Comparing the Argumentum Model of Topics to Other Contemporary Approaches to Argument Schemes: The Procedural and Material Components

See the original example in Perelman and Olbrechts-Tyteca (1958: 115), who declare to report the argument by which Timon highlights the value of pamphlets: “L’orateur parle aux députés, le publiciste aux hommes d’état, le journal à ses abonnés, le Pamphlet à tout le monde… Où le livre ne pénètre pas, le journal arrive. Où le journal n’arrive pas, le pamphlet circule”.

Not coincidentally, the argument scheme, inasmuch as it connects the premises to the standpoint, though being a typical constituent of the argumentation stage, also plays a relevant role in the opening stage as for the premises evoked to support the standpoint.

In our first works on topics (Rigotti and Greco Morasso 2006; Rigotti 2006), we introduced the notion of hooking point to indicate the link between the argument and the standpoint. Later on, revising the medieval treatises on topics, the notion of habitudo has emerged as particularly insightful. The difference, though not substantial, is significant, as assuming the notion of habitudo means assuming a different focus. In fact, the notion of hooking point left implicit the relationship (between the world of the standpoint and the world of the argument) and only focused on one relatum, while habitudo highlights the relationship itself, thus involving both relata. The habitudo, by the way, turns out to precisely coincide with the locus.

The traditional philosophical concept of ontology was meant to deal with questions concerning what entities exist or at what conditions they can be said to exist, and how such entities can be grouped, related within a hierarchy, and subdivided according to similarities and differences (for example, by means of a tool like the Porphyry’s tree, see Rigotti and Greco Morasso 2006). For the AMT, the term “ontological” builds on the following three notions: (1) “ontology of social reality” (Searle 1995), meaning a network of (institutionalized) commitments that create specific interaction fields (see also Rigotti and Rocci 2006); (2) ontology as it is understood within computer science and information science, i.e. a formal representation of a set of concepts within a domain and the relationships between those concepts, that may be used to define the domain and to reason about its (constitutive) properties; (3) the notion of ontology entailed by linguistic semantics, which is particularly relevant for the AMT. The role of ontology in semantics has been discussed in depth in Jackendoff (1983, 1990). Indeed, dealing, for instance, with the problem of reference, natural language semantics needs to postulate an ontology of some sort. Model-theoretic semantics postulate a very sparse formal ontology featuring only individuals, sets and worlds. It can however be questioned whether such an ontology is fully adequate for natural language semantics. The study of referential expressions shows that language(s) seem to require very specific “ontological presuppositions” (Jackendoff 1983) or “metaphysical assumptions” (Bach 1981: 79). For instance, natural languages force us to recognize things such as events, to set them apart from states, and cut even finer distinctions between different types of events (Vendler 1957; Bach 1981). Early linguistic contributions such as Whorf (1997 [1956])—on Hopi vs. English metaphysics—had seen this basic commonsense ontology as eminently culture and language specific, but most contemporary semanticists would maintain with Bach (1981) that at the level of the most basic ‘world furnishing’ we are interested in the commonsense ontology that is inter-culturally shared and is primarily bound to our common experience of the world.

In this relation the terminology adopted by logicians is oscillating: Layman (2002: 20 ff.) uses the term “argument form”, Haack (1978: 201), Barth and Krabbe (1982: 156), Hughes and Cresswell (1996: 25), Epstein (2001: 196) use the term “rule” (e.g.: rule of modus ponens).

Compare Braet (2005: 66): “The topical principles from the Rhetoric [by Aristotle] will be regarded as the core of a modern argumentation scheme. For example the principle ‘If the cause is present, the effect must occur’ (Rhetoric 2.23.25) forms the if–then statement in the causal argumentation scheme ‘If the cause is present, the effect must occur; well then, the cause is present, therefore the effect will occur’”.

In fact, the symptomatic argument, very common in all forms of explanation (for example, in the formulation of medical diagnoses), starts from the ontological relation connecting an effect to its (possible) cause. The relation between effect and cause is normally not necessary, since the same effect might be produced by different causes. Yet, ideally, what a good symptomatic argument should do is to identify the most probable (ideally the unique) cause for a given effect.

In other words, a locus is constituted by a class of maxims generated by the same ontological relation.

Maxims as such are by definition true. Rigotti (2008) introduces the term paramaxims to indicate those tentative or pretended maxims that, as they are false, cannot properly be considered maxims.

Structure-dependency presupposes a distinction between structured and unstructured wholes. Hamblin (1970) introduces an analogous distinction between physical and functional collections. Peter of Spain (Summulae Logicales, 5.7;5.14-5.23; in particular 5.14-5.18) analogously distinguishes between totum universale and totum integrale. Interesting remarks are put fore by Buridan (Summulae de dialectica 6.4.2 ss., see Klima 2002).

In this relation, we hypothesize that the fallacious or sound use of argument schemes is often not determined by their presumptive or probabilistic nature, but by an uncertain definition of their conditions of semantic applicability.

Here, it is first of all important to distinguish between contrary terms and contradictory terms (see in this relation the discussion in Petri Hispani Summulae logicales 3.32, cf. Bochenski 1947: 33). Moreover, even concerning contrary terms, already Aristotle, in his Categoriae, observed that contrary terms not necessarily are the opposite poles of a dichotomy, like odd and even numbers. In some cases, contraries admit intermediate terms, as in the case of white and black, which admit an infinite set of grey shades in between. This has been developed in the medieval doctrine of oppositions, which distinguished among contraria mediata (which admit some intermediate terms) and immediata (see Gatti 2000: 33).

Persuasiveness, indeed, is not exclusively based on the validity of a procedure. To put it synthetically, it is rooted on how much the actual arguments used are anchored in the “material” common ground of the participants in the critical discussion (Rigotti 2006). In this sense, a further component in the reconstruction of the inferential structure of arguments has to be included, accounting for this “aboutness” of the actual use of argument schemes.

Interestingly, also in the case of the locus from analogy, semantic analysis should be invoked in order to identify the proper conditions of its validity, concerning in particular the notion of comparability and the connected notion of functional genus. Comparability, indeed, is not mechanically established: it holds if it focuses on a relevant dimension of the concerned property. For example, in this case, the two celebrations are not claimed to have the same meaning; they are comparable as to the behaviors they provoke. Or, in another domain, a Federal State can be compared to a family, since member States, like family members, are expected to help each other; but we could not say that, in a family-like manner, member States get old and die…

Moreover, an adequate semantic analysis should distinguish, in connection with the different nature of the standpoints concerned, two main types of arguments from analogy: (1) arguments whose standpoints represent factual claims; (2) arguments whose standpoints represent evaluative claims. The maxims should be correspondingly specified: in the first type, the functional genus is justified by the emergence of the same cause (that may be more or less deterministic in nature) in the two compared states of affairs; in the second type, the functional genus is based on the applicability of the same criterion of evaluation. Our example clearly pertains to the first type of analogy. An example of the second type of argument from analogy would be: “the US Federal Government must assist the Detroit’s automakers as Detroit is being hit by an economic hurricane just as New Orleans was hit by a natural hurricane (Katrina)”.

It seems that the tradition of topics indeed neglected the notion of endoxon, perhaps merging it with maxim. But it is hard to imagine that Aristotle attributed to all people or to the majority of them or to the wisest ones etc. the shared explicit knowledge (or belief) of topical rules, even though these rules may become part of the acquired outfit of some of them. The cognitive status of the abstract, general inference rules discovered by argumentation theorists cannot be interpreted in terms of the prevailingly shared opinion. The ignorance of this fundamental component of Aristotelian topics is probably due the fact that Aristotle did not explicitly give any example of what he understood by endoxon. Numerous endoxa can, however, be reconstructed from the examples often given by the author when listing his topoi. Not coincidentally, in our opinion, Braet, aiming to reconstruct an ideal model of an Aristotelian locus, lists as its fourth relevant component, beyond the name, the suggestion of a fair procedure for establishing the concerned type of argument and the topical principle involved, an actual example (in our opinion, inevitably including an endoxon) to which Aristotle often applies this principle (Braet 2005: 69).

In this case, a categorial syllogism is activated.

It is interesting to observe that the term liaison seems to cover the Medieval notion of habitudo, albeit Perelman and Olbrechts-Tyteca seem to ignore the corresponding tradition.

For reasons of uniformity, we here adopt the phrase «argument scheme»; Walton and colleagues, however, prefer “argumentation scheme”.

Research on the AMT is currently developing within a series of projects directly or indirectly bound to the doctoral program Argupolis – Argumentation practices in context (www.​argupolis.​net), financed by the Swiss National Science Foundation - SNFS (Grant: PDAMP1-123089); Eddo Rigotti is the project leader and Sara Greco Morasso is the project coordinator. In particular, one of the research projects in which the AMT is being developed, directed by E. Rigotti, and in which S. Greco Morasso is involved as a consultant, studies argumentation as a tool for resolving conflicts in two quite different contexts: families and publicly listed stock corporations (Grant: PDFMP1-123093). Through the development of the AMT, Eddo Rigotti is contributing to a research project directed by Andrea Rocci and entitled "Modality in argumentation. A semantic-argumentative study of predictions in Italian economic-financial newspapers”, also funded by SNFS (Grant: 100012-120740/1). Rigotti is equally involved in another SNFS project led by Andrea Rocci which investigates the argumentative function and rhetorical exploitation of keywords in corporate reporting discourse (Grant: PDFMP1_124845). Finally, Rigotti is conducting a recently approved project studying the argumentative practices adopted by Swiss banks in order to comply with Anti Money Laundering and Counter Terrorism rules in Finance, while, at the same time, preserving the fiduciary relationship with the suspected client (SNFS, Grant: CR11T1_130652/1).

.