings can also be redefined by use, and by changing the use. For example, the word ‘marriage’ used to mean a sexual relationship, publicly acknowledge, between a male and a female. Recent legal and popular events have changed it to encompass relationships, publicly acknowledged, between people of the same sex. Usage has altered meaning.

Even so, given all that, there is still an underlying reality to much of our language use. This may not be obvious, of course. The logical empiricists tried very hard to get rid of metaphysical and ethical statements, as they had no underpinning in empirical fact. Thus ‘murder is bad’ was reinterpreted as ‘I don’t like murder’ and ‘God exists’ was defined as non-sense, as in not empirically verifiable, which was quickly adjusted to mean nonsense, as in not understandable.

But reality will out. Murder is a bad thing for reasons not only to do with our instinctive aversion to it. The claim that God exists persists, despite the predictions on all sides that God is dead and irrelevant anyway. Re-interpreting statements as nonsense does not make them go away, after all.