INTRODUCTION

You are being asked to work through this booklet as part of a course in the philosophy of education, itself part of your course in education. Why? In the first place, I hope it will be useful for you, not just in coping with your education courses, but also more widely in your everyday life. In the second place, it is, I think, a reasonable, if somewhat unusual, component of a course in the philosophy of education, especially when you won't be getting any other kind of input from philosophy.

The word "philosophy" may well suggest something abstruse and unpractical, and indeed a lot that goes on in philosophy is pretty obscure and practically useless, but not perhaps everything. On the one hand, there is the original motive for doing philosophy - a drive for a general understanding of the world and our place in it and for a reasoned view of what that world ought to be like; on the other, there are the more detailed studies philosophers have undertaken of the nature of knowledge and of the workings, and especially the logical structure, of meaningful language (which is the medium in which much of our knowledge gets expressed). These two aspects of philosophy can, I believe, offer us something of value. The original motive can spill over into education - we can try to see more clearly what we are doing and what we should be trying to achieve - and I hope that it will in the other part of your philosophy of education course. But for now we are going to try to learn from the narrower studies of knowledge, language, and logic.

We are going to start by looking at some simple features of the sort of argument in which someone offers you reasons for believing something or for doing something. We shall go on to dig a little deeper and so see the different things that are happening - the idea being that if you begin to see how arguments work you should be better able to judge them and perhaps better able to construct them yourself. People often try to justify the study of mathematics or Latin or history by saying that they will help you to reason better. There is evidence that studying history helps you to reason better about history, studying mathematics helps you to reason better about mathematics, but that studying these subjects does little for your reasoning outside of the subject, in everyday life or in other subjects. In this course we shall take the bull by the horns and simply study reasoning. Reasoning has to be about something, but I hope you won't be distracted by the subject-matter from the general points being made. (On the other hand, I hope you will find some of the subject-matter interesting; I have tried to avoid invented examples as much as possible, because the aim is for you to deal better with real examples.)

After looking briefly at arguments, we shall look at a few of the things that can go wrong with the statements that make them up. We shall then look at some of the general aims people have in putting forward an argument, in offering reasons. As before, the idea is that if you know what to look out for and can draw some subtle distinctions you are better placed to detect flaws in arguments, your own and other people's.

In this course you are only taking the very first steps towards the study of argument and into the issues that arise in that study - first steps are, however, among the most important. At the end I indicate a few of the further issues that arise and suggest a few of the many books you might use to travel further.

As you will soon discover, there are frequent exercises throughout the text. I suspect you won't learn very much from reading the text unless you also try to do these exercises. I have given my answers, with some supporting remarks, at the end [in this Web version, in pop-up boxes which are also endnotes to each file for those unable to use the pop-up windows]; but please try to do the exercises for yourself first and then try to work out why my answer is different (if it is). I should say now that there is often room for alternative answers. I should be glad to know of any such alternatives, and also of any errors or gross inadequacies in my own suggestions.

I have tried as much as I can to use arguments actually put forward by other people. Such arguments are followed by a number in square brackets and the source can be found by looking up the number in the list of sources at the back.

1. PRELIMINARIES

Some arguments simply consist of people shouting at each other, war pursued by other means; but these are not our concern. We shall be looking at arguments in which someone offers reasons for what she claims. Indeed it might be better to say that we are concerned with reasoning, since we very often offer reasons without pretending to be arguing a point. Consider the following remark:

(1) By the look of those clouds, it'll rain today.

If I said that to you, it is hardly natural to say that I have argued for the conclusion that it will rain today; I have indicated a reason for believing that conclusion true, but we wouldn't normally call my remark an argument. But still, we could make a few changes to transform it into what is obviously an argument for the conclusion. Perhaps we could rewrite it thus:

(2) Those clouds are very dark and heavy, so it is likely that it will rain today.

That is not so natural (nor, perhaps, so persuasive) but it makes explicit what was going on in the first remark; we could say that it is an explicit argument corresponding to remark (1). What is important for us is to see, on the surface of (2) but hidden within (1), a movement from reasons to a conclusion. Whenever we can find such a movement I shall say that we have an argument.

In our example what is at stake is a fact, that it will rain at a certain place and time; since it is a fact in the future we would normally call it a prediction, but people argue for factual claims of all sorts and that is all I am concerned with now. But besides offering reasons for believing that something is, was, or will be so, people also argue for what we can call practical conclusions; they offer you reasons for doing one thing rather than another. "They will have to raise taxes, but we won't, so vote for us!" Here the speaker is trying to get you to vote for her; she might have said "so you ought to vote for us" or "so you should vote for us", but however she expresses herself you can see again the movement we have noted above from reasons (here possibly dubious claims about different parties' future decisions about taxes) to a conclusion.

So then, an argument involves three things: reasons for something, the something itself, and the movement from the reasons to the something. In my first example, the clouds being dark and menacing; the rain I predict; and the movement I want you to make from the first to the second, from the clouds to the rain. Let us agree on some terminology. When we have an explicit argument, we shall be able to distinguish on the one hand one or more premises (some people spell this "premisses") and on the other, one or more conclusions (for convenience we shall usually assume there is only one conclusion to each argument), and we shall say that the premises are intended to support the conclusions.

It is worth noting explicitly that not every connected discourse involves reasoning. A description of something or some event may involve a lot of statements connected together in various ways, but it won't usually involve some of those statements supporting others in the way we are concerned with. Similarly a typical narrative will contain many statements, often with very close connections between them, but while in a narrative the earlier statements may make the later ones intelligible, they won't usually be giving them the sort of argumentative support we are focussing on. Very roughly, a narrative may help you see that the events described by certain claims are intelligible, an argument will give you reasons for thinking those claims to be true or acceptable.

If you read an explicit argument, or listen to one, the premises and the conclusion will be statements, the support will be mainly between the statements, as one might say; it will be indicated as much by how the statements are strung together as by anything else. We shall come back soon to the claim that the premises and the conclusion are statements; for the time being you can regard it as one of our attempts at simplifying our subject-matter. The premises will express the reasons that are being offered for the conclusion. The premises are the starting-points for this argument; the arguer simply asserts them or takes them for granted, in this argument. (Indeed, they may be so taken for granted that they are not actually expressed - we shall come back also to this point later on.) She might argue for them elsewhere, but we must take things one step, one argument, at a time. The conclusion is what the arguer is getting at, what she is trying to prove or establish or defend. And the arguer claims that the premises do support the conclusion, that the conclusion follows from them or is in some other way supported or made more likely by them.

Let us see if we can pick out premises and conclusions from a couple of real arguments. Here is part of a chapter by Bertrand Russell:

(3) The instincts and reflexes with which a child is born can be developed by the environment into the most diverse habits, and therefore into the most diverse characters. Most of this happens in very early childhood; consequently it is at this period that we can most hopefully attempt to form character. [1]

When we want to distinguish premises from conclusions it is often easiest to start by looking for the conclusion. What is Russell getting at? What would he want us to take away from this passage? Russell has in fact set out his argument very clearly and he has given us our answer by his use of the word "consequently". Putting it explicitly, his conclusion is that we can most hopefully attempt to form character in very early childhood. (Note that to make it absolutely explicit we had to look back to see what "at this period" referred to.) What reasons does Russell offer? He makes two points. The first is contained in the first sentence: "The instincts and reflexes with which a child is born can be developed by the environment into the most diverse habits, and therefore into the most diverse characters." The second is the claim of the second sentence that most of this environmental moulding happens in very early childhood. These two points join together to support his conclusion. If we want to formulate Russell's argument absolutely explicitly we would have to do a little rewriting, expanding the word "this" a couple of times, but in this case there is very little such manipulation to be done. Regimented in text-book fashion, with premises in a column, followed by the conclusion, Russell's argument might appear thus:

(4) The instincts and reflexes with which a child is born can be developed by the environment into the most diverse habits, and therefore into the most diverse characters.

Most of this development happens in very early childhood.

So, we can most hopefully attempt to form character in very early childhood.

Let us move on to a less straightforward example, taken from Dr. Johnson (with spelling modernized):

(5) Many of the books which now crowd the world, may be justly suspected to be written for the sake of some invisible order of beings, for surely they are of no use to any of the corporeal inhabitants of the world. [2]

This is less straightforward partly because Dr. Johnson does not exactly mean what he says; he is being sarcastic. But let us take him at his word (which is a necessary part of appreciating his sarcasm) and try to sort out his reasoning. We can rephrase his points in the interest of brevity and clarity. What is the conclusion of the literal argument? Many of the books that now crowd the world are written for some invisible order of beings. Why? They are of no use to us. The explicit argument is then:

(6) Many present-day books are of no use to us.

So, those books are written for some invisible beings.

Note that in Johnson's argument the conclusion comes first. He begins with what he is getting at, and then tells us why we should accept it. Notice also that he indicates the conclusion by using the phrase "may be justly suspected to be" which I have changed to "are". This phrase is really part of the scaffolding of the argument and is not really part of the content of the conclusion itself; it is Johnson's way of telling us that we may reasonably move from the fact that many books are of no use to us to the conclusion that they are written for an invisible order of beings. In less literary English there are other words, e.g. "must", which often play a similar role. Just now I could well have said "the conclusion that they must be written for an invisible order of beings". But really the "must" is not a part of what is concluded; it indicates that it is the conclusion.

In general it is worth noting explicitly that you will often be able to rephrase and usually simplify what a person writes. But be careful; some apparently minor parts of what is written can be of the utmost significance.

In looking at our examples I have relied on your ability to recognize the major components of each piece of reasoning. In learning how to describe arguments you must be able to do this well, so it might be useful to stand back to see how you do it, or at least how you might go about doing it when the going gets rougher. What we need to know is how to pick out the various statements that make up an argument. The first step is easy, at least if you have to contend with a written argument: pick out the sentences. But several of the examples we have already looked at show you that you will often need to do more, since sometimes (as with (2) and (5)) what is grammatically one sentence may contain a whole argument and in many other cases what is grammatically one sentence will contain more than one of the component parts of an argument (as with the second sentence of (3)). The second step is then to pick out some of what in grammar are known as the clauses in each sentence. Very roughly this is a matter of picking out finite (inflected) verbs; each such verb belongs to a different clause. So if we are faced with the sentence "OPEC is in disarray, so oil prices will fall" we can find two finite verbs, "is" and "will" (or "will fall" if you want to include the whole of the verbal phrase). The first gives you one component, "OPEC is in disarray," the second the other, "Oil prices will fall." Note that equally helpful in parcelling out the sentence are the conjunction "so" and the punctuation mark.

Unfortunately when you start looking at what you have been doing without a care in the world, it often turns out to be frightfully complicated. It is in this case. I said the second step is to pick out some of the clauses in each sentence, because we have been ignoring some of them in our preceding discussions. Look, for instance, at the second sentence of (3). Here we find three finite verbs, "happens", "is", and "can...(attempt)", but we only counted two important components. In this case we ignored the way Russell had focussed on the period of time and rewrote it making "can attempt" the only finite verb in the conclusion. Once again our division of the sentence is supported by the punctuation and the conjunction "consequently". Nor have we made any separate mention of clauses which simply describe things talked about in other clauses, such as "with which a child is born" in (3), or "which now crowd the world" in (5). (But note that in one case we kept the clause in our reformulation while leaving it out in the other, though Johnson's clause certainly contributes to what he says, but not to the content of the reasoning contained therein.)

I have already alluded to another kind of complexity in the analysis of the sentences used to convey an argument when we looked at Johnson's phrase "may be justly suspected to be". To connect this to our suggested second step of counting clauses, let us suppose Johnson had written "We may suspect that many books have been written..."; here we would count two finite verbs, "may (suspect)" and "have (been written)" but I have suggested that we ignore the first as merely an ornamental comment on the content of the second, which is the real meat of the argument. Sometimes similar comments are more important and should not be ignored in reformulating the argument, as in (2) where "it is likely" makes a difference to the reasoning. (For our purposes we can regard the conclusion of (2) either as "It will rain today" with "It is likely" qualifying the amount of support given that conclusion or as "It is likely that it will rain today", though for other purposes the former approach is preferable. But many such comments will have to be treated in the second way.)

The moral of all this is that we don't have automatic procedures for picking out the crucial parts of an argument. Isolating the clauses can be useful, but don't do it uncritically.

So far I have suggested two steps: pick out the sentences, then pick out their component clauses. We shall see in some later examples that it may be useful to go on even further, in particular to reconstruct some of the clauses that might underlie phrases in your statements. Suppose someone said "The present disarray in OPEC strongly suggests a fall in the price of oil." This has only one finite verb "suggests", but it is arguably a close paraphrase of the reasoning we looked at earlier, so in analysing it we would translate "the present disarray in OPEC", a noun phrase, into the sentence "OPEC is now in disarray", and similarly treat the other noun phrase, "a fall in the price of oil". More generally still, paraphrases in terms of clauses (and clause connectives) can often be illuminating - so one might suggest translating "All bald men are discontented" into something like "If any man is bald then he is discontented."

One last suggestion for sorting out the appropriate components of an argument: look to see whether what you have listed as distinct components cannot be merged, either because they are simply stylistic variants, different ways of saying the same thing, or because one is the opposite of another, i.e. what you have listed as "p" and "q" are really "p" and "not p", to use the conventions we shall adopt. Considering these two possibilities can often do a lot for the clarity of your description of an argument, but be careful! What may seem slight differences may be crucially important, and you need to make sure that you have real opposites - "No men are selfish" is not the opposite of "All men are selfish."

On the basis of examples (3) and (5) we can also see that the two functionally different components of an argument, premises and conclusion, can be arranged in various ways. You can't automatically write down the structurally significant parts of the argument from the sequence of sentences in front of you, just as we have seen the bare sentences do not necessarily yield all the significant statements within the argument. You have to ask: "What is the arguer getting at?" and then "What is she offering in support of this conclusion?" And as we have seen, the order alone can't decide these questions. But we have also seen that some words can act as clues to the answers to these questions. Words like "and so", "consequently", "implies", or "therefore" usually join premises to a conclusion in that order. On the other hand, words like "because", "for", "since", or "may be inferred from the fact that" tend to link a conclusion to its premises in that order. Again we noticed that the conclusion of an argument may be marked by the presence in it of a word like "must". All these clues can help you discover what is going on in any particular argument; but in the final analysis, you have to use your total grasp of the language to arrive at the most likely structure of argument in a passage.

While it's not automatic, I've suggested you use the following strategies to reveal what is going on in an argument:

(a) pick out the grammatical sentences;

(b) pick out the conclusion;

(c) pick out salient clauses;

(d) paraphrase to release other salient clauses;

(e) let conjunctions, punctuation reveal structure.

One thing you will soon realize is that your total grasp of the language often reveals how complicated even the simplest arguments are. For one thing, as we have already noted, an enormous amount is left out - you may well have felt that in both (1) and (5) the arguers were relying on connections that they didn't state. For another, there may be hints of complexities both inside the argument and in its external relations to the rest of what someone is saying. For instance, when we looked at Russell's argument (3), I said nothing about the occurrence of "and therefore" in the middle of what I called the first point. But I went on to say that words like "therefore" usually link premises to conclusions, so you might well ask what it is doing here. If we were to look closely at Russell's argument, I think we could say that what I called the first point itself contains a little argument - from the premise that instincts and reflexes can be developed by the environment into the most diverse habits, to the conclusion that instincts and reflexes can be developed by the environment into the most diverse characters; or perhaps we should say that the people whose habits they are will, therefore, exhibit the most diverse characters. Russell certainly relies on some such connection between habits and character. In the passage I quoted, he may be relying on some earlier, more detailed presentation of this connection; in this argument (3) it is almost a single claim, but for completeness' sake there is nothing to stop us recognizing this complexity in his first premise. We can analyse arguments with different degrees of detail; the appropriate amount depends on our purposes.

A different sort of complexity is displayed by (5), since in its context this is not seriously meant as an argument for the conclusion that many books are written for some invisible order of beings, but simply as an exaggerated way of saying that many books are of no use to us. But if we want to see exactly how Johnson makes his point, then we can isolate it from its context as we did above. But our final verdict on the argument must wait upon our examination of the wider context. I do not apologize for starting with examples that are not completely clear or cut and dried. Just as physical science uses simplifications and idealizations to help us understand the tremendous complexity of the actual world, so we are beginning to use some simple tools for dealing with the tremendous complexity of actual thought and argument. Understanding usually requires simplification.

One simplification, or perhaps regimentation, of actual arguments that we shall adopt has been mentioned already. We shall regard all arguments as made up of statements. Statements are the sorts of thing you use declarative or indicative sentences to express. The test for a declarative or indicative sentence is: can you grammatically put "Is it true that....?" around it? If you can, it is declarative; if you can't, it isn't. So if someone answers a question by saying "Rain" or "Stop asking me such silly questions" he is not using declarative or indicative sentences, so we could not use them, as they stand, as parts of an argument. But if instead he had said "It is raining" or "I wish you would stop asking such silly questions" or "You are to stop asking such silly questions", then he is using declarative sentences, and so, on our terms, making statements that could appear in an argument. These examples may also suggest how many actual cases can be handled to fit our requirement. In particular, facts are often not stated (as in our very first example (1)) - we require them to be set out explicitly; and often the conclusion of an argument or piece of reasoning is a command or an action - we can find ourselves a statement that the action is to be done. (You could handle commands directly, but in this course there is a very great deal we shall have to omit or approximate.)

To sum up what we have covered so far, when we tidy up a piece of reasoning it consists of a sequence of statements, with one or more premises supporting one or, occasionally, more conclusions.

EXERCISE A

(I) Which of the following are indicative or declarative sentences?

1. Can you lend me five dollars?

2. You will lend me five dollars.

3. Cloud, and scattered showers over the hills.

4. If only John would come!

5. If India wins the test in Antigua they will still lose the series.

6. Breaking promises is wrong.

7. I wonder what became of Rimbaud.

[Answers.]

(II) Rewrite the following passages as explicit arguments:

1. The two strongest determinants of completed family size are education and occupation. So, achieving the goal of replacement fertility must encompass improvements in education and increased job opportunities for women to be effective. [3] [Answer.]

2. Time spent on dialogue should not be considered wasted time. It presents problems and criticizes, and in criticizing, gives human beings their place within their own reality as the true transforming Subjects of reality. Even when we regard the work of the agronomist-educator as limited to no more than the teaching of new techniques, there is no comparison between dialogue and anti-dialogue. Any delay caused by dialogue means time saved in firmness, in self-confidence, and confidence in others, which anti-dialogue cannot offer. [4] [Answer.]

3. If the French Revolution were to recur eternally, French historians would be less proud of Robespierre. But because they deal with something that will not return, the bloody years of the Revolution have turned into mere words, theories, and discussions, have become lighter than feathers, frightening no one. [40] [Answer.]

2. THE ELEMENTARY STRUCTURES OF ARGUMENT

Now that we have agreed how to talk about arguments, let us move on to examine how they are constructed. Actual arguments exemplify an unlimited number of structures, shapes, or forms; but all these variations are built up out of a few simple patterns, and it is with these that we shall be concerned. To make life easier, and to save space, we shall use some simple notation: little letters, a, b, c,... to stand for the statements that make up the premises and conclusion, and an arrow --->, to stand for the support the premises are meant to give the conclusion. It doesn't matter which little letters we use, just so long as in each argument the same little letter means the same statement each time. We shall also allow ourselves to write "not a" for the statement that is the opposite of "a". We shall occasionally modify our simple little letters by other words such as "never" and "it is possible that" to convey meanings that should be obvious from the context. We shall also use brackets and asterisks to help make the structure clearer. When we draw a picture of an argument using our little letters (with their interpretation) and arrows, we shall call it a labelled structure diagram.

To illustrate our way of displaying the structure of an argument, consider the following piece of reasoning:

(7) Mary must be at home because there's a light in the kitchen.

The premise here is "There's a light in the kitchen" which we can symbolize "a", the conclusion is "Mary is at home" (note again our change of "must be" to "is") which we'll write "b". So the labelled structure diagram for example (7) is:

(8) a = There's a light in the kitchen.

b = Mary is at home.

a -----> b

That is an example of the simplest structure of all: one premise supporting a conclusion.

Not merely is it simple, it is not very persuasive. Suppose the arguer offers an expanded version:

(9) Mary must be at home because there's a light in the kitchen and she never leaves the kitchen light on when she goes out.

This time we have a third statement, which we can label "c", "Mary never leaves the kitchen light on when she goes out." What is this extra statement doing? Surely, it is working with the statement we have labelled "a" to support the conclusion; it's joining together with the statement "a" to help reach the conclusion. We need to show the way "c" is joined to "a"; to do this we shall use brackets, as in diagram (10):

 (10) a ) 
                   ) -------> b 
      c ) 

(12) a ---\ 
                \--- > 
                           b 
                /--- > 
     d ---/ 

a ) 
           ) -----> c 
b ) 

 b ---\ 
             \----> 
                          a 
            /----> 
 c ---/ 

 a ) 
                     ) --> c --> d --> e 
 b ) 

 not b --\ 
                 \----> 
                             a 
                 /----> 
      c --/ 
[Answer.]

 b ) 
   ) 
           c ) ------> a 
   ) 
 d ) 

 b ) 
             ) ------> a 
 c ) 

 c ---> d )
 
                         ) ---> a ---> b 
        e )

  b)
 
   )
 
  c) ---> a)
 
   )       ) 
  d)      d) ------> f)
 
           )          )
 
          e)         h) --> g
 
                      ) 
                     i)