Snoeck Henkemans published Analyzing complex argumentation in 1992 as a doctoral dissertation. The second edition appeared in 1997. The puzzle she set about solving is mostly solved by the end of the book, and the manner in which this is done is quite compelling. However, some small yet not clearly unimportant questions are still in the air. Let us see.
The big issue is this: how does an analyst produce a satisfactory identification & analysis of more sophisticated or elaborated forms of arguing. It is obvious to any student that the two premises - one conclusion story is something of an Aristotelian myth. More often, arguers choose to put forward something which, as the title says, can be considered an instance of complex argumentation. To be sure, even if this did not happen as often as it does, the question would still arise. One might choose to boil it down to the following question: When more than one argument is put forward, how do we represent the way in which it is put forward? In this form, it is clear that the answer to this question will act upon the evaluative part; different analyses might give rise to different evaluations.
First, as the academic modus operandi goes, Snoeck Henkemans identifies the approaches that, in the past, have dealt with the same problem. She divides them into structural and functional approaches, according to whether they distinguish complex acts on the basis of form (i.e. structure of the product “complex argumentation”) or role (i.e. function of the process “complex argumentation”). The authors discussed do not always fit this nice division, but I guess the point of it is to be practical rather than entirely correct.
Now, Snoeck Henkemans finds some flaw or another in each of the members of the two camps (5 in the first, 3 in the second). In most cases, these flaws follow from vague theorizing such as saying that one premise is “dependent” if it offers “some” “support” to the conclusion, or that the three or more premises “converge”. I am not going to go into everything that is written as a commentary to other theories. Amongst them, Freeman’s Thinking logically (1988), which is presented as an “elaboration” of Toulmin’s The uses of arguments (1958), seems to be the closest to one idea Snoeck Henkemans takes as a starting point. The idea is that in order to decide which argument structure one is dealing with, one should search for the (imagined or explicit) critical questions that each argument in the structure responds to.
Nothing could be more handy for the pragma-dialectical approach than explaining argumentative elaborations dialogically. In this sense, the subject of complex argumentation – if the functional approach is fully endorsed – is itself very… pragma-dialectical. The definition goes like this:
Advancing complex argumentation amounts to performing at least two argumentative speech acts. If these speech acts constitute one attempt at convincing the listener of the acceptability of the standpoint, the defence is a case of coordinatively or subordinatively compound argumentation. If the speech acts constitute more than one such attempt, the argumentation is multiple (Snoeck Henkemans, 1997, p. 73)
The differences between coordinative, subordinative and multiple argumentation are explained with reference to the role the arguments involved in such structure play within the discussion. I mentioned this interactional approach briefly – and, looking back, in a rather rococo fashion – in this post. Again, as the academic practice goes, since at that time there was already a model halfway through this idea on the market, Snoek Henkemans discusses it thoroughly. The model in question consists of Edmondson’s exchanges (“Proffer”, “Satisfy” & “Contra”), developed in Spoken discourse: A model for anaylsis (1981).
The most simple instance of argumentation goes like this:
(1) A: This blog, Argumentics, is kinda crappy
B: Why is that?
A: Well, it hasn’t published anything in 52 months!!!
B: Oh yeah, you’re right. Booooo.
This example falls outside the scope of Snoeck Henkemans’ research. But take the next one:
(2a) A: This blog, Argumentics, is the best thing that ever happened to human race
B: Geez, calm down. Why is that?
A: Well it reviews stuff so tremendously ingeniously!
B: No it doesn’t!
A: Oh yeah, you’re right. But hey, at least is free!
In the second example, A is not withdrawing his standpoint, B is maintaining criticism, and A is advancing a different argument. This is an instance of multiple argumentation. It is called multiple because the arguer “undertakes more than one separate attempt to defend his standpoint” (1997, p. 81), either because his previous argument is rejected and he accepts this, or because he’s going for an overkill, like in:
(2b) A: This blog, Argumentics, is the best thing that ever happened to human race
B: Geez, calm down. Why is that?
A: Well it reviews stuff so tremendously ingeniously!
B: Oh yeah. You know what. I too think it is the best thing that ever happened to human race
A: I freakin’ know! And it is free and sooper cool!
But, unlike multiple argumentation, it can occur that the arguer (A) does not retract his first argument. When this happens, compound argumentation occurs. This is of two types: subordinative and coordinative. As promised, these are distinguished by their function within the discussion.
Let’s start with the subordinative argumentation. (I should say first that the full name of the sub-type is subordinatively compound argumentation)
(3) A: Lady Gaga is so amazing
B: Why do you think so, dad?
A: Well, for one thing she has all this brill gear, man!
B: WTF? Meat-costume, brill gear?
A: Yeah, man, that’s art, man, art.
B: Dad, I’m 5. Why do you keep calling me “man”?
A: I don’t know, son. I’m sorry. I got a bit carried away. But I tell ya: the schmutter is utter!
Rather straight forward. The argument is advanced by A, the argument is rejected by B, the argument is supported by A with another argument, and this support (one might say, the support of the support) gives rise to the subordinative composition.
The coordinatively compound argumentation goes like this:
(4) A: Rocky II is a very retarded movie
B: I wouldn’t say it is very retarded. Why do you say that?
A: We’re looking at a guy training his head off for 45 minutes. Don’t you think that qualifies as very retarded?
B: Not really. This is the idea of the movie.
A: But he is in a freaking meat factory!
B: Meat packing factory…
What happened there is that, for some reason or another, A’s argument was rejected. B does not withdraw the argument, nor does it support it with a sub-argument, but advances another argument which claims sufficiency along with the first one. Remember: The difference between multiple and compound is not in the structure (since in both cases two arguments have been put forward), but in the interaction (since, in the case of multiple argumentation a withdrawal has been performed during the interaction).
The rest of the book is the most interesting part. It is concerned with the passage from the raw text to these fixed structures. How do we recognize a SCA, how do we recognize a CCA? Sometimes, it is the standpoint (modality, quantification), sometimes it is a qualifying adjective, sometimes it is the explicit (mentioning of a) criticism. Yet again, sometimes, it is the bigger context of the dialogue: e.g. the acceptability of one of the speech acts put forward during the interaction. All these are studied. Attention is also paid to the manner in which felicity conditions (translated in pragma-dialectics as the two Principles – of Interaction & Communication) help us see what is going on in a piece of discourse. But this is the story is studied in greater detail in van Eemeren, Grootendorst, Jackson & Jacobs (1993). Here, Snoeck Henkemans only focuses on how such pragmatic knowledge (for the analyst’s part) might help to pinpoint this or that structure.
The biggest objection I have to all this is the suggestion that, in the end, all argumentation is reducible to these structure. I think one could very well agree with the fact that all complex argumentation is reconstructible as MA, SCA & CCA, but this wouldn’t mean that even more elaborate versions might benefit from the same patterns. In a way, I am merely continuing Snoeck Henkemans train of thought: What suggested to her that, as opposed to the Toulminian form, more elaborate forms of argumentation need to be identified in order to account for some interactional aspects, might be taken a notch further and turned into the suggestion that even more elaborate forms could be identified if more sophisticated interactional aspects are left aside. The outcome would be the same: It doesn’t mean that one cannot reduce the very-very complex structures to the (merely) complex pre-defined forms, it just means that the passage from very-very complex structure to the very-very complex pre-defined forms would be more beneficial, i.e. would explain more.
