ISSA Proceedings 2014 – Towards A Foundation For Argumentation Theory
Abstract: I shall present and analyze numerous principles that argumentation theorists do agree upon (and some closely related one which they do not) and argue that the set presented here offers at best limited grounds for cross-theoretical evaluation.
Keywords: Acts, expressions, informational content, reasons, arguments, repeatable, abstract object
Argumentation theorists disagree about many things. For example, is conductive reasoning distinct from deductive or inductive reasoning? Could a painting or a judo flip be an argument? How many types of fallacies are there? Are there any enthymemes? Is relevance an independent condition of a good argument? Can a non-virtuous arguer give a good argument? Are arguments better construed as acts or as propositions or as sentences? Are all arguments dialectical? Answering these sorts of questions are among the current challenges of argumentation theory.
One impediment to answering these questions is that differing answers are often grounded in different theoretical frameworks. Hence, the issue is not merely one of trying to marshal ‘the best’ reasons for a particular answer to one of these questions, but rather to produce ‘the best’ overall theory. But now a new problem emerges – how do we assess, across theories, whether theory X is right for saying an argument can have an infinite number of premises say, while theory Y is wrong for saying an argument cannot? We could of course try to adjudicate theories in the standard way in terms of simplicity, explanatory depth and breadth, etc., but such comparisons rarely generate a neat linear ordering. One theory may have advantages in one area of explanation, but do worse in another. Even worse, the theories may not agree on even the basic ontology and not agree on what sort of thing an argument is (or could be). Hence, one might doubt that it is possible to construct a fully adequate theory of argumentation.
My concern here is to at least begin to explore the possibility of adjudicating basic ontology issues in argumentation theory. What, if anything, are the constraints on an adequate theory of argumentation at the basic ontological level (at least from the perspective of argumentation theory)? Are there any substantive principles that are accepted by all theories that might serve as grounds for adjudicating amongst competing theories? In this paper I shall present and analyze numerous principles that argumentation theorists do agree upon (and some closely related ones which they do not) and argue that the set presented here offers at best limited grounds for cross-theoretical evaluation, though I shall also point to some possible paths forward.
2. Background agreement
Argumentation theory does not take place in a vacuum. Indeed, for there to be a recognizable argumentation theory (as distinct from say particle physics or pre-Imperial Roman history or basket weaving or World Cup football) there must be much that is at least tacitly agreed upon, such as at least: there are thinking beings, there are material objects such as chairs, buildings, stars, etc. The thinking beings perform various kinds of actions and have various kinds of goals, beliefs, and desires. There are languages which thinking beings use to communicate information with each other. There are various academic disciplines that categorize this information, etc.
I am not claiming that these tacitly agreed upon items are definitely known or true or unchallenged. Paul Churchland (1981) doubts there are beliefs. Trenton Merricks (2003) argues that there are no macro-sized non-conscious material objects while Jason Turner (2011) argues there are no composite objects at all. All I am suggesting is that, as argumentation theorists, we presuppose that argumentation is a human activity that occurs within the context of human beliefs and desires and goals within a world of tables, chairs, buildings, etc.
So there is a vast swathe of propositions that I suspect we agree upon and take for granted when we are doing argumentation theory. But much of this that we presuppose does not itself impact or help us adjudicate disputes in argumentation theory since it is against this presupposed backdrop, when trying to understand the human activity of argumentation, that the disputes themselves arise. Hence, even if it turns out that Merricks is right that there are no baseballs (or any other non-conscious composite objects), but merely atoms arranged baseball-wise, then, while a part of our presupposed background is not quite accurate, the inaccuracy is not something that affects our argumentation theory. We can argue about whether baseballs were in the strike zone just as easily as whether atoms arranged baseball-wise were in the strike zone. So despite the existence of large-scale agreement, we have not necessarily made much progress in terms of helping adjudicate theory disputes in argumentation theory, since it is against the large-scale agreement that the disagreements arise.
3. Substantive agreement
Is there anything substantively relevant to argumentation theory that all argumentation theorists agree upon? (or at least should agree upon?) At the very least it seems hard to be counted as doing argumentation theory if one does not accept:
(1) There are acts of arguing
Hard, though perhaps not impossible. Could there be a world in which people give/express arguments (and so there is a need for argumentation theory, and yet there is no arguing)? Perhaps they give arguments as a form of poetry or entertainment. The question of course is whether what the people give should in fact be called ‘arguments’ (or whether even if called ‘arguments’, the study of them should be called ‘argumentation theory’). If we say ‘yes’ because historically they once used them to argue, but now do not, then the world is not a world in which there are no acts of arguing. If we say ‘yes’ because what they give/express correspond with what we give/express when we argue, then the matter is inconclusive since it may be that it is the usage of the giving/expression to argue that allows the giving/expression to be called an argument. So without the arguing, the giving/expressing in our hypothetical world would not be the giving/expressing of an argument. Regardless, even if it really were a possibility that one could do argumentation theory without there being acts of arguing, that possibility is quite remote from the situation in which we actually find ourselves – one in which there are acts of arguing.
Given the plausible background assumption that action theory and argumentation theory are not the same thing, we should also accept:
(2) Not all acts are acts of arguing
(2), unlike (1), is not a precondition for doing argumentation theory, but rather a fact about the background world that is presupposed and yet is relevant to argumentation theory. Given the world of agents with beliefs and desires, and goals and wants and needs who act on those beliefs and desires to achieve their goals in a world of tables and chairs and money, etc., there are in fact acts that agents perform that are not acts of arguing. My sitting down before turning on the computer was not an act of arguing. Your eating of breakfast this morning (assuming you ate breakfast this morning) was not an act of arguing. In general acts of poety reading, prophesying, walking, etc are, most of the time anyway, not acts of arguing. This of course leaves open where the line is between acts of arguing and acts that are not acts of arguing. For example, are acts of persuading (or attempted persuasion) all acts of arguing or not. Are at least some acts of explaining also acts of arguing? Is proving a type of arguing or not?
While we may disagree on where the line is, we agree that there is a line to be drawn. For the notion of arguing to be a relevant sub-class of action, then there need to be examples of action that do not fall into the sub-class – otherwise arguing and acting start to look like two different names for the same thing. Hence, any theory that ultimately claimed that all acts (or none) are acts of arguing is to be rejected.[i] So what to make of the critical thinking textbook – Everything’s an Argument? Despite the title, the actual claim of the book is that every instance of language or symbol use is a form of argument, which, even if stronger than most argumentation theorists are willing to accept, is still much weaker than the claim that all acts are acts of arguing.
(2) is not to be confused with the related:
(Z) Not all acts could be acts of arguing.
Put another way (Z) is: there is some act that could not be an act of arguing, or there is some act for which it is impossible that it be an act of arguing. While I suspect that many argumentation theorists agree with (Z) – there just are some acts that could never be acts of arguing, I am not sure that such agreement is justified. Indeed, if exemplifying, providing an example to show a certain kind of object, act, or state of affairs is possible, is a kind of arguing and any action could, in the right circumstances, be an act of exemplifying, then every act could be an act of arguing.[ii] (This does not mean that there is a possible world in which every single act in that world is an act of arguing – it merely means that for every act x, there is some possible world in which x is an act of arguing.)
Some argumentation theorists hold that there must be a linguistic component for an act to count as an act of arguing. Others disagree – consider for example, Michael Gilbert’s (2003) judo flip example. Regardless, if it is true that an act of arguing must involve a linguistic component, then any act with no linguistic component is not and (assuming it could not be the same act if it had a linguistic component) could not be an argument. But since argumentation theorists do not universally agree on whether an act of arguing must involve a linguistic, or even symbolic, component, we cannot use such an appeal to ground accepting (Z).
While argumentation theorists disagree about what is and is not an act of arguing and disagree about whether there are boundaries to what acts could be arguings, theorists at least agree that:
(3) At least some acts of arguing involve the expression of reasons
Stipulate that to express reasons it to give a symbolic representation of the reason. For many those expressions are limited to linguistic expressions – for others, pictorial expressions with no linguistic component will also count as expressions of reasons. But given the stipulation, Gilbert’s judo flip may be the giving of a reason, but not the expressing of one. Hence, I cannot say that argumentation theorists agree that all acts of arguing involve the expression of reasons. But what of:
(A) All acts of arguing involve the giving of reasons.
According to Tony Blair (2003), “[e]ven the broadest definitions of argument, such as those of Willard (1989 ) and Gilbert (1997 ), presupposes some element of reason-using.” Is there then no arguing if one is just giving the conclusion without reasons for it? While plausible, I am not sure that all argumentation theorists agree. For example, Maurice Finocchiaro (2003), argues that in at least some instances an argument is merely a defense of its conclusion from objections even if no reasons are given for that conclusion. Others allow the possibility of zero-premise arguments and if one thinks that for every argument there is a corresponding potential arguing, then again it seems one is committed to the possibility of an act of arguing that does not involve the giving of reasons.(See Goddu 2014) So as plausible as (A), I hold off from adding it to list of agreed upon principles. [It may turn out that resolving the Finocchiaro case or the zero-premise argument case will ultimately vindicate (A). In the former, one might hold that the rejection of objections to a given conclusion themselves constitute reasons for that conclusion, whereas in the latter, perhaps one might reject that for every argument is a corresponding potential arguing. Regardless, I leave (A) off the list for now.]
Could you have an expression of reasons that was not part of an act of arguing? I suspect so. When I give an example of a reason, I express it, even if I do not argue. If I merely repeat someone else’s reasons, I express them without arguing with them. A computer that generates complexes of sentences in the form: “A, B so C” may express reasons without any act of arguing happening. So I suspect we have evidence for:
(B) Not every expression of reasons is part of an act of arguing.
But I put (B) aside on the grounds that there may be some dispute about what counts as the expressing of a reason.
Finally, it is part of our background presuppositions about language and symbols and representations in general that they have meaning or content. Hence, all argumentation theorists should agree that:
(4) Expressions of reasons have informational content
Of course we may disagree about how to capture the notion of informational content – say in terms of propositions, or some primitive ‘same content as’ property, or something else. Regardless, we still agree that there is informational content that is distinct from the expression – “x is a bachelor” and “x is an unmarried male of marriageable age”, or “x = 25” and “x = 5 squared” may have the same informational content, but are definitely not the same expressions.
Argumentation theorists, as far as I can tell, agree on (1)-(4). At the very least they act and write as if they do even if they have never explicitly uttered or written them. I suspect most would assent to (A) and (B) as well, but for the moment I am putting those aside. (Though what follows does not change if (A) and (B) are put in the mix.) If I am wrong and argumentation theorists do not even agree on (1)- (4), then the prospects for moving forward are quite limited. If we cannot even agree on the basic constituents out of which the data we are trying to explain are constructed, then we will certainly never agree on any attempt to explain and organize that data. But is agreement on (1) – (4) enough for any progress? I turn to that question in the next section.
4. Any payoff?
Does (1)-(4) provide us enough agreement to make progress on our disputes? I suspect not, since the background presuppositions and (1)- (4) are currently consistent with:
(Y) There are no arguments.
Proof: Suppose the word ‘argument’ were stricken from our language as a myth, say on the par of ‘subluminous ether’ or ‘phlogiston’. Could one still do argumentation theory with the ontology presupposed in (1)-(4)? Yes. There would be acts of arguing which we would try to distinguish from acts that were not acts of arguing. At least some of those acts of arguing would involve the use of expressions that had informational content. One could still debate whether the act or the expression or the informational content was the most important aspect of what was going on. One could still distinguish combinations of actions and expressions that in a certain context for a certain audience would be more likely to achieve assent than other combinations of actions and expressions in that context. One could talk of the logical properties holding between different pieces of informational content. One could ask whether the actions or the expressions or the informational content could be partitioned into various categories such as good, bad, rational, irrational, deductive, inductive, conductive, abductive, enthymeme, fallacy, convergent, divergent, virtuous, etc. One could, in short, I suspect recapitulate much of argumentation theory without the word ‘argument’ referring to anything at all.
One might claim that all this shows is that the word ‘argument’ is ambiguous – sometimes it is used to refer to the acts of arguing, sometimes to reason/claim expressions, sometimes to the informational content of those expressions. Granted. But I was not trying to show that (1) – (4) entail that there are no arguments – I was merely trying to show that (1) – (4) are consistent with there being no arguments. The fact that (1) – (4) would also be consistent with ‘argument’ being a disjunctive ontological category, i.e. x is an argument iff x is an act of arguing or a reason/claim expression or the informational content of a reason/claim expression is beside the point. Put another way, (1) – (4) is consistent with none of the three contenders being arguments and with all of them being types of arguments. Nothing in (1) – (4) privileges one possibility over another. But note that even if one accepts that the word ‘argument’ is ambiguous, the word could still be excised for clarity’s sake with no ontological loss – in other words, at the very least one could be a reductionist about arguments – they are nothing over and above acts of arguing or reason/claim expressions or the informational content of reason/claim expressions (and if the ambiguity was causing theoretical problems, then for the sake of accurate theory we might decide to excise the word anyway.)
But if (1) – (4) are consistent with there being no arguments, or with just acts being arguments or with all three ontological categories including types of arguments, then agreement on (1) – (4) alone will not help us adjudicate disputes concerning the nature and types of arguments. We cannot resolve disputes concerning enthymemes or fallacies or whether there are deductive, inductive, conductive, and abductive types of arguments if we cannot agree whether there are arguments at all, or if there, are what ontological category they fall into. Suppose, however, that, in addition to ~(Y), i.e. there are arguments, we add:
(C) Arguments are repeatable
to our list of agreed upon principles. Roughly speaking, repeatable entities can happen, exist, or be instantiated more than once. On most views, material objects are repeatable, but the temporal slices of material objects are not. Your desk chair is probably the same chair as yesterday. Even if the person in the next office is sitting in the same type of chair as you – they are not sitting in the very same chair. Similarly, on most views properties are taken to be repeatable even if the particular instantiations of them are not.
Argumentation theorists write and act as if arguments are repeatable. We worry about how to correctly extract the arguments from given texts, we expect our students to give us Anselm’s argument and not their own muddled version of it, we speculate about how an argument would fare when given in different situations or to different audiences, and so on. This is not to say that we agree on the identity conditions of arguments – by no means. But argumentation theorists do not take the identity conditions to be so stringent that arguments are not repeatable.
But holding to (1) – (4), ~(Y), and (C) has significant consequences for argumentation theory. Assume that the only three plausible candidates for arguments are some sort of act, expression, or abstract object. I know of no attempt to define argument that does not fall into one of these three categories (though I can find you various works where a given definition in one place puts arguments in one ontological category, but in another place puts arguments in a different ontological category – oops!). But given (C) we should also accept, what I take is a controversial claim in argumentation theory, viz.:
(*) Arguments are abstract objects.
The reason is simple. Neither acts nor expressions are repeatable. I raise my hand. I raise my hand again. While I performed two acts of the same type, I did not perform just one act – one act happened before the other and temporal location is one of the identity conditions of acts. Similarly for expressions: the first ‘the’ on this page may be the same type of symbol as the second ‘the’, but the two ‘the’s are not one and the same expression – they are located in different places and composed of different molecules of ink. Abstract objects of various stripes, on the other hand, are repeatable – informational content construed as propositions say, or act types or expression types which are properties. Hence, adding (C) to our list of agreed upon principles brings with it a commitment to arguments being a kind of abstract object.
Note that it does not commit us to a particular type of abstract object. Hence, those who favour act talk might opt for act types over propositions. I suspect that such solace will be short lived, for though I will not argue it here, I strongly suspect that any appeal to act types, to get the typing correct, will ultimately appeal to the informational content. For example, my giving Anselm’s argument in a high falsetto in English while someone else presented Anselm’s argument in booming Danish will count as instances of the same act type, for the purposes of identifying arguments, in virtue of the informational content presented since most of the other act types these two particular acts fall under do not overlap.
Regardless, I am not here trying to argue for the truth of (*), but merely to show that given (1) – (4), commitment to ~(Y) and (C), short of finding another ontological option for arguments beyond the three standard ones used in argumentation theory, commits one to (*). If arguments as abstract objects cannot be tolerated, one is free to reject that arguments are repeatable (and live with the consequences) or even to reject ~(Y) and just give up on arguments altogether and focus, in one prefers, on, say, arguings and types of arguings instead.
On the one hand I have made no progress on the list of issues I used as examples at the beginning of this paper. The principles we, as argumentation theorists, agree upon so far, are too minimal to help us resolve these issues. But I do hope that I have at least provided four possible avenues for moving forward. Firstly, we could try to find more principles that argumentation theorists agree upon. (Perhaps one might try to appeal to the principles offered in George Boger’s “Some Axioms Underlying Argumentation Theory”? I suspect however that the tenets he gives are not generally agreed upon or non-contentious, even if widely accepted within one strain of argumentation theory.) For example, I strongly suspect that argumentation theorists also agree on some principles roughly like the following:
(D) All arguings involve the expressing/giving of a claim.
(E) All arguers have some goal to be achieved by arguing.
(F) Some arguings happen for the purpose of changing belief, promoting action, convincing, persuading, demonstrating.
One can hope that finding more agreed upon principles will generate a better basis for adjudicating disputes. Note however, that even adding (D) – (F) to our list of agreed upon principles does not change the results of section 4.
Secondly, we could deny that there are arguments and focus instead on arguings, reason/claim expressions, and the informational content of such expressions (and the relationships and uses and types) of each and see if dissolving talk of arguments also dissolves the original problems. Thirdly, we could deny that arguments are repeatable and trace out the consequences for argumentation theory. Fourthly we could accept that arguments are repeatable and focus on arguments as abstract objects and trace out the consequences of that. For example, it is not at all clear that arguments as abstract objects can have missing premises – perhaps the expressions of the arguments in texts can have missing components (given the arguments we take those expressions to express), but the arguments themselves cannot. Hence, commitment to (*) might also commit one to ‘enthymeme’ not being a property of arguments at all. I leave it up to you which path you shall follow.
i. John Woods (1992) appeals to similar principle with regards to relevance – any theory of relevance that makes everything relevant to everything or nothing relevant to anything is to be rejected.
ii. The issue is made more complicated by the problem of trying to type acts or identify the identity conditions of an act – could act x have happened two minutes later and still be the same act? On some theories of the nature of acts the answer is ‘no’, but on others it is ‘yes’.
Blair, J.A. (2003). Towards a Philosophy of Argument. In J.A. Blair, et. al. (Eds.), Informal Logic @25: Proceedings of the Windsor Conference (CD-ROM). Windsor, ON: OSSA.
Boger, G. (2005). Some Axioms Underlying Argumentation Theory. In D. Hitchcock (Ed.), The Uses of Argument: Proceedings of a conference at McMaster University, 18-21 May 2005 (pp. 40-47). St. Catherine’s, ON: OSSA.
Churchland, P. M. (1981). Eliminative Materialism and the Propositional Attitudes. Journal of Philosophy, 78, 67–90.
Finocchiaro, M. (2003). Dialectics, Evaluation, and Argument. In J.A. Blair, et. al. (Eds.), Informal Logic @25: Proceedings of the Windsor Conference (CD-ROM). Windsor, ON: OSSA.
Gilbert, M. A. (1997). Coalescent Argumentation. Mahwah, NJ: Lawrence Erlbaum Associates.
Gilbert, M.A. (2003). But why call it an argument?: In Defense of the Linguistically Inexplicable. In J.A. Blair, et. al. (Eds.), Informal Logic @25: Proceedings of the Windsor Conference (CD-ROM). Windsor, ON: OSSA.
Hitchcock, D. (2009). Informal Logic and the Concept of Argument. In D. Jacquette (Ed.). Philosophy of Logic (pp. 101-129). Amsterdam: Elsevier.
Lunsford, A., Ruszkiewicz, J., and Walters, K. (2013). Everything’s an Argument, 6th ed. New York: Bedford/St. Martin’s.
Merricks, T. (2003). Objects and Persons. Oxford: Oxford University Press.
Turner, J. (2011). Ontological Nihilism. Oxford Studies in Metaphysics, 6. Oxford: Oxford University Press.
Willard, C. A. (1989.)A Theory of Argumentation. Tuscaloosa and London: University of Alabama Press.
Woods, J. (1992). Apocalyptic Relevance. Argumentation 6, 189-202.