John Hospers' book "An Introduction to Philosophical Analysis" has chapters on various different philosophical themes. Chapter 1 is on language and it is this chapter this essay examines. I shall argue that while Hopsers' arguments provide a useful framework for analysing our various uses of language and rebutting the more primitive theories of language, nevertheless there are some unquestioned assumptions he makes that enables us to go beyond his conclusions.

Hospers first looks at the relation of words to things. He points out that there is no necessary relation between a name and its object, "a rose by any other name would smell as sweet". This seems like an obvious point, but in fact is at the root of a number of differences of opinion. Hopsers gives the example of a group of people arguing over the following problem. A hunter is trying to catch a squirrel which is on the side of a tree, the hunter is on the other side of the tree to the squirrel and as the hunter moves around the tree the squirrel moves round too, always keeping the tree between himself and the hunter. The hunter runs round and round the tree but always the squirrel keeps running just quickly enough to stay on the opposite side of the tree to the hunter. The question in dispute was, does the hunter go around the squirrel ? Hopsers says this isn't a real dispute but simply one of how you define "go round". If you define it by saying the hunter goes round the squirrel if he first goes above, beside, beneath, beside and above again (eg north,east,south,west,north) then the hunter does go round the squirrel, however if you define it as facing the squirrel, then moving round to the back of the squirrel then continuing round to the front of the squirrel again the hunter has not gone around the squirrel because the squirrel was always facing the hunter. Its all a question of how you define your terms. Hospers gives another example:

"If someone said to you "There aren't really any material objects in the world - there are only spirits," you might be surprised at this "information" then skeptical, then inclined to deny it vehemently. But your surprise would vanish if you found that he was using the word "spirits" in such a broad way as to include trees, houses, planets and so on - the very things to which common usage already assigns the phrase "material objects". Often people, perhaps without knowing what they are doing, will flaunt one of their assertions as a new discovery about the universe, whereas in reality they are merely manipulating words and employing them in violation of common usage without informing the hearers of the fact." (p.9)

There is no law that prevents you from using the noise "lamp" to refer to any object you like. However you will almost certainly be misunderstood if you move outside the common usage and hence if you wish to communicate you must stick to established convention.

Some words, of course, are not so simple to define as "chair" or "lamp" and hence many disputes are simply arguments about what we mean by a term such as "God", "society", "moral" etc.

Hence Hospers first point is that words are simply labels for things, like bottles or containers, and have nothing to do with the objects themselves, we may give these objects any "names" we like but if we wish to be understood we must clearly define any problematic terms we are going to use.

Next Hospers moves onto classifications. There are millions of different objects in the world and if each object had a different name it would be very difficult to communicate. Hence we distinguish between proper names such as calling your parrot "Polly" or your dog "Rover" and class words such as parrot and dog and each of these words is used to stand for many particular things. But having a word which stands for many particular things brings about difficulties. If I use the term "human being" must I list every human being to define the word ? No, I simply have to show some characteristic or group of characteristics which they all have in common. As people become aware of resemblences among objects they give them group names. The fact that various languages do not have equivalent class names shows how many ways people can verbalise their experiences. The disadvantage of class words is that they tend to make us forget the differences that still exist between members of the same group, we fossilize our conception of an ever changing world. Another disadvantage is ambiguity, "we went to look at the construction" way mean we went to look at something being constructed or something already constructed. Ordinarily the word "sound" is not misleading, but when someone asks "if a tree falls in a forest and nobody is there to hear it, is there a sound ?" it is misleading because the answer depends whether you mean sound in the sense of sound-waves (the answer would be yes) or auditory sensation (the answer would be no).

Hospers also points out that figurative language is also ambiguous and potentially misleading. For example there is nothing wrong with saying a student is sharp, even though we usually use sharp in the sense of a sharp knife because we can explain what we mean in a non-figurative way: he is able to comprehend new material quickly. However what does someone mean when they say the entire world is but a shadow ? Can they explain this in a non-figurative way ?

Hospers then moves on to look at definitions. He points out we began by saying words stood for things, but also saying class words stand for characteristics, which is correct ? He explains the solution by distinguishing between essential properties and accidential ones. For example a triangle may be large or small, scalene or isosceles and still be a triangle, so these are accidental properties, but a triangle cannot have two or four sides and still be a triangle, so having three sides is a necessary property of a triangle. Hospers shows that mathematics gives us very simple definitions, and that real life examples are usually a lot harder to work out, for example what are the essential properties of a zebra or a elephant ? Most people would quickly say the zebra has stripes and the elephant a trunk, but suppose you were able to remove the stripes from a zebra or the trunk from an elephant, would they stop being a zebra or an elephant ? The solution to this is to recognise perhaps there is no single essential peoperty, but rather that there are a number of properties, most of which will occur in the object but not necessarily all of them. Many words have this vague definition whereby all we can say is that several properties will be present in a denotated object and it would be misleading to try to make the definition more exact than that (for example defining "neurotic" or "murder" will always have verious qualifications, special cases and varying characteristics).

Hospers gives two simple examples of why we have vague definitions for many words. Firstly he points out that in practise when we look at a dog or a cat or a fox all we can say is that they "look different", we don't automatically verbalise our sensory impressions into perfect descriptions and so can be stuck with an implicit recognition of difference without being able to say explicitly where the difference lies. Secondly there is the problem of the continuum. Suppose we ask does Colchester lie between Ipswich and London ? Since you travel down the A12 to get to London go past Colchester I'd say yes. The suppose you take a village a few miles further away from the line between Ipswich and London, but near Colchester, and asked if that was between Ipswich and London. Since you said Colchester was then you'd probably say the village was. And suppose you continue this, moving a few miles further away each time, it seems random to say, this village/town is between Ipswich and London while another village/town a couple of miles down the road isn't. Yet continuing like this you end up saying Newcastle is between Ipswich and London just because you can't find THE point at which a position stops being between Ipswich and London ! There are various example of this random stopping point, why should someone who gets 60 in a test pass and someone who gets 59 fail ? We have to be aware of definitions which have a hazy area in which its difficult to say which term (between, near, large etc) applies.

Hospers finishes the third section on definitions by an interesting discussion on words defining words. He starts by showing that most people are happy with definitions such as "X is the brother of Y means X is a male having the same partents as Y", which is called a contextual definition, or a definition of "Bird" was sparrows, hens, starlings etc, which is a denotational definition. However we are simply defining words in terms of other words, are there some words which we don't define in terms of other words ? In other words, how do words hook onto the world ? Non-verbal definitions are called ostensive definitions, and are where you show or confront someone with the object, for example showing someone a beech tree would be giving them an ostensive definition of a beech tree.

Hospers concludes the chapter with a section on sentenses and propositions. We don't use words by themselves, but in a certain order, in a sentence. A sentence describes a state-of-affairs, for example the book is on the table. But the sentence can stand for a possible state of affairs, not just an actual state. The sentence, or proposition, can be either true or false, depending upon whether it stands for an actual state-of-affairs or not; the state-of-affairs itself cannot be true or false, a table simply is or is not, it cannot be said to be true or false.

Non-assertive sentences. Not all sentences assert something: questions, imperitives and exclamations may not assert anything but still be meaningful. Some sentences have a emotive meaning, "what dress is gorgeous" has little cognitive meaning but clearly has emotive meaning. Another group is persusaive definitions, for example suppose someone says "true culture is acquaintance with science and technology" they are using the emotive meaning attached to "culture" but shifting the cognitive meaning, the new definition is based on the emotively persuasive character of the word. Other things to watch out for are word-pictures, for example "his is a cold, austere philosophy" creates a mental picture that may make the object described more or less attractive; argument from analogy (politics is like riding on a bike, don't lean too much to the left or right; those who stay in the middle of the road get run over etc).

Hence to conclude our summary of Hospers' position, he sees the function of language as primarily representational, where the word/noise-object relation becomes the "cornerstone" of what language is. From this basis the rest of the theory follows: the individual objects are grouped together with group words each with their own definition, these words are combined together in sentences, and the complex meanings generated through sentences enable figurative, emotional and analagous language forms.

The first point we shall examine is the alternative theories of language Hospers presents us with. Typically of much of bourgeois-western thought Hospers presents the alternatives as a natural theory or an individual theory. The natural theory of language corresponds to the metaphysical- religious view where questions about the good, catagories, meaning, cause etc are all seen as concerning objectively existing entities. Hospers criticises this view as mistakenly imagining that there is a "real" or "true" meaning to a word which we have to discover. He argues against this by claiming that man simply invented noises to correspond to objects, and hence any metaphysical problem may be resolved by clearly defining the terms in dispute and mutually agreeing on a suitable definition. Thus Hospers replaces a natural explanation by an individualistic explanation whereby some ideal human subject chooses words to use. Clearly this is a false set of alternatives since it ignores the social as a catagory of explanation.

Hospers often shows that words could have many meanings, for example the term "solid" could be defined in a scientific way, so as to say "nothing is solid" but he then argues against this by saying such a definition would not be practical. The crucial question to ask is "practical for whom ?", thus we come back to the classic Marxist question "for whom ? against whom ?". When man looks at the world in terms of tools to sustain him, we understand he is seeing the world in a certain way - dividing it up into use-value. He asks of objects "is this food ?", "is this a weapon ?", "is this a shelter ?" etc. However this in itself does not explain language because why should primitive man wish to say what he is doing ? Only in a more developed society where symbolic exchange is necessary (for example asking for bread, asking someone to fix your plough, asking someone for three eggs etc) does language become a requirement. Thus language is an expression of a society, and in society there are power structures and power relations. Should we not then find these also represented in language ? Hospers problem lay in seeking to give an individual explanation to language, and having found the "simple individual origin" to preceed from this to construct his whole theory of language. Instead a more correct explanation sees language as a product of society and thus rather than beginning with nouns, or proper names or ostensive definitions it begins as an expression of a society, as a totality dividing the world up according to the dominant set of values in that society. Thus leaders positions have to be validated and workers positions justified, and because any society prefers the compliance of the dominated rather than forcable subjection. Thus metaphor is inherent in language because it represents the power structures of a society. Think about what a child's first words are "Mummy" or "Daddy". These are apparently recognition words, but note the complex set of power relations they represent. To call someone "Mummy" or "Daddy" is to recognise they have certain rights and privilages, but these rights are not inherent, they have to be communicated to the child (or back to the real example, the rights of the leader have to be understood by the people, and they must accede to these rights). How does the leader explain these rights that are not natural but social ? He must explain them in natural terms, as if they were a part of the natural world, he must go back to the natural world and construct the meaning of his social position in terms of the natural world. The leader may take various titles to link him with the natural world: bringer of health, light, food etc or the leader is like an animal: a lion, a wolf, a bear or the leader might be a Father and the tribe his children etc.

To give certain jobs male titles or to tell tales about how different parts of society came to be (the roles of the women for example) language has to be useful to that society, not just to one individual person. In keeping the explanations of language individual all the power relations that language preserves, justifies, validates etc are ignored. Why should a woman choose to give a job a male title (eg chairman etc), why is that useful to her ? The notion of an ideal individual who chooses the names of words for the universal use of mankind is a myth, the words were chosen by a certain class, to be useful for that class (the dominant ideas in any society are the ideas of the dominant class). If anyone chose the term "chairman" it must of course have been a man, to confirm the idea that the woman was not to work as men work but to conform to a certain female stereotype.

Similarly were does the use of animal names as a term of abuse come from ? To call someone a pig or a cow implies that pigs and cows are very bad and horrible, and thus that they have no rights. Thus animalist language backs up the injust and horrific treatment of animals, in the same way the use of the term "Jew" to denote someone who is miserly backs up the idea that Jews don't deserve the same rights as "normal" (ie non-Jewish) people because they are morally inferior. Words come into being not through some cool individual choice but rather they are pulled, twisted, cut and knocked into use by the dominant literate group. When language was developed it was the ruling class who read and produced language, the working class was denied access to the means of literary production. We can be sure that language is functional enough to let us work and obey orders, but there are no guarentees that it will tell us the truth about the world.

It is for this reason that some philosophers have moved on from simply analysing words and their meanings and come to recognise that language itself has an underlying metaphysical theory based on the primacy of identity, presence, truth and singularity. This has resulted in the idea of single truths, unique revelations of the truth of an object (eg to see it is to know the truth of it), and thus the exclusion of all that is different, whose truth is differed, with no multiplicity or complexity. Yet science itself "reveals" that the truth is not present and revealed in the way metaphysicians believed, but rather certain aspects of the "truth" may be seen through certain languages, bringing parts to light, but on the understanding that other parts necessarily remain hidden. Similarly we recognise (for example through the structural interpretation of language that we can only define words in terms of words) that in order to have some identity, we must define it in terms of those objects that do not have an identity, (ie not this, not this) and thus the object we sought to make primary cannot be first because we need other objects in order to define it.

Heidegger tried to radically analyse language and western metaphysics in this manner. He recognised that if we define Being as that which appears to us, we fail to get to the complexity of language, because language only allows certain aspects of "what there is" to come to light (ie through various activities such as working, eating etc), thus we can only "see" what there is when we are in a certain "mode" and since we cannot be in all modes at once we can never say that we finally come to know all there is that could "stand out", thus we cannot finally come to a final knowledge of Being as the metaphysicans with their erroneous concepts picked up through language thought we could. Yet Heidegger didn't persue this line to analyse the political implications of the understanding of possibilities and the way power structures in society hide and distort "Being", thus we can see both that whilst we may never exhaust the possibilities of Being, nevertheless we can pull down the distortions and injustices perpetuated in language and help to let the scales fall from our eyes and see that language is a flawed tool with which to think with.