Analysing and systematizing Walton’s critical questions
Annie de Jong
Universiteit van Amsterdam
Faculteit der Geesteswetenschappen
Communicatie en informatiewetenschappen: Rhetoric, Argumentation and Philosophy 10 June 2019
Supervisor: Dr J.H.M. (Jean) Wagemans Second reader: Dr. B.J. (Bart) Garssen
Table of contents
1. Introduction 3
2. Walton's view on critical questions in context and other theories of critical questions 7
2.1 Theories of critical questions 7
2.2 Walton's argument schemes and critical questions 10
2.3 Problems with Walton's view 13
3. The Periodic Table of Arguments as an analytical tool 16
3.1 Propositional content and justificatory force 16
3.2 The Periodic Table of Arguments 17
4. Analysing critical questions 21
4.1.1 Argument from expert opinion version 1 21
4.1.2 Argument from expert opinion version 2 24
4.2 Argument from pragmatic inconsistency 26
4.3 Argument from sacrifice 28
5. Conclusion 31
1. Introduction
In the theory of argumentation, different ways of classifying types of arguments have been put forward. These ‘types’ of arguments are also called argument schemes. Argument schemes are a schematic representation of an argument, built up from any varying number of statements. Of these, one statement usually functions as the conclusion, standpoint, or claim. The other statement(s), also called premise(s), function(s) as support for this conclusion. How these supporting premises defend the conclusion differs among the various distinguished types of arguments, as well as between theories. Prominent authors to write about argument classification and argument schemes are Hastings (1963), Perelman and Olbrechts-Tyteca (1969), Schellens (1985), Kienpointner (1992), Walton (1996), Blair (1999), and van Eemeren and Grootendorst (2004).
Besides the classification of arguments, one of the central questions in the theory of argumentation asks how we decide which arguments are good ones and which ones are not. Evaluation of arguments is important if we are to make any sort of distinction between what are good reasons to change our beliefs or actions, and which are not. However, there is a lot of disagreement on what makes arguments good ones, and the rationale we choose to judge by affects whether we do or do not find arguments acceptable.
One of the proposed methods of evaluation is using critical questions. Critical questions are questions put forward when a dialogue partner has doubt about some part of the argument. This tool for evaluation is especially used when judging arguments on the level of the scheme. Some authors formulate very general questions, that can apply to multiple argument schemes. Others assume each type of argument, and therefore each argument scheme, has its own specific set of questions.
When it comes to the research done on argument schemes and critical questions, Douglas Walton is arguably one of the most prolific authors. He has written handbooks on argument and debate1, as well as being influential in the domain of artificial intelligence. His theories are used a considerable amount in computational argumentation (Visser et al., 2018, p. 314). His work in general has been prominent enough to license a book reflecting on his thought.2 Blair mentions that argument scheme theory, which he views as “a combination of the Toulmin notion of warrant as inference license, the Perelman and Olbrechts-Tyteca emphasis on argument schemes, and Hastings idea that critical questions can be associated with argument schemes and serve as the basis for evaluating arguments that instantiate them” are “all developed most fully by Walton.” (2012, p. 132)
Although Walton has written a lot on argument schemes and their associated critical questions, a problem that arises with his view is that there is no systematic way in which critical questions are connected to their respective argument schemes. Hoppmann notes that as of yet, there is no system to critical questions and that “[i]t remains a strong theoretical desideratum to revisit, contrast and unify [Hastings’s, Schellens’s, Kienpointer’s and Walton’s] lists in an inductive approach, or (preferably) to find a deductively sound way to deduce more systematically reliable lists of scheme-dependent Critical Questions.” (2013, p. 9) A necessary
1 See for example Walton (1989), (2005), (2008). 2 See Reed & Tindale (2010).
step then, to discover how critical questions are systematically connected to argument schemes, is to find out what aspect of the argument (scheme) critical questions pertain to.
A generally accepted distinction in the literature on evaluation is that of (the content of) the premise supporting the conclusion, and the relation between this premise and the conclusion.3 These two concepts are sometimes referred to as propositional content and justificatory force (see for instance the description of pragma-dialectics in Hitchcock &
Wagemans, 2011, p. 201). Propositional content is that aspect of the argument that in the classical logical tradition is called the ‘minor premise’. Other names for the relation between premise and conclusion are “inference license”, “warrant”, “major premise”, “argumentative mechanism” and “unexpressed premise”.
These terms identify the two concepts that are generally the objects of the evaluation process when looking at the microlevel of the argument scheme. The examination of the premises is done in light of their propositional content. The relation between the premises and the conclusion is judged for its justificatory force (to what extent do the premises offer a justification of the conclusion). Taken together, many theorists agree that propositional content and justificatory force are (at least among) the relevant elements by which an argument’s goodness can be judged.
The central question in this research is what aspects of an argument scheme critical questions relate to. The goal is to sort all critical questions under review into the categories of propositional content and justificatory force. If it transpires a question does not relate to either of these, I will label it ‘other’. In this way, it should become clear whether the critical questions systematically connect to argument schemes in virtue of these two aspects, propositional content and justificatory force, which occur in all argument schemes.
To find out whether it is the case that propositional content and justificatory force are what critical questions relate to, we need a framework to which we can ‘translate’ the schemes and questions which makes the distinction between these two different aspects. A tool that is able to make this distinction is the Periodic Table of Argument (PTA) (Wagemans, 2016). The PTA systematically classifies types of arguments. At present, the PTA can be used as a descriptive tool, but not yet as an evaluative one. By using this system of classification to study the critical questions, I believe it will be possible to begin to understand what aspect of an argument the critical questions relate to. Once we grasp this, we may start to create a typology of critical questions. In this way, this research contributes to the theoretical grounding of critical questions, connecting them to their related argument schemes in a systematic way.
The reason the Periodic Table is useful for our purposes, is that it systematically differentiates conclusion, argument, and the relation between these two.4 The argument is directly represented in the Table by the premise. The relation between the argument and conclusion can be derived and is described by the argumentative mechanism. For instance, an
3 This distinction descends from the soundness criteria in traditional logic, where premises are judged for their truthfulness, and the relation between the premise and the conclusion for its validity. Many normative theories of argument, such as dialectical theories, informal logical theories, and epistemic theories of argument use this framework in their ideas on argument evaluation, albeit with their own interpretations of these aspects. 4 Note that I use the word ‘argument’ to denote what others may call a ‘premise’. In my use, the word argument
generally does not include the point of view being argued for. This point of view I call a standpoint or conclusion, as above.
argument from effect could be the following: The suspect was driving fast, because he left a long trace of rubber on the road. (Wagemans, 2019, p. 65) In this case, the conclusion is “The suspect was driving fast.” The argument is directly stated: “Because he left a long trace of rubber on the road.” The relation, or argumentative mechanism, can be derived: “Leaving a long trace of rubber on the road is an effect of driving fast.”
It is in theory therefore possible to ‘translate’ other theories of argument in terms of the PTA, which also allows for comparison between classifications. Translating the critical questions in terms of the PTA should allow us to see whether these questions pertain to propositional content or justificatory force, or neither of these. In this way, the Table functions as a sieve, allowing us to see which critical questions are systematically connected to the argument scheme and which, though perhaps relevant, are not appropriate at the microlevel of the scheme.
It is important to realize this difference between the different ‘levels’ of evaluation. It is possible to take contextual elements into account, for example, when evaluating an argument. However, it is commonly accepted that when it comes to the micro-aspect of the argument, namely the scheme, the relevant elements to evaluate are the propositional content and justificatory force of the argument. The question, of course, is if the critical questions indeed do test these aspects. It is very possible there will be questions which cast doubt on some contextual aspect. This is another reason to use the PTA as our ‘sieve’, since it will separate these types of questions from the ones that relate to propositional content and justificatory force, if any.
It is not feasible to do such an analysis for all of the schemes Walton lists within the scope of this paper, since they number around 60.5 Rather, this paper will be explorative, in the sense that I select three examples of Walton's schemes, and apply the proposed method of analysis above to these examples. These are not examples of the three categories into which Walton sorts his argument schemes. Instead, they are selected on the basis of considerations pertaining to criticism that has been levelled against critical questions. This critique concerns what aspect of the argument critical questions relate to, as well as questions concerning the way in which the critical questions are phrased, what types of answers are required, and how many questions are sufficient.
In analysing these three problem groups, I focus on finding an answer to the question what aspect of the argument the questions relate to. Understanding how the questions are connected to the schemes might point us in the right direction where to look for answers to the other questions that still surround critical questions. In this way, this paper can be seen as a pilot study which may eventually be extended to cover all schemes as formulated by Walton.
The first group consists of schemes that have a relatively long list of critical questions, perhaps because they have been discussed more often in the literature on argument schemes. An example of this is argument from expert opinion, which Walton et al. list as having six associated questions (Walton, Reed & Macagno, 2008, p. 310). This is quite a large amount, considering many others have ‘merely’ between one and three questions specified.
The second group consists of schemes that have critical questions which seem to relate to very different aspects of the argument. For example, argument from pragmatic inconsistency
has a critical question that pertains to the rhetorical force of the speaker (CQ1); a question that relates to the locution of the argument (CQ2); and a question that probes the utility of the argument scheme (CQ3) (Walton, Reed & Macagno, 2008, p. 336 – 337). Hornikx (2013, p. 134) points out the latter is not functional for differentiating a good version of an argument from a bad one.
The third and last group that can be identified consists of schemes for which no critical questions are listed at all, for example argument from sacrifice (Walton, Reed & Macagno, 2008, p. 322). Does this mean these schemes cannot be evaluated? Presumably, this is neither the case nor something Walton contends, which means questions must still be formulated for this category. The question is, what should these questions be and what part of the argument should they call into question?
In chapter two, I will provide the context of this research, describing what other authors have written about schemes and especially critical questions. It should become clear what ideas there exist about the role of critical questions in different theories. I then set out in more detail Walton’s view of argument schemes and critical questions, so that we can understand where his views fit in, and where they coincide and differ with others’ perspectives. The chapter concludes with a discussion of problems that exist with Walton's theory. Our central problem thus begins to come into view.
In chapter three, I propose a method for solving the issue under consideration. I explain the theoretical concepts of propositional content and justificatory force, and how we can use the Periodic Table of Arguments to help classify the critical questions. With the problem clearly stated and having suggested and defended a method to solve this problem, we can begin the analysis.
In chapter four, this analysis will be carried out for three argument schemes’ critical questions. As mentioned, there will be an example for each problem group identified. For the first group, which has a relatively large amount of questions, I will examine argument from expert opinion. The second group, which has questions that relate to multiple aspects of the argument, will be represented by argument from pragmatic inconsistency. For the third group, which has no questions listed as yet, I will take argument from sacrifice.
The results of this analysis are discussed in the final chapter. After a brief summary of the findings, I situate these in the broader context of research on critical questions. I present some solutions to problems with Walton’s view by hypothesizing some rules for systematizing his schemes and questions. I also go into implications this research has for extending the PTA with an evaluative component. Finally, I offer some suggestions for further research.
2. Walton's view on critical questions in context and other theories of critical questions
2.1 Theories of critical questions
There are many authors who have written on the topic of argument schemes. The way they are distinguished from one another can differ considerably, as well as how many different schemes are postulated. Pragma-dialectical theory for instance lists three kinds of schemes (as well as subtypes and variants of these; see Hitchcock & Wagemans, 2011), which Kienpointner does as well (see Walton & Macagno, 2016, p. 4), but based on entirely different criteria. Walton’s method of differentiating schemes is similar to pragma-dialectical theory, yet he lists around 60 individual schemes.6
What researchers do agree on, is that argument schemes are a formal way of describing a ‘type’ or a ‘kind’ of argument. Components that a scheme is generally held to consist of include one statement that functions as a conclusion. This statement, which a protagonist does not assume will immediately be accepted by a dialogue partner or audience, is then in need of support. This support is the second component of an argument scheme and consists of one or multiple other statement(s) which function as arguments or ‘premises’. These statements will be directly acceptable to the audience, at least so the protagonist supposes.
In a dialectical theory of argument, these two parties - the protagonist putting forward the argument and an antagonist or audience – are essential to the working of an argument. For, the theory goes, if there is no doubt or objection to what someone has stated, there is no need for argument. These objections thus play an important role, especially in the evaluation of an argument (see also Blair & Johnson, 1987, p. 45). Objections are conceived of in a number of different ways, but generally they are assertions meant to dispute or at least cast doubt on what the protagonist has put forward.7
As a way of judging arguments, the concept of critical questions has been put forward as a tool. Critical questions are usually somewhat differentiated from objections in the sense above, in that they are put forward by an antagonist in interrogative from. As such, researchers tend to conclude that they have slightly different argumentative significance and that they lead to other consequences. Not everyone agrees with this distinction, however.8
An example of an argument scheme is the following (Walton, Reed & Macagno, 2008, p. 315):
Argument from analogy
Similarity premise: Generally, case C1 is similar to case C2. Base premise: A is true (false) in case C1.
6 Although, as we will see below, Walton divides these 60 into 4 categories, it cannot be said that these four categories embody a ‘main’ type of argument, of which the 60 are all subtypes, as is the case for some other theories.
7 See for instance the pragma-dialectical conception of ‘difference of opinion’ in van Eemeren & Grootendorst (2004, p. 59).
8 See for instance Krabbe & van Laar (2011), who do not explicitly name critical questions, but discuss critical reactions in such a way that indicates critical questions can be understood to be a part of that category.
Conclusion: A is true (false) in case C2.
We see here that the first component, the conclusion, is supported by the second component, which consists of two premises. For this scheme, the following questions have been put forward (Walton, Reed & Macagno, 2008, p. 315):
CQ1: Are there differences between C1 and C2 that would tend to undermine the force of the similarity cited?
CQ2: Is A true (false) in C1?
CQ3: Is there some other case C3 that is also similar to C1, but in which A is false (true)? The questions are meant to test some part of the argument, doubting precisely that of which the protagonist wants to convince the questioner.
Besides what purpose the questions serve in the dialogue, another unresolved issue is how precise the critical questions should be. Some authors feel that each argument scheme should have its own individual list of critical questions (see for instance Pinto, 2001, p. 104; Garssen, 2001, p. 91), whereas others feel it is not possible to formulate fixed lists of questions (see for instance Blair, 1999, p. 12). This difference also reflects on ideas about how many critical questions a scheme should have.
Critical questions appear to find their origin with Hastings (1963). In his Ph.D. thesis, he specifies a number of argument schemes. Along with these schemes, Hastings listed a number of corresponding critical questions. These questions are meant to probe the strength of the argument, thereby being able to judge whether the argument is an acceptable one. According to Walton et al., “[t]hese features set the basic pattern for argumentation schemes in the literature that followed.” (2008, p. 9)
Although much has been written in the literature that followed, there is hardly a consensus on what critical questions are or should be. Blair suggests that the function of critical questions is to “outline in a general way for any scheme the principle classes of exceptions to it, and so to provide useful general, if not universal, guidelines for the employment of that scheme.” (1999, p. 11) He believes that these questions cannot be formulated ‘universally’, but that they can be used practically in a given situation, as long as we keep in mind the same questions may not have bearing on another situation in which the same argument scheme is used.
Blair later writes that critical questions can serve multiple functions, testing 1) whether the premises are true or otherwise acceptable, 2) whether that type of reasoning is prima facie plausible, 3) whether the inference from the premises to the conclusion in the given case is actually warranted, 4) whether there are independent reasons for rejecting the conclusion, and 5) whether the argument employed is appropriate in the situation in question (2012, p. 130 – 131, emphasis in original).
Hoppman (2013) discusses various authors, including van Eemeren and Grootendorst (1992), Walton (1996), Govier (1999), and Schellens (1985). He writes that in the pragma-dialectical view, each type of argument has a fixed set of critical questions, which question the connection between the conclusion and the argument. This theory has in common with Walton’s approach that the questions arise from the protagonist’s burden of proof, which they have
incurred as a result of putting forth a standpoint. According to Hoppmann (2013), Govier (1999) distinguishes between a number of different types of objections, which are assertions rather than questions, and as such, antagonists putting forward these objections incur their own burden of proof, which – according to the pragma-dialecticians for instance – an antagonist asking critical questions does not. Finally, Schellens (1985) formulates closed critical questions that can be answered either ‘yes’ or ‘no’, whereas other theories also formulate questions that need gradual answers (such as when asking to what extent a source for an argument might have knowledge of the proposition claimed), or other kinds of answers.
Pinto (2001) divides criticism into two types. The first relates to defects in the argument such that the inference “should not have been made in the first place” (2001, p. 103). In this case, the proponent fails to shift the burden of proof to their interlocutor. Critique that can be levelled of this type either attacks the first expressed premise, or the unexpressed generalization (2003, p. 102) – in other words, this criticism is levelled at the propositional content and justificatory force of the argument. The second type of criticism either undermines or overrides the inference, again meaning that the critique applies to the justificatory force of the argument.
Posing the questions has no normative force according to Pinto, but rather serves to guide the respondent where to look for overriding or undermining evidence (2003, p. 112). It seems to follow he believes the questions should relate to justificatory force only. It further follows, as he also points out, that critical questions do not have the function of shifting the burden of proof back to the proponent.
Overviews such as Hoppmann’s, as well as the one I have given above, show that clearly, there is no one way of conceiving of critical questions. As Hoppmann points out, this makes it difficult to use critical questions to evaluate arguments (2013, p. 1), as we cannot be sure which (kinds of) questions we should in fact be asking. In his article, Hoppmann suggests a
taxonomy of critical questions which are meant to help applied arguments critics to evaluate
arguments. Whilst his suggestions are indeed useful for this purpose, this does not solve the issue that we still do not have a theoretically motivated typology of critical questions. Hoppmann himself acknowledges this, as we have seen in the introduction.
Issues pertaining to critical questions in Walton's theory are also unresolved in other theories, for instance in pragma-dialectical theory. A problem that can be identified in the pragma-dialectical theory of critical questions is that, although they suggest that critical questions pertain to the connection between the argument and the standpoint, in some cases the questions ask after the propositional content of the premise (i.e. is what is asserted in the argument (factually) true, or at least acceptable if, for instance, it concerns a value judgement?). Garssen for example writes that “[t]he most central critical question pertains directly to the relation between the premise and the standpoint. The other critical questions pertain to additional factors that may hinder transference of acceptability.” (2001, p. 97) It is clear that, much like with Walton's theory, there is some ambiguity as to what aspect of the argument critical questions pertain to.
2.2 Walton's argument schemes and critical questions
Together with Chris Reed and Fabrizio Macagno, Walton authored Argumentation Schemes, in which is listed 60 argument schemes, along with their associated critical questions. In the book, the authors suggest a categorization of the schemes along three lines: 1) reasoning, which “is taken in a broad sense to include different kinds of sequences in which there is a chaining of inference” (2008, p. 348); 2) source-based arguments, which “comprises all those arguments where the argument is dependent on a source, an agent that is in a position to know something” (2008, p. 348 – 349); and 3) applying rules to cases, which “relate to a situation in which some sort of general rule is applied to the specifics of a given case[.]” (2008, p. 349)
Walton and Macagno later revised this categorization, dropping the class of reasoning and introducing in its place two new categories, ‘discovery arguments’ and ‘practical reasoning’ (Walton & Macagno, 2016). They point out that the criteria for a classification system depend on the purpose of the classification, echoing Blair, who more than fifteen years earlier already wrote that “[s]ystems of classification are relative to their purposes.” (1999, p. 20) The authors explain that there is a number of different ways in which one can go about such a classification and choose the way which they call ‘according to general category’ (2016, p. 7 – 9).
Though the description is somewhat vague, it seems on close reading to suggest that this way of classification is based on the relation between premise and conclusion used in the scheme – especially when contrasted with the other types of classification, which have more to do with the content of the premises, the nature or strength of the conclusion, and the purpose of the arguer. This manner of classification is in line with how argument schemes are distinguished in pragma-dialectical theory (Garssen, 2001, p. 91). Blair also agrees that “van Eemeren and Grootendorst are also right to use the evaluative function of reasoning schemes as their principle of classification[.]9 This is another way of using the warrants used in reasoning schemes as the basis of their classification, and recognizes the central role of these inference licenses in the conception of a reasoning scheme.” (1999, p. 18 – 19)
As for critical questions, Walton is of the view that they play two roles (Walton & Godden, 2005, p. 476). On the one hand, they have a part in the justification of argument schemes. On the other, they have a role to play in the evaluation of what he calls ‘schematic arguments’, which are individual instances of argument schemes. As for the first role, this seems to relate to what was said about the relation between the premise and conclusion above. The argumentative mechanism would be what differentiates between types of arguments and formulating the mechanism as a ‘bridging premise’ completes the argument scheme. Critical questions then test this mechanism.
Walton also refers to the critical questions as ‘topoi’, “offer[ing] the user ... a choice among strategies for probing into the weak points in an argument” (Reed & Walton, 2003, p. 202, cited in Walton & Godden, 2007, p. 280). This ties in with what Pinto suggests is their function (2003, p. 4). In this sense, critical questions could have a part in the justification of
9 Blair (1999) makes a distinction between reasoning and arguing. His reasoning schemes, however, do not differ importantly from what we so far have been calling argument schemes. Walton on occasion refers to
‘argumentation schemes’, but since I believe argumentative mechanisms pertain to a single argument, rather than an argumentation complex, I will refer to them as argument schemes. One argumentation complex can, in my view, comprise of more than one argument scheme.
argument schemes. However, since Walton does not explicitly go into this subject much further, it is difficult to say what his exact thoughts are on the matter.
As for the second role, he argues with Godden that critical questions provide necessary conditions for the acceptability of arguments in the case of schematic arguments, but not in the case of an argument scheme, since there may be context-dependent and practical considerations which have bearing on the decision to raise critical questions (2005, p. 478). Since the questions are not necessary conditions for argument acceptability, it follows that there need not be a fixed set of questions per argument scheme.
Further, Walton and Godden write that critical questions are not a sufficient condition for the acceptability of an argument either (2005, p. 479). Firstly, this is because these types of arguments are defeasible, and always subject to review in light of new information. Secondly, whilst critical questions contribute to the evaluation of arguments, they are not exhaustive of the evaluation process. There may be other aspects affecting the acceptability of the conclusion of an argument, even if all critical questions have been answered satisfactorily.
Another aspect of critical questions Walton discusses is whether or not they have a burden of proof attached to them. According to him, they can be posed in a weak way and a strong way. The weak way has no burden of proof attached to them for the respondent, and serves to shift the burden of proof to the proponent who has to reply to the question(s) successfully in order to regain the presumption for their argument. The strong way of posing them is as an objection, which rests on an implicit assumption. In this case, Walton argues the burden of proof lies with the respondent, as they have to defend the assumption on which their objection rests (Walton & Godden, 2007, p. 283 – 287).
He also relates the critical questions to Blair and Johnson’s relevance, acceptability and sufficiency (RAS) criteria for assessing arguments (2007).10 He writes
Critical questions are not supplied as an alternative to the R.S.A. standard; rather they are best seen as an application of it to arguments of particular types—arguments that involve distinctive patterns of reasoning—deployed in unique dialectical circumstances. Thus, the legitimacy of a critical question derives from the fact that it tests some aspect of its target argument against one of the R.S.A criteria. (2007, p. 277 – 278)
From the above, we may presume that Walton sees critical questions as specifications of the RAS criteria, which means that the questions pertain to (the acceptability of the) propositional content of the premises on the one hand, and the sufficiency and relevance of the premises as support for the conclusion on the other (which are ways in which justificatory force can be critiqued). As mentioned above, these two aspects are what is usually considered in the evaluation of arguments at the microlevel of the argument scheme.
This assumption is strengthened by the fact that Walton and Godden describe critical questions as functioning as rebuttals and defeaters (2005, p. 483), and later say of ‘undercutting defeaters’ and ‘rebutting defeaters’ that they can be seen as “applications of the sufficiency
10 Though I list the criteria as relevance, acceptability and sufficiency (RAS) and Walton lists them relevance, sufficiency and acceptability (RSA), the order does not make any difference.
criterion of cogency” (2007, p. 277). Presuming they are writing about the same concepts in both articles, this would entail that critical questions relate to justificatory force.
However, Hoppmann mentions that though Walton's critical questions pertain specifically to the argument scheme level, Walton also writes about criticism at a ‘dialectical’ and ‘global’ level (2013, p. 5). The dialectical level supposedly relates to the complex argumentation, presumably the larger context of dialogue in which the singular argument takes place. The global level purportedly pertains to the argumentative process. Hoppmann suggests that though Walton does not refer to the criticism raised at these levels as critical questions, it functions in a ‘highly parallel’ way (2013, p. 5). This is an important clue, as it suggests multiple aspects of the argument to which the critical questions might pertain, rather than just the scheme-level, where propositional content and justificatory force would be appropriate.
Yet this clear-cut view is not apparent in Walton’s (1996) book. Though Walton indeed writes about the dialectical and global level (1996, p. 14 – 15), it is ambiguous whether these two are indeed distinct, as he mentions both levels have to do with the dialectical situation and the larger dialogue. Either way, he seems to contend that the critique at the argument scheme level, or ‘local’ level, pertains to justificatory force and propositional content: “At the first level of criticism, one is mainly concerned with the premises and conclusion of the argument – what might be called its inferential structure and content.” (1996, p. 14, my emphasis) Inferential structure is where we locate justificatory force and content of course refers to the propositional content of the argument. As I mentioned in the introduction, this microlevel of the scheme is what I concentrate on in this paper.
It seems that Walton has contradictory opinions on the matter, and what he does in his analyses are also different to some of the things he writes. It is not clear, for example, what aspect of an argument he thinks critical questions relate to. With Reed and Macagno he writes, “The asking of a critical question throws doubt on the structural link between the premises and the conclusion.” (2008, p. 15) This evidently refers to the justificatory force of an argument. Yet a little further on it is also mentioned that “[i]n many cases there is a range of assumptions, all of which can be seen as acting as implicit linked premises.” (2008, p. 17) Linked premises in this case means coordinative, in the sense that one without the other is not sufficient to support the conclusion: both arguments are necessary in order for the argument to ‘work’. This suggests that critical questions relating, for instance, to the speaker are on the same level as it were as questions probing the propositional content and justificatory force.
In an analysis of argument from expert opinion, he states, “[I]t shows you what you
basically need as the core of the appeal to expert opinion.” (p. 20, my emphasis) The version of
the scheme is given below (Walton, Reed & Macagno, 2008, p. 19). Appeal to Expert Opinion
Major premise: Source E is an expert in subject domain S containing proposition A.
Conditional premise: If source E is an expert in a subject domain S containing
proposition A, and E asserts that proposition A is true (false), then A may plausibly be taken to be true (false).
Conclusion: A may plausibly be taken to be true (false).
This reads very much as if propositional content and justificatory force are what ‘counts’ at scheme level after all. However, when we look at the questions, there is a question that runs: CQ2: Field question: Is E an expert in the field that A is in? At first glance, this question is highly context-dependent, since what counts as an expert is different for each area of expertise.
The clearest answer given comes yet a few pages later, where it is explained that all the critical questions can be built into the scheme as additional premises (2008, p. 19 - 21). This means that there is no distinction between scheme level, dialectical level and global level, but that everything is on the scheme level. However, the authors prefer the option not to build them all into the scheme, but to leave the critical questions as they are, and that they can be used as topoi for analysts evaluating arguments. This ties in with Walton's thinking that not all questions need to be raised in every situation, as explained earlier. This does not mean Walton ignores the dialectical component in the evaluation of argumentation (2008, p. 218), and it seems he assumes critical questions can be raised for all three components: the scheme, the larger dialogue and the global context.
A final important point, that seems to contradict some of the things above, is that Walton recognizes that different critical questions relate to argument schemes in different ways (2005, p. 481). This means that there is no systematic way in which the critical questions are associated with their respective argument schemes. At the same time, he also believes critical questions probe types of arguments in a distinctive way and that there are standard critical questions appropriate for each scheme (2016, p. 2). If critical questions are specific to the argument schemes they are associated with, there must be some reason that they are specific to them, meaning there is some sort of order or method by which they are connected.
2.3 Problems with Walton's view
As mentioned above, there are some serious problems still unsolved when it comes to Walton's theory of critical questions. First, how the questions are connected to the schemes and which aspects of the arguments the questions relate to is still ambiguous. It is unclear which aspect of the argument is called into question by the critical questions, or if there are multiple aspects that are doubted. It is also uncertain if each questions connects to the argument scheme in the same way. Second, the questions are formulated in various ways, which means it varies what kinds of answers are required to answer the questions properly. This makes it more difficult to judge how the answers contribute to or detract from the quality of the argument. Third, critical questions are not always particular to a specific scheme. It is unclear whether or not Walton believes this should be the case.
As for the first problem, Wagemans (2011b) discusses Walton’s version of the scheme for argumentation from expert opinion. He points out that “from a systematic point of view, some criticisms can be made.” (2011b, p. 334) He argues it is not clear how the questions relate to the
scheme, that is, to which aspects or premises of the argument the questions appertain. Hornikx (2013) also wonders how the questions relate to the standpoint and how they cohere among one another.
Hansen (2010, p. 8) comes up with his own typology, based on Walton's (1996) and (2006) versions of argument schemes. He categorises Walton's critical questions into 5 types: those that pertain to either 1) premise acceptability; 2) premise sufficiency; 3) defeasibility; 4) the nature of the conclusion; and 5) none of the above, creating a ‘leftovers’ category.
He also notes a problem with the completeness of the sets of critical questions. According to him, all the listed schemes have associated ‘P-questions’ (premise acceptability questions), but only about a third have associated ‘S-questions’ (sufficiency questions). Furthermore, he “wonders why S-questions would be included [since the schemes are supposed to be structures that provides prima facie premiss sufficiency].” (2010, p. 11)
This idea that argument schemes provide prima facie support for their conclusions arises from Hansen’s notion that these schemes are in a ‘positive mood’. He writes,
If schemes are in the positive mood then they are such that any argument that instantiates a scheme (and has acceptable premises) will make its conclusion prima
facie acceptable. Such schemes, we noted, should not include S-questions since a
measure of premiss sufficiency is guaranteed in virtue of being an instance of the scheme. Neutral-mood schemes, by contrast, do not confer prima facie acceptability on their conclusions. To compensate for this, they must include S-questions along with other critical questions. (2010, p. 11)
This would suggest that though critical questions pertaining to propositional content (premise acceptability) are necessary to pose, questions relating to at least one part of justificatory force (sufficiency) are not. Both aspects, however, are listed in Hansen’s categorization.
Lastly, D-questions (defeasibility questions) “are directly related to presumptive generalizations [the warranting premise]” (2010, p. 14) and Hansen suggests there should be one D-question associated with every scheme (2010, p. 11) This is of note, since he appears to believe that there should then at least be some sort of question relating to justificatory force for each scheme. Whether or not one agrees with Hansen’s conclusions, his point remains that Walton's formulation of the critical questions is incomplete and this reflects in the issue that it is unclear to what aspect of the argument critical questions relate.
The second problem has to do with the form the questions take. Wagemans (2011b) points out some questions are formulated in a way that requires a ‘yes or no’ answer, whereas others are answered by something to a certain degree. Hornikx (2013, p. 134, my translation) raises a number of questions to this effect: 1) Are the questions formulated sufficient to be satisfied that, if answered positively, the argument is indeed a good one? 2) Does a positive answer imply a qualitatively better argument? 3) Are we able to answer the questions properly?
The third issue is pointed out by Hornikx (2013, p. 134) when he writes that if argument schemes are indeed to be distinguished by means of the argumentative mechanism involved, critical questions should be particular to the argument schemes for which they have been
formulated. He points out that this is not always the case and that sometimes questions are relevant for more than one scheme or that they concern the utility of the argument scheme itself.
Finally, Hornikx (2013, p. 134) explains that the normative status of the questions has not been clarified yet. I would argue that Walton (2007) goes into some detail on the normative foundation of critical questions, linking them to the RAS criteria, as explained above. He writes,
Thus, critical questions are a kind of evaluative topoi, providing a list of individually necessary conditions for the success of particular schematic arguments. On the other hand, because they represent only commonplace ways in which arguments of some schematic-type can default, they are not jointly sufficient conditions for the success of a schematic argument. (2007, p. 269)
Still, even though he provides a theoretical foundation for the normative status of the questions, I agree with Hornikx that it is still unclear how a positive answer to a critical question exactly contributes to the quality of the argument. As is clear from the quote, even if the critical questions have all been answered satisfactorily, there may be other factors impugning the success of the argument. This means there are other aspects that are important for the quality of the argument, but what those are and what the relation is between them and the questions is as yet uncertain. As for Hornikx’s remaining questions, I believe that thus far, adequate answers to them are still outstanding.
In the next chapter, I outline what I hypothesize the critical questions relate to and how we can use a model for argument classification to help us sort the critical questions in accordance with these aspects.
3. The Periodic Table of Arguments as an analytical tool
Now that we have seen what problems exist with Walton's approach to argument schemes and critical questions, we turn to the central question of this paper and how to solve it. As I have explained, the main question this research focuses on is how the critical questions relate to the argument scheme. It is inconclusive as to how Walton views this connection. I hypothesize that a way in which critical questions can be systematically connected to argument schemes, is to sort them in terms of propositional content and justificatory force. To do so, I utilize the Periodic Table of Arguments (PTA), a system of argument classification proposed and developed by Wagemans (2016).
In what follows, I shall first briefly discuss argument classification, as how I treat the critical questions later depends to some extent on what approach I take to argument classification. Next I turn to the concepts of propositional content and justificatory force and how these relate to critical questions. Finally, I explain how the Periodic Table of Arguments gives us the framework we need to be able to systematically categorize Walton's critical questions according to the theoretical concepts of propositional content and justificatory force.
3.1 Propositional content and justificatory force
Perhaps even more so than critical questions, much has been written on the classification of argument schemes, specifically on how many there are or should be, and how they may be distinguished from one another. Though that is not the focus of the current research, this issue is undeniably connected to the problems surrounding critical questions, which is why I will say a few words about argument schemes here.
Blair (1999, p. 18) comments that Walton’s list of schemes contains a much smaller number of what he calls subsets (which I would call supersets, given that what Blair refers to is a more general argument scheme rather than a more specific one). He points out that these were distinguished by means of a single general warranting principle. As we have seen in the previous chapter, this warranting principle, or argumentative mechanism, is what many theorists believe should be the means of categorizing arguments. What we generally mean when talking about a type or kind of argument is the way in which the acceptability of the argument is transferred to the standpoint, which is expressed by such a principle.
The method of evaluating arguments that instantiate these ‘types’ of arguments thus often include an assessment of the argumentative mechanism, or warranting principle: that part of the argument scheme that connects the argument to the standpoint, and in doing so, transfers the acceptability of the former to the latter. Though not all researches consistently follow up on this view in their analyses, it is nonetheless widely held that this ‘license’, ‘warrant’, or ‘argumentative mechanism’ is crucial when it comes to judging the ‘goodness’ of arguments. The research that is to follow is done with this theoretical choice in mind.
Given this manner of classification, I hypothesize that critical questions are connected to argument schemes in virtue of their relation to the propositional content and the justificatory force of the argument. These two concepts are generally accepted in the literature as delineating the features of an argument that are subject to scrutiny. Propositional content is that information
that is conveyed by the proposition which functions as the argument. Justificatory force has to do with the capability or capacity an argument has for transferring the acceptability of itself to the conclusion. At least at the level of the argument scheme, these concepts are generally held to be relevant to the evaluation of an argument.
I contend that it will therefore be useful to see if it is not possible to sort Walton's critical questions into these two categories. In that way, we will get a clearer view of what aspect of the argument the critical questions call into question. To help us with this sorting task, I propose to use the Periodic Table of Arguments. I believe this model can help us to systematically specify which part of the argument (scheme) relates to propositional content, and which to justificatory force, for any given pair of statements expressing a premise and a conclusion. This is because the PTA utilizes one premise, as opposed to Walton's schemes, which often have multiple premises. The propositional content is thus immediately represented in the PTA. The justificatory force can be derived. The upshot of the PTA in that sense is that it can determine for any arbitrary combination of premises what the structure of the argument is. Having these aspects laid out can in turn help us identify how the critical questions connect to these two aspects of a scheme.
3.2 The Periodic Table of Arguments
How can we determine propositional content and justificatory force in the PTA? To explain this, I will first briefly outline the workings of this model.
The Periodic Table of Arguments is a system of classification developed specifically with a descriptive aim in mind. To this end, it relies on three distinctions that are used to separate out statements on the basis of formal-linguistic properties and propositional content. The three parameters used are 1) the distinction between subject and predicate arguments, 2) the distinction between first-order and second-order arguments, and 3) the combination of types of propositions. I will not go into the third parameter here, as it is not relevant to our purposes. Note that the argumentative mechanism is not one of the distinguishing features used to characterize arguments (though it is implicitly present in the Table).11
The first parameter is the distinction between subject and predicate arguments, which is based on the linguistic distinction that can be made between the two elements of a categorical proposition: the subject, or that about which something is said, and the predicate, or that which is said about the subject. The PTA utilizes one premise (argument) and a conclusion (standpoint). The combination of a proposition containing an argument and a proposition containing a standpoint may either have a subject that is the same in both propositions, or a subject that differs. The same holds for the predicates of both propositions. Four combinations of arguments and standpoints can therefore be made: those with 1) same subject-same predicate, 2) same different predicate, 3) different same predicate, and 4) different subject-different predicate.
Wagemans (2019, p. 60) argues it is important that there is a ‘common term’, which is shared by both statements. This means that either the subject of both statements should be the same, or the predicate of both should be. This common term is what allows us to establish a link
between the premise and the conclusion. In virtue of having a common term, the argument can transfer acceptability of itself to the conclusion it is meant to support. In some cases, this common term is immediately visible. In other cases, it needs to be brought out by making the commitment of the speaker explicit. This is done by means of the second parameter.
The second parameter is a distinction between first-order and second-order arguments. A first-order argument is an argument at the level of the proposition: ‘a is X, because b is X.’ We see here that the common term shared by both statements is ‘X’, the predicate of both subjects. A second-order argument, on the other hand, is described as a case where the subject or predicate of the conclusion does not relate directly to the subject or predicate of the premise, as is the case with first-order arguments.
In the case of second-order arguments, the argument is related to the acceptability of the standpoint as a whole, rather than to one of its constituent elements. What this means is that the conclusion should not be analysed at the level of the proposition, but rather at the level of an assertion: the claim the speaker makes that what they state is true.
An example of this is the argument from authority, where the acceptability of the conclusion is defended by means of arguing some authority has stated the conclusion to be true. In that case, the predicate ‘is true’ is predicated of the standpoint as a whole. The standpoint, consisting of a subject and a predicate itself, then becomes the subject of the new proposition, with it ‘being true’ predicated of it.
For ‘same subject-same predicate’ arguments, “[t]he analysis then reveals that the argument draws on yet another relationship between the subject terms, namely the identity of [the subject of the premise] and [the subject of the conclusion].”12 (Wagemans, 2019, p. 63) In the case of arguments where the premise and conclusion differ entirely, both propositions can be analysed as assertions and the argument depends on another kind of relation between the subjects of both statements.13
A second important job of the common term is that it “helps the analyst in formulating the underlying mechanism of the argument.” (Wagemans, 2019, p. 63) Already in (2011a), Wagemans has shown how the combinations of premises and conclusions yield the argumentative mechanism. This mechanism is an expression of the relation between the elements of the premise and conclusion that differ from one another. Where the common term is what allows for a transfer of acceptability from premise to conclusion, this relation is where the transfer takes place. Wagemans (2019) sketches what these mechanisms look like. Below is an overview of the systematic types of arguments (characterized only by the first two parameters) with the respective mechanisms and concrete examples of each.
12 An example of an argument of this type is a logical tautology: “A, because A.”
13 An example of an argument where the content of the premise and the conclusion differ entirely could be what is called a disjunctive argument: “Megan must have been at school, because her sister wasn’t home.” The reasoning behind such an argument may be something like, “Normally her sister looks after Megan when she (Megan) is at home (meaning the sister is at home too), but if her sister is not home, then Megan must be elsewhere (for instance, at school).”
Name Abstract representation
Concrete example
Abstract lever Concrete lever
First-order predicate argument
a is X, because a is Y
The suspect (a) was driving fast (X), because he (a) left a long trace of rubber on the road (Y)
relationship Y-X leaving a long trace of rubber on the road (Y) is an effect of driving fast (X) First-order subject argument a is X, because b is X Cycling on the grass (a) is forbidden (X), because walking on the grass (b) is forbidden (X) relationship b-a walking on the grass (b) is analogous to cycling on the grass (a) Second-order
subject argument
q is T, because r is T
He (a) must have gone to the pub (X) (q) is true (T), because the interview (b) is cancelled (Y) (r) is true (T) relationship r-q either the interview is not cancelled (¬r) or he went to the pub (q) Second-order predicate argument q is T, because q is Z
We (a) only use 10% of our brain (X) (q) is true (T), because that (q) was said by Einstein (Z) relationship Z-T being said by Einstein (Z) is authoritative of being true (T)
Table 1. Overview of argumentative types and their corresponding argumentative mechanisms (the ‘lever’) (Wagemans, 2019, p. 65).
How then can we see the concepts of propositional content and justificatory force in the Table? Firstly, the propositional content simply refers to the information contained in the argument. As we have seen, the argument is clearly one of the features in the PTA. For each systematic type, a combination of the conclusion and the premise functioning as the argument is listed. In that way, finding the propositional content is not difficult. An example of this is, “Cycling on the grass is forbidden, because walking on the grass is forbidden.” In this case, the conclusion is “Cycling on the grass is forbidden.” The argument used to support this conclusion is “Because walking on the grass is forbidden.” The propositional content of the argument is just the information that walking on the grass is forbidden.
As for justificatory force, this is expressed in the argumentative mechanisms which are expressed by the relation between the argument and the conclusion. The subject arguments
have argumentative mechanisms that are based on referent transfer (a relation between the subjects of the premise and conclusion), whereas the working of predicate arguments rely on argumentative mechanisms that utilize predicate transfer (a relation between the predicates of both statements).
In our example, there is a common term, namely the predicate of both the premise and the conclusion: being forbidden. This means there must be some sort of relation between the subjects of the two statements, which licenses the attribution of ‘being forbidden’ to both. This relation between the subjects is expressed in the argumentative mechanism, “Walking on the grass is analogous to cycling on the grass.” This is thus an example of referent transfer.
In the next chapter, I will analyse three of Walton's argument schemes and critical questions in terms of this framework. We will then see to what extent the critical questions Walton has formulated can be traced to these two concepts of propositional content and justificatory force.
4. Analysing critical questions
In this section, I will analyse the three examples of argument schemes: argument from expert opinion, argument from pragmatic inconsistency, and argument from sacrifice. The question I am attempting to answer is what the relation is between the premises included in Walton's argument schemes and the critical questions he formulates for them. At first glance, it may appear strange that there is no one-to-one correspondence between these two. By this I mean that pre-theoretically, one may reasonably wonder why there is not one critical question for each premise, that directly questions what is stated in that premise.
To find out what this connection between premise and critical question might be, I first reconstruct each scheme in terms of the Periodic Table of Arguments. To do so, I review each premise in Walton's schemes. Often, there is more than one premise included in the argument scheme, and for each I try to determine what the (justificatory) function of it is. I describe the relation between premises and critical questions in terms of the framework of propositional content and justificatory force, as I hypothesize that it is appropriate for the questions relate to either of these aspects when it comes to the evaluation of the argument scheme.
I then reconstruct each scheme using the pragma-dialectical notation system. This system uses numbers to express what ‘level’ the statement is on, allowing us to see which premise supports which. The standpoint is indicated by 1, the argument directly supporting the standpoint by 1.1, and the argumentative mechanism by 1.1’. Further supporting arguments are represented by adding further numbers (1.1.1 thus indicates an argument supporting argument 1.1; an argument supporting the argumentative mechanism 1.1’ is indicated by 1.1’.1). Coordinative arguments are indicated by adding letters to the notation (e.g., 1.1a and 1.1b).
For some schemes, in order to properly find the right place for each premise in the reconstruction, it is needed to reformulate premises or add new ones which express previously unexpressed elements. Transformations of this type are common in reconstruction of argumentative discourse and also perfectly acceptable as long as the analyst’s choices are accounted for (see, e.g. van Rees, 1996). In this way, I find the systematic type of argument for each scheme, according to the PTA.
By doing this, it becomes easier to see which premise directly supports the conclusion and what the argumentative mechanism at work is. Once these two aspects are clear, we can more easily see what the critical questions relate to. If they relate to the argument, they in some way probe the propositional content of the argument. If they relate to the argumentative mechanism, they question the justificatory force the argument has.
In what follows, I see whether it is possible to label all of the critical questions in this way, using the PTA as a sorting mechanism.
4.1.1 Argument from expert opinion version 1
A first problem to be noted with argument from expert opinion is that Walton, Reed and Macagno formulate four different versions of this scheme (2008). The first version is a more standard one, consisting of two premises and a conclusion, thus instantiating something of an ‘informal syllogism’. Other versions include modified or extra premises. For simplicity’s sake, but also
since the PTA works with only one premise, I will here analyse the simplest version, which is also the one listed in the compendium (Walton, Reed and Macagno, 2008, p. 310).
Argument from expert opinion
Major premise: Source E is an expert in subject domain S containing proposition A.
Minor premise: E asserts that proposition A (in domain S) is true (false). Conclusion: A is true (false).
The first step is to reconstruct this scheme in terms of the PTA. The PTA utilizes a combination of only one premise and a conclusion. Traditionally, the minor premise is taken to be the argument in defence of the case in point. Further, when someone asserts, “A is true”, and we were to ask, “Why is A true?” the answer surely is “Because E says it is true” and not “Because
E is an expert in subject domain S containing proposition A.” This indicates it is what the authors
here call the minor premise which is the actual support for the conclusion. Therefore, in reconstructing the scheme in terms of the PTA, it seems correct to work with only the ‘minor premise’. The scheme then becomes as follows:
STP: A is true (false).14
ARG: E asserts that proposition A (in domain S) is true (false).
As Wagemans has shown (2016; 2019), these kinds of arguments can be reconstructed as second order predicate arguments. They can schematically be represented as follows:
STP: A (q) is true (T).
ARG: A (q) is asserted by E (Z).
We then have the argument form ‘q is T, because q is Z’.
Wagemans also describes the type of argumentative mechanism by means of which the argument ‘works’ (2019, p. 65). In the case of a predicate argument, there is some sort of relationship between the two predicates of the conclusion and the premise. It is this relation which transfers the acceptability of the latter to the former. In this case, there is a relationship between T (being true) and Z (being said by E). Wagemans expresses this as “being said by [E] is authoritative of being true” (2019, p. 65). This relation is indirectly expressed by the ‘major’ premise in the original scheme. We will see below how it works to support the argumentative mechanism.
According to our framework, critical questions probe either the propositional content or the justificatory force of the argument. The propositional content is what is stated in the argument, here “A is asserted by E”. Questions asking after the information stated in the argument then should relate to propositional content. The justificatory force relates to the
14 In this case, the conclusion is already expressed as an assertion, rather than as the simple proposition ‘A’. It therefore isn’t necessary to add the predicate ‘is true’, but it is to be considered a second order argument, since the conclusion is expressed as an assertion.
mechanism, here “being said by E is authoritative of being true”. Questions asking after this relation then should relate to justificatory force.
The critical questions accompanying this argument scheme according to Walton, Reed and Macagno (2008, p. 310) are the following:
CQ1: Expertise question: How credible is E as an expert source? CQ2: Field question: Is E an expert in the field that A is in?15 CQ3: Opinion question: What did E assert that implies A?
CQ4: Trustworthiness question: Is E personally reliable as a source?
CQ5: Consistency question: Is A consistent with what other experts assert? CQ6: Backup evidence question: Is E’s assertion based on evidence?
Of all of these, CQ3 is the only question that seems to ask after the facts of the argument – the premise stating A was asserted by E. CQ3 is the only question that asks whether this indeed was the case.16
CQ1 directly probes the linking mechanism, in that it questions whether E’s word is indeed ‘a good reason’ to accept the conclusion. This question calls for further justification of the argumentative mechanism, thus leading to subordinative argumentation (see 1.1’.1 below).
CQ2 again directly questions the argumentative mechanism, since it questions whether
E’s say-so is an acceptable reason to accept the conclusion. This too calls for further argument
defending the argumentative mechanism, thus rendering subordinative argumentation (1.1’.2). This premise is similar to the ‘major’ premise in the original argument scheme.
CQ3, as mentioned above, directly questions the propositional content of the argument. Essentially, what someone asking this question would want to know is whether it is indeed true that E asserted A. An answer to this question is a defence of the propositional content of the argument, meaning subordinative argumentation (1.1.1).
CQ4 directly relates to the argumentative mechanism, since again what it questions is whether E’s saying so is good enough to accept the conclusion. It can be interpreted as asking whether the expert has spoken solely from their position as expert, and not from other interests. This too leads to subordinative argumentation in support of the argumentative mechanism (1.1’.3).
CQ5 also relates to the argumentative mechanism, but in a different way than CQ1, 2, and 4. In this case, the question is whether E’s assertion is sufficient to justify accepting the conclusion. If it is, it wouldn’t matter what others have had to say on the subject, but if the
15 One could say this question relates to the major premise, if you take ‘field’ to be identical to ‘subject domain’. The latter is suggested by Blair (2012, p. 130).
16 On Wagemans’s (2011b) analysis, CQ 1 and 2 relate to the propositional content of the argument. However, in that version of the scheme, the fact that E is an expert is included in the premise that functions as an argument. In that case, the questions indeed ask after the ‘facts’ of the premise. A special consideration here is that CQ1 is interpreted as “is E really an expert?” In that sense, the question probes the propositional content of the argument. On my interpretation, firstly, the fact that E is an expert is not part of the premise. Secondly, the question as is seems to emphasize the credibility of the expert. Clearly, the question is ambiguously phrased. However, if it is interpreted in the latter sense, the credibility of E contributes to the argument “being a good reason” to accept the conclusion. This is why I believe the question relates to the argumentative mechanism, rather than the propositional content of the argument.
question is raised, then E’s opinion alone clearly is not enough. I have reconstructed a possible answer to the question as support for the justificatory force (1.1’.4).17
CQ6 pertains to the argumentative mechanism in a similar manner to CQ1, 2, and 4, which ask whether it is relevant that it is E who states the conclusion. This question questions E as a source, and asks for a further substantiation of E’s knowledge, again leading to subordinative argumentation (1.1’.5).
To show how the critical questions can lead to additional arguments, I have added a number of premises which may function as answers to the questions to the reconstruction below.18
1 A (q) is true (T).
1.1 A (q) is asserted by E (Z). [1.1.1 E asserted A.]
[1.1’ Being asserted by E is authoritative of being true.]
[1.1’.1 E is generally acknowledged to be a credible expert.]
1.1’.2 Source E is an expert in subject domain S [= field] containing proposition A. [1.1’.3 E is personally reliable as a source.]
[1.1’.4 A is consistent with what other experts assert.] [1.1’.5 E’s assertion is based on evidence e.]
4.1.2 Argument from expert opinion version 2
As mentioned above, Walton, Reed and Macagno (2008) list a number of variations of the argument from expert opinion. We have seen in the previous analysis that critical questions can inspire additional premises. The version of the scheme listed below incorporates this idea (Walton, Reed and Macagno, 2008, p. 19). I discuss it for this reason, and also to show that it is not essentially different from the version discussed above.
Appeal to Expert Opinion19
Major premise: Source E is an expert in subject domain S containing proposition A.
Minor premise: E asserts that proposition A (in domain S) is true (false).
17 There is another way of reconstructing an answer to this question, where the new premise becomes one that is coordinative with the original argument. Taken together, these arguments then would be sufficient to support the conclusion. However, this leads to problems with the reconstruction when it comes to formulating the
argumentative mechanism. I am aware of this issue, but this is not the place to solve it. 18 Premises I have formulated are distinguished by square brackets: […].
19 Note that the name of this version of the argument differs to the first version (‘appeal to’ versus ‘argument from’). This is their differentiation, not mine.
