Motivation

You’ll recall that I mentioned that the Ancient Greeks also studied Critical Reasoning in their own way, though they called it Rhetoric. Here is a quote from a dialogue of Plato which describes how its champions (whom we may identify with the Sophists) justified it. In this dialogue Socrates is questioning Gorgias, a great sophist of the time who is visiting Athens. He has said that he can teach the art of rhetoric to anyone who will take the time to learn, but Socrates wants to know just what it is that he claims to be providing when he does so.

SOCRATES: … And so come, Gorgias, imagine you are questioned by these men and by myself as well, and answer what it is you claim to be the greatest blessing to man, and claim also to produce.

GORGIAS: Something, Socrates, that is in very truth the greatest boon, for it brings freedom to each man, and to each man also dominion over others in his own country.

SOCRATES: And what exactly do you mean by that?

GORGIAS: I mean the power to convince by your words the judges in court, the senators in Council, the people in the Assembly, or in any other gathering of a citizen body. And yet possessed of such power you will make the doctor and the trainer your slave, and your businessman will prove to be making money, not for himself, but for another, in fact for you who can speak and persuade multitudes.

                                                (Plato, Gorgias, 452d-e)

The art of rhetoric according to this is just the art of being able to persuade others to believe what you want them to believe simply by the power that lies in the words that you use. And because you can change their beliefs you can get them to do things that you want them to do, because what they do depends only on the desires that they have and the beliefs that they have. Obviously this is a very important thing, especially in a society like that of Athens where the people in council made all the important decisions.

Pretty clearly, arguments as we usually understand them are one very important way of persuading people, and that’s why we’re going to start looking more closely at the nature of arguments in general. But we don’t start from zero: we already have some idea of what we’re talking about when we talk about arguments. In the last lecture I mentioned almost in passing a way of understanding arguments just in the course of showing that arguments are fundamentally linguistic events – a point of view that we can see Gorgias agrees with. We might, for example, take the following as a rough and ready description of an argument. It is not to be taken too literally, and it’s definitely not a definition. It is simply a statement of our (almost) first thoughts on what we think we know about arguments.

F.         An argument is a set of claims called the premisses, and another claim called the conclusion, such that a person puts forward the premises to try to persuade others to accept the conclusion on the grounds that the latter is guaranteed by the former.

The Pragmatic View of Arguments

Persuasions

This is an example of treating arguments as sets of statements, but as I’ve also mentioned before, there’s now general agreement that the approach to the analysis of arguments which is most likely to yield useful results is a ‘pragmatic’ approach. A pragmatic approach to arguments sees arguments as a particular type of social interaction, the type of social interaction we’re having when we’re trying to persuade someone of something. I don’t know if here’s a general name for this sort of thing so I’m just going to call it a Persuasion Situation.

We’ll get back to arguments in just a moment but to do so we’ll have to know how to talk about some of the things that are important in these so-called ‘persuasion situations’ There are at least four things that we need to name[1]:

D1.          The Proponent. The person(s) attempting the persuasion. We don’t have to assume that this person believes what they’re trying to get the other person to believe.

D2.          The Respondent. The person(s) being subjected to an attempt at persuasion. It is not necessary that the respondent previously disbelieve the thing that the proponent is trying to persuade them of. Indeed, there doesn’t need to be any attitude towards it at all, as long as she does not already have the attitude of belief. (If the ‘respondent’ already believes it then the situation is one of confirming a belief set rather than altering/augmenting it.)

D3.          The Inducement. The technique of persuasion which the proponent uses.

D4.          The Target. The thing that the proponent is trying to persuade the respondent to accept. Typically, if something is to be presented to a respondent as a candidate for belief it has to presented by a sentence in our natural language. I shall also use ‘target’ to refer to that sentence and I will rely upon context to make clear which is intended.

Argument as a Type of Inducement

Contexts of Presentation of the Target

According to this, for something to count as an inducement it just has to be a way of making the target acceptable to the respondent, or, to put it another way, it involves putting the presentation of this target into some context which can affect the disposition of the respondent to accept it. Now, the popular view of arguments sees them as an entirely linguistic technique of persuasion, so that the context of presentation of the target in an argument would be linguistic in nature. You can see that this is what Gorgias had in mind in his explanation to Socrates. But linguistic contexts don’t exhaust the possibilities: there are inducements that involve extralinguistic elements and there are those that do not. In the former class we would include such things as brainwashing with drugs, the influence of peer pressure, and so on. In the latter class we would find all the ways that ideas can be imposed by mere language users. This class corresponds to those techniques anciently taught as Rhetoric; and we know that this class, while it includes everything that we would call an argument, also includes much that we would not call argument. An education in Logic, as we understand it, would have formed only a small part of training in rhetoric.

The Power to Affect Dispositions to Believe

A more interesting way of looking at inducements is in terms of their power to affect dispositions to believe on the part of the respondent. This power is possessed by inducements in all the various ways in which one mind can be imposed upon by another. In the example of brainwashing, for example, the mind of the respondent is crudely attacked in such a way that the normal procedures that mediate belief adjustments are disabled. The more typical case, I suspect, is that the inducement merely preferentially activates these procedures to achieve the desired result. When a belief is induced by peer pressure, for instance, we may suppose there is some (evolutionarily plausible) psychological mechanism that prefers beliefs held by the relevant group, and that the inducement activates this mechanism while possibly suppressing others.

In the case of arguments, however, it appears from our untutored understanding of arguments that the form of inducement characteristic of this class is simply the presentation to the respondent of a set of statements, including the target, in which the statements are supposed to have some sort of peculiar relationship to each other so that the acceptability of the target is in some way increased by the acceptability of the other members of the presented set, and the presentation itself is such that the respondent is affected by this relationship in the appropriate way.

An Intuition about Arguments

Now we have no very convincing story to tell about what really brings about this improvement in the ‘acceptability’ of the target for the respondent. All we can really say here is that there must be some psychological mechanism involved which causes the change in the disposition of the respondent to accept the target if the other propositions are accepted. For the presentation itself to be such that such that the respondent is affected by this relationship in the appropriate way is just to say that the presentation is such that it causes the activation of this psychological mechanism. We would probably experience this sort of thing as an intuition that the conclusion follows from the premisses. Or perhaps we would be able to distinguish several types of intuition for several types of argument. It might be that the intuition which tells us that

‘All men are mortal and Socrates is a man, so Socrates is a mortal’

is a good argument is different from the intuition that makes us accept that

‘The sun has always come up in the morning so I can expect the sun to come up tomorrow morning’

is a good argument. But let’s not dwell on these matters. Let’s suppose that there is just one Argument Intuition which tells us when an argument is good or not.

Characterization of Argument

We are now in a position to attempt a characterization of this technique of persuasion, and it seems appropriate and useful to begin by giving the meaning of some terms which shall be used specifically to describe aspects of this technique.

D1.     An Argument Formation is a sequence of statements in which

a.        the final member is called the Conclusion, and

b.        the other members are called the Premisses.

D2.      An Argument Base for an inducement is an argument formation in which the conclusion is the target of the inducement. An argument base defines premisses and conclusion for the inducement.

First Characterization

Given this terminology, let’s make a first stab at defining arguments. How about this:

D3.      An Argument specifies an argument base.

So for example, someone may say the following:

Arg.A.      “All men are mortal and Socrates is a man so Socrates is mortal.”

and in a statement like that we can identify the argument base as the following:

F.A:         < ‘All men are mortal’,

   ‘Socrates is a man’,

   ‘Socrates is mortal’ >

But we’re interested in being able to determine the ‘good’ arguments, so, following the discussion above let’s use that definition of an argument to define ‘good’ arguments in the following way:

D4.      An Effective argument is one such that just recognising the argument base causes the respondent to feel the Argument Intuition, so that the respondent is more disposed to accept the target if the other statements are accepted.[1]

‘Specifying’ the Base

Now when I say an argument specifies a base, I mean something like ‘can reasonably be interpreted in terms of a specific base.’ This is a pretty significant sort of claim and we need to understand what is involved in it. We’ll start that now, but there’ll be more on this topic later. The first thing to notice here is that if you say of some event – say some conversational event – that it specifies an argument base, you’re actually claiming that it presents for your consideration statements that can be ‘premisses’ in an argument form and a statement that can be a ‘conclusion’ in an argument form. And, as the definition of ‘good’, effective arguments makes clear, the whole purpose of this business is that the premisses are supposed to support the conclusion, so the presentation of the statements should naturally suggest which of the statements get to be premisses in the base and which gets to be the conclusion.

It’s not hard to think how this happens. Our language – any language – is full of little clues about how we should understand the statements that are made. The most obvious clues occur when we see special words attached to the statements that are presented in a possible argument event. Some words indicate that the attached statements are premisses in an argument, and some indicate that the attached statements are to be conclusions. These are the clues that we use to decide where in an argument base the presented statements are to be put. They’re also, of course, amongst the clues that we use to conclude that some event may contain (or be) an argument.

           Premiss                 A word or phrase to signal that a local passage contains statements

           Indicator                that can play the role of premisses.

                                                EG: Because, since, for, as we see from the following.

NB: In cases where there is ambiguity or uncertainty in determining whether a word or phrase is acting as a premiss indicator, replace with ‘because’ and determine whether the meaning has changed. If it has not then it was a premiss indicator.

                                                ‘Since I’ve repaired the car I can drive to Melbourne.’

‘Since I repaired the car I’ve driven to Melbourne.’

           Conclusion            A word or phrase to signal that a local passage contains a statement

           Indicator                that can play the role of a conclusion.

                                                EG: therefore, hence, so, implying, thus, consequently.

NB: In cases where there is ambiguity or uncertainty in determining whether a word or phrase is acting as a conclusion indicator, replace with ‘therefore’ and determine whether the meaning has changed. If it has not then it was a conclusion indicator.

                                                ‘I’ve repaired the car, so I can drive to Melbourne.’

‘I repaired the car so that I could drive to Melbourne.’

           Argument             A word or phrase to signal that a local passage contains an argument.

           Indicator                Premiss and conclusion indicators are argument indicators.

Be aware that words never have just one function. Sometimes a word may act as an indicator, and at other times it may not. And be aware, too, that indicators are not actually necessary for us to be able to specify an argument base. In many cases passages are interpreted as if they contained premisses and conclusions even though there are no words or phrases that could plausibly be counted as indicators. That’s language for you.

Arguments in Standard Form

Before we get back to the real topic of finding out just what is meant by an argument, I have to mention that what I’ve identified as the ‘argument base’ corresponds in most textbooks to something they call the ‘standard form’ of an argument. Whether we call it the base or the standard form we can see that it is a useful concept because we can see that the same argument can be presented in many different ways.

E.g. Socrates is mortal since Socrates is a man and all men are mortal.

                Because Socrates is mortal and all men are mortal, therefore Socrates is mortal.

                Given that all men are mortal, Socrates is mortal since Socrates is a man.

Since one and the same argument can be presented in a variety of ways it is helpful to have some standard way of presenting them.

D5.           An argument is said to be in Standard Form just in case it looks like this:

                                Premise 1

                                ...

                                Premise n

                    \        Conclusion

(The symbol '\' represents the conclusion marker 'therefore' or 'hence' — the inferential move in an argument from supporting premises or assumptions to the argument's conclusion that is taken to be supported by them. We will usually omit the symbol '\'. In such cases we can assume that the line separating premises from conclusion, '—', represents the inferential move.)

E.g.                 All men are mortal

                                Socrates is a man

                                Socrates is mortal

We’re going to see now why this is not such a good way to think of arguments.

Inadequacy of the First Characterization

In the case of the ‘Socrates’ example I’ve been working with I think it’s obvious that we do believe that recognition of the argument base alone is sufficient to cause the respondent to feel the Argument Intuition, so that the respondent is more disposed to accept the target if the other statements are accepted. If you don’t see this immediately then perhaps you lack this intuition entirely and if so then there is no way that any argument like this can be convincing to you. Or perhaps it’s a matter of educating your intuitions. I’m sure that I, with my long experience of dealing with these things, can see the effectiveness of some arguments more readily than you who are just beginning to think systematically about these things.

On the other hand, we mustn’t suppose that all such failures are due to mental deficiency or ignorance. It is not hard to see that those obvious sorts of arguments are only a tiny subset of what we usually mean by ‘good’ arguments and that if we’re going to account for all the good arguments we’ll need to change our definitions. Why do I say that this is the case? Well, I’ll show you now that even for an argument that is only slightly more complex than that example, it becomes plain that more is required to be specified by an argument than a simple record of the premisses and conclusions in order to achieve this acceptance.

Consider this argument:

Arg.B.      “All men are animals and all animals are mortal and Socrates is a man so Socrates is mortal.”

which I think can be said to specify the argument base:

F.B:         < ‘All men are animals’,

   ‘All animals are mortal’,

   ‘Socrates is a man’,

   ‘Socrates is mortal’ >

Here mere recognition of the argument base is conceivably insufficient to satisfy the condition required for it to be effective. Because the argument is just too long, the respondent may simply not feel the argument intuition connecting the premisses and the conclusion. And yet we want to be able to say that this looks like a pretty good sort of argument, one which the respondent should be able to feel the force of. We even think that we know how she could be made to see its force using nothing but the resources already present in the argument base.[2]

Second Characterization

One way of understanding what lies behind this belief, is to consider what our response could be if the respondent denies the force of the argument. In the case of argument Arg.A, for example, if the respondent claims not to see how the conclusion depends upon the premisses there seems to be nothing that we can do to make the relationship any clearer. Either the respondent can see the connection or she cannot. In this respect the argument base in Arg.A may be taken as an ‘irreducible’ type of argument. It is worth giving this sort of thing a name, so:

D6.      An Effective Argument Formation for the respondent is an argument formation whose recognition causes the respondent to feel the Argument Intuition, so that the respondent is more disposed to accept the conclusion if the premisses are accepted.

Note that in this definition effectiveness has been made relative to the respondent. If we are going to talk about arguments being good or bad as means of persuasion then we have to be aware that different people will find different things to be persuasive. If we account for the ‘goodness’ of an argument in terms of intuitions we have to accept that intuitions are events in an individual, and may differ from person to person.

Now let’s turn our attention back to Arg.B. We find that, unlike in Arg.A, we could reasonably respond to a claim of incomprehension by showing that a subset of its premisses form an effective argument formation with a statement not included in the argument base occupying the role of conclusion, and that this statement together with some other premisses from the argument base are the premisses to an effective argument formation with the target occupying the role of conclusion. If that sounds complicated it is really just a way of saying that the supposed argument can be broken up into smaller parts with intermediate steps. To be even more explicit, the argument Arg.B can be associated with an ordered pair of argument formations that are effective for the respondent, like this:

F.1:         < ‘All men are animals’,

   ‘All animals are mortal’,

   ‘All men are mortal’ >

F.2:         < ‘All men are mortal’

   ‘Socrates is a man’,

   ‘Socrates is mortal’ >

The effectiveness of F.1 means that the acceptability of the argument premisses warrants – for the respondent – the acceptability of F.1’s conclusion, and the effectiveness of F.2 means that the acceptability of the argument premisses and of F.1’s conclusion warrants the acceptability of the argument’s conclusion. Therefore, granted the transitivity of “warrant”, if the respondent becomes aware of this sequence, that is sufficient to cause the respondent to feel the argument intuition, so that the respondent is more disposed to accept the conclusion if the premisses are accepted.

Such a process is obviously generalisable. For any argument sufficiently complex that conviction does not follow immediately upon recognition of its base we can attempt to show that a sequence of formations can be discovered such that conviction is immediate upon its recognition. We can use this to improve our definition of argument.

First, let’s name those longer sequences as Argument Completions. To be precise:

D7.      An Argument Completion associated with an argument is a sequence of argument formations where:

a.        Each premiss of the base appears as a premiss in at least one formation in the sequence,

b.        The final formation in the sequence has the target as its conclusion.

Now we define an argument:

D8.      An Argument specifies:

a.        An argument base,

b.        An argument completion.

And in order to talk about ‘good’ or ‘successful’ arguments we need to define the following:

D9.      An Effective Argument Completion for the respondent is an argument completion whose recognition causes the respondent to feel the Argument Intuition, so that the respondent is more disposed to accept the conclusion of the argument if the premisses are accepted.

The example above (of <F.1, F.2>) suggests that a completion in which every formation is an effective formation for the respondent is thereby an effective completion. On the other hand, we just don’t know whether an effective completion contains only (or even any) effective formations. Whatever our suspicions may be, nothing which has been said makes it apparent that some organization of ineffective formations could not be itself effective.

Given all this we can now propose a definition of a ‘good’ argument.

D10.    An Effective Argument for the respondent specifies a base and an effective completion for the respondent.

I’d be interested to hear any examples that you can come up with of arguments which seem to be ‘good’ arguments but which don’t satisfy the criteria of effectiveness. And likewise for arguments which are effective in this sense but which don’t seem to be ‘good’ arguments.

The Recognition of Arguments

If no such counterexamples are forthcoming then it looks like we’ve worked out a set of characteristics which any argument has to have. If that’s right then these are the necessary conditions for anything to be what we would wish to describe as an argument. I think it’s even more plausible that they’re the sufficient conditions. If an inducement is such that it specifies a base and a completion then I cannot imagine that we would not consider it to be an argument.

Specification and Interpretation

You’ll recall that we’ve already looked at what is meant by saying that an argument specifies one of those structures that we’ve been talking about when we were talking just about its specifying an argument base. That turned out not to be too difficult – though we certainly didn’t go into it in much depth. But if the test of whether or not an inducement is an argument depends whether or not the inducement specifies a completion then we are going to have to make it a good deal clearer what is intended by the term. Again, it’s not difficult to do in an informal way. When we say that an inducement specifies a particular structure of statements we don’t have to mean anything more than that it is possible for the respondent to discover that structure because of the information provided by the inducement, or by using that information.

In a very general sense, then, if the respondent is able to discover a base and a completion in this way then she may declare the candidate inducement to be an argument, but if she fails then she cannot consider the inducement to be an argument. On this understanding the test is really rather a matter of interpretation: an inducement is said to specify a base or a completion if the inducement can be interpreted in such a way that it is taken to provide information which leads to the definition of a base or a completion. Understanding the process in this way emphasizes the fact that whether or not an inducement is an argument is not solely dependent upon some essential quality of the inducement – as an understanding in terms of an inducement’s power of ‘specification’ would suggest. It is, instead, dependent upon qualities of both the inducement and the respondent in varying degrees, and upon the relationship that exists between those two. Our understanding of this relationship can be improved by considering some specifics of the process of interpretation.

For the purposes of this discussion we shall use the following definition:

D11.    An argument Candidate is an inducement selected by some means – it doesn’t matter how[1] – for examination to determine whether or not it is an argument.

Intention, Fidelity, and Charity

Suppose that we have an argument candidate. Let our first step be to determine whether the candidate may be interpreted so as to specify an argument base. Since the requirements for an argument base are pretty minimal, consisting of the target itself as the conclusion and one or more other propositions as the premisses, just about any act which is more than a bare assertion of the target may be interpreted so as to specify a base. This being so, the more important part of the test must be whether the candidate can at the same time be interpreted so as to specify a completion. But, in fact, any candidate that can be interpreted as specifying a base can conceivably be taken to specify a completion, for the base itself may be interpreted as a completion.

However, we need not be discouraged by this because the mere fact that a thing may be interpreted in a certain way does not mean that we are justified in interpreting it in that way. As a general rule, in the interpretation of actions attention must be paid to the intention of the actor. In the particular case of candidate arguments, if we are free to interpret the inducement without reference to the intention of the proponent then there are literally no constraints upon the interpretations that may be proposed. Any principled process of interpretation therefore depends upon a principle that we call the principle of Fidelity.

Fidelity                  A principle of interpretation that demands that a passage be interpreted in a way that remains faithful to what the author actually states.

In the case of candidate arguments recall that the process of interpretation attempts to establish that the candidate makes the existence of a particular relationship amongst propositions knowable, and if it is inconceivable that the proponent could have intended that the function of the candidate was to be achieved under some proposed interpretation then that interpretation is quite unlikely to be acceptable.

Fidelity, however, is only one half of the requirement of reasonable interpretation. The other half lies in the application of the principle that we call the principle of Charity.

Charity                  A principle of interpretation that demands that a passage be interpreted in a way that does most credit to the rationality of the author.

To illustrate this type of application of the principle of charity and to make what is unclear clear let us consider again the example of argument Arg.B above. In that example there was an obvious interpretation by which a base could be identified. Let us suppose that the base in that case was also to be interpreted as the completion. Could we judge this to be an acceptable interpretation? Is it likely that the proponent could believe that the base itself constituted an effective formation? It is possibly so, but if we know or suspect of the proponent that he is unlikely to make this assumption then we should reject that as a plausible interpretation and seek to derive further information from the candidate which might suggest a more satisfactory interpretation.

In this case we do not find any further explicit information of that nature. In fact it is probably typical that an inducement intended by the proponent as an argument underdetermines the compÿÿtiÿÿ which the argument is intended to ÿÿ specify. In such a case the recognition of the intention of there being an argument may prompt an attempted construction of a satisfactory completion by the respondent using her own resources, including her experience of arguments past and her knowledge of the sorts of things that proponents typically consider to be plausibly effective formations and completions. The danger in this case is that her charity will be excessive. A completion derived in this manner can only with difficulty be attributed to information in the candidate: although it may, I suppose, be claimed to have been derived ‘because of’ the information in the candidate it can hardly be claimed to have been derived ‘using’ that information.

We will have much more to say about the application of fidelity and charity in later lectures.