RMA Thesis - June, 2018
University of Amsterdam Graduate School of Humanities
Communication and Information Studies
Research MA - Rhetoric, Argumentation Theory and Philosophy
The Role of Analogy in Imaginary Scenarios
Analogical Reasoning and Argumentation in Scenario-based Thought
Table of Contents
1 Introduction
2 Thought experiments and imaginary scenarios
2.1 The nature of thought experiments and the use of imaginary scenarios
2.2 The epistemic and argumentative status of scenario-based thought experiments 2.3 Reasoning and argumentation
3 Reasoning analogically in scenario-based thought experiments
3.1 Reasoning schemes: the missing premise problem
3.2 Reasoning quasi-dialectically in destructive thought experiments
4 The role of analogies in scenario-based argumentation
4.1 Argument schemes: the missing premise problem in arguments from parallel reasoning
4.2 The role of structural similarities between imaginary and real-world settings
5 Conclusion
References
Thought experiments and imaginary scenarios are widely used in a variety of knowledge-related and action-related domains. In the academic field, for example, this subject matter has been of interest especially for scientists and philosophers, who have resorted to scenario-based thought experiments as illustrations or testing grounds for their hypotheses for centuries. Furthermore, thought experiments are extensively employed also within the business, military and political domain, where they are used to aid decision making regarding strategies and courses of action to adopt (Aligica, 2005, p. 816). Problem solving and decision making, in these cases, is guided by imagining the potential consequences that certain strategies may entail, by transposing them onto fictitious settings.
Thought experiments and imaginary scenarios are very tightly linked to one another, hence their individual characteristics are often difficult to identify independently from each other. Generally, thought experiments can be understood as special types of experimental processes that are not implemented in physical reality, as they are performed within the mind and involve mental manipulations of the elements involved (Brown, 2011, p. 1). Their object is the external world - or theories regarding the world - and they proceed through the method of pure reasoning (pp. 15-16). Imaginary scenarios, on the other hand, are best defined as invented or fictional situations, often used in thought experiments, that exploit conceptual possibilities (Ward, 1995, pp. 182-185).
In the present study, I will focus on thought experiments that make use of fictitious narratives in the academic domain. In such cases, a target theory is transposed onto an imaginary setting and, through the process of thought experimentation, its implications and consequences are reflected upon. This process appears to reflect that of an analogical mapping, taking place between two domains: the imaginary scenario in which the theoretical claim is transposed and implemented, and the real life situation in which the original hypothesis is meant to be applied. Hence, the objects and conclusions belonging to one domain are considered as systematically corresponding to objects and conclusions in another domain. The notion of analogy, therefore, emerges as a key concept in the performance of scenario-based thought experiments analysed in this study.
Within their various domains of application, scenario-based thought experiments can be employed with different purposes. As a consequence, a debate has developed with regards to their epistemic value. The current dispute revolving around these linguistic and intellectual phenomena focuses on the cognitive weight that should be ascribed to them, and to whether - and to what extent - they can be regarded as lending actual support to a claim under discussion (Angelucci, 2014, p. 204). Such debate has driven argumentation theorists to focus on scenario-based thought experiments, as from an argumentative perspective imaginary narratives are, sometimes, considered useful to test and - in some cases - to convince an
audience about the (un)acceptability of a hypotheses under discussion, by transposing it onto an imaginary situation and considering its consequences and implications (Ward, 1995, p. 182; Gendler, 2004, p. 1152; Popa, 2017, p. 9). However, different conclusion are drawn by different scholars as to their epistemological force.
One common attitude is to consider scenario-based thought experiments as heuristic tools that elucidate and spell out an academic claim, a policy or a strategy under discussion in terms that are more easily understandable thanks to their vividness. From this perspective, imaginary scenarios are seen as (im)possible worlds in which one can illustrate a claim or simulate a strategy for purposes of clarification (Aligica, 2005, pp. 818-819). Also, they may be used heuristically to aid the reasoning process that leads to the actual formulation of a theoretical hypothesis (Brown, 2011, p. 35). Alternatively, others have been investigating the argumentative features of thought experiments, starting from the ascertainment that, although not necessarily linked to any empirical data and often fabricating scenarios that are impossible to enact, they seem to have an epistemic weight as they produce new knowledge (Aligica, 2005; Häggqvist, 2009, p. 55). In the academic domain, for example, the argumentative character of thought experiments is derived by the fact that they can help a scholar to justify her acceptance or rejection of a theoretical claim under discussion. In these perspectives, they are said to sometimes “serve the scholars’ aim of resolving a difference of opinion regarding the acceptability of an academic claim” (Popa, 2016, p. 49), and “when employed as part of an academic dispute, an imaginary scenario is said to be used argumentatively” (p. 155).
In order to clarify this dispute regarding the epistemic value of scenario-based thought experiments, I believe it is important to underline the fundamental features that distinguish the process of reasoning from the activity of argumentation. The reason for this is that the heuristic or argumentative nature of this phenomenon depends on the purpose of employing thought experiments based on imaginary scenarios. As a consequence, a heuristic use will be guided by reasoning purposes of illustrating or designing a theoretical claim; an argumentative use will be based on justificatory purposes of rejecting or accepting a hypothesis under discussion. In both cases, scholars have identified certain recurring structural aspects in the performance and designing of thought experiments. Häggqvist (2009), for example, outlines a pattern of reasoning underlying certain types of thought experiments, and three argument schemes that correspond to its dialectical progression. Also Popa (2017) identifies common argumentative traits in scenario-based argumentation - in the form of discussion structures (p. 88) - and in the reasoning process that underlies its construction and performance.
However, it seems like none of the accounts analysed in the present research focus systematically on what kind of reasoning and on what kind of argumentation is at play in the
performance of scenario-based thought experiments. In my opinion, when reconstructing such experiments - as well as in their dialectical progression - and when reflecting upon their underlying structure, the centrality of relations of analogy between the elements of these phenomena emerges. What guides this ascertainment is that what results from scenario-based thought experiments is that a main - target - claim is projected onto a parallel claim and onto an imaginary setting in which its implications are considered. At the reasoning level, the experimenter applies the conclusions drawn in the fictitious narrative onto the target claim, implemented in the real world; in the argumentative dimension, the listener is asked to do the same, and judge the (un)acceptability of the claim at issue. The perspectives considered in this study do not deny the role of analogies in scenario-based thought experiments, and some have also explicitly resorted to thought experiments as examples of arguments by analogy (Juthe, 2005, p. 11; Walton, 2014, p. 27). However, the crucial function of analogy in scenario-based thought experiments seems to often be presupposed, or not investigated in enough detail.
In the present study, my overarching goal is to focus on the fundamental role of analogies in scenario-based thought experiments through the analysis and evaluation of current approaches to the subject matter. Therefore, I will analyse how analogy is embodied in theories of reasoning and of argumentation in relation to thought experiments. In respect to such theories, I will analyse how analogy is expressed in the reasoning schemes and in the argumentative reconstructions suggested in the literature. Moreover, I will evaluate how adequately the role of analogy is accounted for in the reasoning and the argumentative patterns that are associated with thought experimental processes.The reason for considering both dimensions, of argumentation and reasoning, is that the debate regarding their epistemological status is mirrored in the different ways they can be employed: heuristically or with justificatory purposes.
In order to achieve my research goal, in Section 2 I will outline the nature of thought experiments and imaginary scenarios, and provide an overview of the current debate surrounding their epistemic status. Moreover, I will show how this debate is connected to the different uses that can be made of scenario-based thought experiments, on the basis of the distinction between reasoning and argumentation. In Section 3 I will investigate the role of analogies at the reasoning stage of designing a thought experiment, and offer a critical assessment of the reasoning schemes identified in the literature. In Section 4 I will explore the role of analogy in scenario-based argumentation, and evaluate the suitability of two forms of schemes for arguments based on relations of analogy or comparability. Finally, in the concluding section, I will provide an overall evaluation of how the notion of analogy is embodied in theories of reasoning and of argumentation that deal with scenario-based thought experiments. In this Section, I will also offer some suggestions for future research in the field.
In the present Section I will clarify the nature of thought experiments and of imaginary scenarios. Through addressing some of the main aspects that characterize the debate surrounding this subject matter, I will focus on distinguishing between these two intellectual tools. Secondly, I will address the matter of the epistemic value of thought experiments, hence distinguishing between their heuristic and argumentative use. Finally, on the basis of this distinction, I will focus on the difference between the process of reasoning and the process of argumentation, useful to better understand the uses that are made of scenario-based thought experiments.
2.1 The nature of thought experiments and the use of imaginary scenarios
Providing a clear-cut definition of a thought experiment is not an easy task, as “there is no consensus among scholars about their nature” (Picha, 2011, p. 154). As a consequence, many scholars interested in the debate around thought experiments have underlined the peculiar characteristic of this method of investigation that, although systematically employed in many academic and non-academic fields, appears very hard to spell out precisely (Aligica, 2005; Brown, 2011; Häggqvist 2009; Picha, 2011, p. 154; Souder, 2003). However vague the term has been said to be, definitions have been suggested that mainly focus on the characteristics that are commonly identifiable in what we label as thought experiments. One frequent tendency has been to distinguish the class of thought experiments from that of regular experiments, yet still definiding the legitimacy of the former method in spite of its fundamental difference with the latter: one occurs in the mind, the other occurs within the boundaries of empirical reality. A popular metaphor used to refer to thought experiments is, for example, that of experiments performed in the laboratory of the mind (Brown, 2011, p. 1; Souder, 2003, p. 205) which aim at reaching their goal without the need of being empirically executed (Sorensen, 1992, cited in Souder, 2003, p. 205). Brown (2011, p. 1) elaborates on this metaphorical definition by characterizing thought experiments further as actual experimental processes that, however, are often impossible to implement due to the lack of necessary equipment - such as technological apparatuses - or to the fact that they hypothesize situational contexts that appear impossible in principle - such as a universe devoid of matter. However, because of their performance within the mind, they are subject to mental manipulations and forms of “imaginistic reasoning” (Gendler, 2004, p. 1155) that make the whole process observable - or somewhat perceivable - to the thinker, implementing the experimental process in a whole new way.
In my opinion, an interesting and effective way to define thought experiments is through the goal they seek to achieve, and through the object they focus on. It is true that
thought experiments are employed in a variety of different fields, such as the academic domains of physics, biology and philosophy (Brown, 2011, p. 31), as well as in the dimension of decision-making and strategy-oriented activities such as management, public administration, business and military (Aligica, 2005, pp. 815-816). Nevertheless, they always seem to be aimed at confirming or disconfirming a certain hypothesis or theory (Gendler, 2004, p. 1154; Häggqvist, 2009, p. 58), and their object is always an aspect of the external world - or implications of a certain theory regarding the external world (Brown, 2011, pp. 15-16). As Brown points out, thought experiments are not experiments about thought, as these consider thought as their primary object. An example can be to think about the sentence “colourless green ideas sleep furiously” (p. 15) and conclude whether it is a grammatically correct proposition. Thought experiments, instead, are experiments regarding the world that take thought to be the method through which we learn something about it (p. 16).
Although they share these common, general features, different types of thought experiments have been identified, according to their more specific aim. One first and main distinction is that between ‘constructive’ and ‘destructive’ thought experiments (Brown, 2011, pp. 32-35). Constructive thought experiments are those aimed at achieving a positive result. This goal may lead to the formulation of a claim in order to explain and account for what is happening in the scenario depicted in the thought experiment - therefore, to the performance of a ‘conjectural’ thought experiment (pp. 38-39); it may lead to draw a conclusion from a specific and articulated theory by illustrating aspects of it that may seem contradictory - thus, to the production of a ‘meditative’ thought experiment (p. 35); or, it may lead to considering an unproblematic event that leads at some new conclusion, not accounted for in an already given theory - resulting in a ‘direct’ thought experiment (pp. 39-40). On the other hand, destructive thought experiments are arguments directed against a given theory, that aim to a reductio ad absurdum of a claim under discussion (p. 33); to demonstrate its logical invalidity (p. 49); or, to deny the phenomena depicted in the thought experiment (‘counter thought experiments’; p. 50). A famous example of destructive thought experiment is the Chinese room argument by John Searle (1980), a thought experiment aimed at rejecting the theories of strong A.I that take computer programmes to be intelligent and able to exhibit understanding.
Suppose that I'm locked in a room and given a large batch of Chinese writing. Suppose furthermore (as is indeed the case) that I know no Chinese, either written or spoken [...] Now suppose further that after this first batch of Chinese writing I am given a second batch of Chinese script together with a set of rules for correlating the second batch with the first batch. The rules are in English, and I understand these rules [...]. Now suppose also that I am given a third batch of Chinese symbols
together with some instructions, again in English, that enable me to correlate elements of this third batch with the first two batches, and these rules instruct me how to give back certain Chinese symbols with certain sorts of shapes in response to certain sorts of shapes given me in the third batch. [...] As far as the Chinese is concerned, I simply behave like a computer; I perform computational operations on formally specified elements. [...] it seems to me quite obvious in the example that I do not understand a word of the Chinese stories (Searle, 1980, pp. 417-418).
An example of what I believe to be a constructive thought experiment, instead, is the one of Thomson’s violinist, formulated to support the moral permissibility of pregnancy terminations.
[...] let me ask you to imagine this. You wake up in the morning and find yourself back to back in bed with an unconscious violinist. A famous unconscious violinist. He has been found to have a fatal kidney ailment, and the Society of Music Lovers has canvassed all the available medical records and found that you alone have the right blood type to help. They have therefore kidnapped you, and last night the violinist's circulatory system was plugged into yours, so that your kidneys can be used to extract poisons from his blood as well as your own. [...] To unplug you would be to kill him. But never mind, it's only for nine months. By then he will have recovered from his ailment, and can safely be unplugged from you." Is it morally incumbent on you to accede to this situation? (Thomson, 1971, p. 48).
Through this experiment, Thomson seems to be rephrasing her claim through an illustration that is meant to guide our intuitions more directly. In this case, accepting the implications in the scenario is not as compelling as in the previous example, because they do not necessarily lead to absurd conclusions. One could, potentially, reject the fact that unplugging the violinist is morally acceptable. Also, this thought experiment is used to formulate a claim, not to reject it. According to Brown’s aforementioned classification, thought experiments that hold justificatory force are those aimed at the rejection of a target claim.
The examples of thought experiments presented above both employ imaginary scenarios as testing grounds for the claims under consideration. Imaginary scenarios represent new and made up situations that help us to interpret the concepts under discussion by evoking intuitions about them (Ward, 1995, p. 181). They can be seen as fictitious settings in which we apply the concepts under discussion in order to guide our intuitions in a more direct manner (p. 182). To say this with Popa (2016, p. 9), “an imaginary scenario is a more or less fictional story about fictional entities and events. The story is employed as a testing ground for the claim under discussion”. In the case of the Chinese room, the imaginary scenario depicted is that of a man (perhaps Searle himself) situated in a solitary room, receiving
batches of Chinese and English scripts that he has to manipulate on the basis of given instructions. In the violinist thought experiment, the scenario is represented by the fictitious situation of having a violinist’s body attached to one’s own, that the bodies are co-dependent from one another, and so on. In destructive thought experiments (Brown, 2011, pp. 33-35), the role of the fictitious setting is that of “showing that one of the consequences that the rival theory would be committed to if the state of affairs described in the scenario were to actually occur falls short of matching our current intuitions” (Angelucci, 2014, p.210). Therefore, the relation between performing a thought experiment and employing an imaginary scenario during this process is very tight, and it is surely a hard task to define one without somehow referring to the other. In fact, it is frequent to find definitions of thought experiments as processes through which one contemplates an imaginary scenario (Gendler, 2004, p. 1152), through which we test our intuitions in unfamiliar scenarios (Ward, 1995, p. 183), or through which we create fictitious scenarios (Angelucci, 2014, p. 210). In some cases the difficulty of separating these two concepts is such that we can find definitions that seem to overlap them. An example can be found in Aligica (2005) that appears to define scenarios in terms of thought experiments: as a form of reasoning from suppositions to extract a conclusion in the domain of public administration or strategy planning. Although he seems to differentiate the two by stating that a scenario “is a special instance of thought experiments”, Aligica defines it as
an attempt to draw an instruction from a process of hypothetical reasoning that proceeds by drawing out the consequences of an hypothesis which, although may be anchored in well established facts, refers to future (that is possible) developments. Thus it consists in reasoning from a supposition whose truth status is uncertain but is assumed provisionally as certain in the interests of developing a broader image of its implications, consequences and assumptions (2005, p. 816).
However, these two concepts are not to be confused because not all thought experiments involve forms of “imaginistic reasoning” based on fictitious scenarios (Gendler, 2004, p. 1155). Thought experiments, therefore, represent the method of reasoning about a certain hypothesis that does not involve empirical experimentation, whilst imaginary scenarios are narratives that depict a fictitious situation in which the implications of the hypothesis are put to the test.
Because of the fact that scenario based thought experiments can, usually, only be performed within the mind and do not involve empirical data, much debate has developed around the issue of whether they can be said to carry epistemic value. The question is, if scenarios instantiate situations that are not likely to be realized and they do not rest on any empirical findings, how can they involve new knowledge about things that concern the outside world? In other words, the debate revolving around the epistemic status of thought experiments addresses whether it is possible “to accept an imaginary scenario as a source of justification, that is to say, to accept it as a good reason to embrace an opinion” (Picha, 2011, p. 155). Indeed, despite the defining feature of these forms of mental experimentation - namely, that of escaping empirical grounding - they represent a successful and frequently used method of investigation in many academic and non-academic domains, however impossible to implement they may be. Nevertheless, many question why we proceed to draw conclusions on the basis of speculation regarding unreal scenarios; how can considering a fictional situation produce new knowledge; how can speculation about fiction provide a legitimate and reasonable justification for a claim.
According to Aligica (2005) scenarios are not that separate from reality after all. He believes that such mental models are, in fact, grounded in experience because “they are abstracted and extrapolated from experience and knowledge of the world” and are manipulated and transformed “in a manner that is itself realistic, since it is guided by the experience and knowledge of how the world works”. It is important to keep in mind that Aligica (2005) is not dealing with philosophical or scientific scenario-based reasoning but, rather, with the way scenarios are employed in decision-making regarding public, military or business policy making. From this perspective, scenarios are simulations run on mental models of actual real-life situations and provide further understanding of the conditions under which the model would hold (p. 821) - representing, therefore, “possible possibilities” (Ward, 1995, p. 186). Academic thought experiments, instead, often involve more “conceptual possibilities” (p. 185), that are unlikely to occur in physical reality - such as men with two heads, brains in vats or violinists attached to women’s bodies. Ward believes that it is not enough for scenarios to represent “logical possibilities” in order to make sense to us - i.e. to not present self-contradictory elements - because being logically valid does not imply that they may be physically instantiated (p. 185). What makes us take scenario-based reasoning seriously, according to Ward, is the presence of “black boxes” (p. 186), elements of the narrative that provide an additional context or specify how the conceptual possibilities envisioned could be realized - by the use of advanced technology not yet invented, for example. From this perspective, the imaginary scenario can actively influence our intuitions because it is considered as a potentially real possibility and in this way it makes sense for us.
The epistemic value of scenario-based thought experiments also strongly depends on the aim that guides its performance. As mentioned above, Brown’s distinction between constructive and destructive thought experiments (2011, pp. 32-40) allows for a first categorization. In the first case, the proponent of the experiment aims at achieving a positive result, namely at the formulation of a claim or at a clarification and illustration of problematic aspects of a claim. In both instances, a certain form of reasoning is used to formulate a hypothesis, hence operating at a heuristic level. On the other hand, in the case of destructive thought experiments, the process is used with justificatory purposes, in order to defend the acceptability or unacceptability of the hypothesis in question. However, this distinction is far from exhaustive, because also among those that defend the epistemological force of thought experiments there are divergent opinions as to how they should be categorized and to how they achieve new knowledge. Some claim that they can be considered as deductive arguments that lead to new knowledge, as “a new configuration of knowledge emerges out of the exercise [of thought experiments] in spite of the fact that no original empirical findings are involved” (Aligica, 2005, p. 820). Others (Häggqvist, 2009; Souder, 2003) focus on their counterfactual nature, tightly linked to arguments but not identifiable with them. Häggqvist (2009) underlines the strong causal connection between counterfactual use of scenarios and arguments on the basis that the former aim at formulating counter instances of a certain hypothesis, and the latter make it useful for assessing the validity of a claim. In other words, “performing a thought experiment causes thought experimenters - inventor or audience - to hold certain non-observational statements as true, which may subsequently be employed in arguments concerning the theory to be tested” (p. 62). Also Angelucci (2014) considers the use of thought experiments in rejecting a hypothesis, as many of them apply such a theory to an imaginary scenario and show that, if the theory under discussion were to be accepted, unacceptable implications would follow in the scenario (p. 210).
The type of scenario-based thought experiments that Häggqvist (2009, p. 63) is mainly interested in are the aforementioned destructive thought experiments, which systematically aim at providing a counter example to the claim at issue. This systematicity is identified also in terms of the structure that the experiment follows, namely that a counterfactual scenario is presented as (conceptually) possible (◇C); then, the hypothesis under consideration (T) is transposed onto the scenario (C), and connected to a certain consequence or event occurring within it (W); finally, it is shown that such a consequence does not emerge in the scenario, and the falsity of the claim at issue is derived. The scheme identified is represented as follows:
◇ C
C⬜➝ ¬W
________________ ¬ T
Fig. 1 Häggqvist’s reasoning scheme (2009, p. 63).
This scheme is meant to be the representation of the cognitive process of a counterfactual thought experiment, specifically of the intentions of the thought experimenter that performs it. This scheme can be causally related to the arguments used in defence of the claim regarding the unacceptability of (T). Häggqvist mentions also other three forms of argument schemes connected to the performance of a thought experiment (2009, pp. 65-67). These are schemes that represent possible defences of (T) and rejections of the plausibility of the scenario depicted (C), or of the consequences that should follow within it, and will be dealt with in Section 3.
Another important question that characterizes the debate around scenario-based thought experiments is whether they can be considered as reasonable forms of argumentation aimed at accepting or rejecting a claim. Popa (2016) deals with this aspect extensively from a pragma-dialectical perspective on argumentation, specifically focusing on academic debates. When thought experiments are used in trying to resolve a difference of opinion between scholars, one of the main procedural starting points is agreeing that both parties are allowed to justify or reject the claim at issue based on what occurs in a imaginary scenario (p. 53). This agreement is crucial, because without it the thought experimental process would carry no epistemic value and no justificatory force for a claim. The propositional content of “scenario-based argumentation” (p. 71), therefore, describes what would happen in an imaginary scenario if the theory under discussion were to be applied in it. In accordance with Häggqvist’s (2009, p. 63) structural reconstruction of certain types of thought experiments - namely, destructive ones - also Popa (2016, p. 71) identifies a systematic structural feature of such scenario-based thought experiments, namely that they present the form of modus tollens. In fact, the falsity of the theory at issue is deduced by assuming the theory to be true, but then showing that false or absurd consequences would follow from it - of course, within the scenario. However, Popa underlies that this type of reconstruction does not take into account the dialectical progression of the thought experimental process used argumentatively in academic disputes, as it only represents the contribution of one party, not accounting for the communicative features that are relevant for their argumentative status. For this reason, Popa introduces a reconstruction of scenario-based argumentation using the tool of “discussion structures” (p. 79), in order to represent clearly every move and contribution made in the discussion. For the purpose of the present study I believe that it is enough to report Popa’s reconstruction of a simple form of scenario-based argumentation, first in the form of modus
tollens (a) and then reconstructed through the use of discussion structures (b). The example is that of Searle’s Chinese Room argument mentioned above.
(a) 1. ¬ p 1.1 ¬ (I ➝ E)
1.1’ p ➝ (I ➝ E)
Fig. 2 Popa’s reasoning scheme (2017, p. 83).
This general structure, that resembles Häggqvist’s reconstruction (2009, p. 63), shows that an academic claim (p) is false, because in the agreed upon imaginary scenario (I), a certain event (E) does not occur, although it should occur if (p) were true. By applying this structure to Searle’s argument, we would obtain the following reconstruction:
(b) 1. Digital computers do not exhibit understanding
1.1 It is not true that in the scenario, I understand Chinese
1.1’ If digital computers could exhibit understanding, then, in the scenario, I would understand Chinese
Fig. 3 Popa’s reasoning scheme applied to Searle’s Chinese room (p. 85).
By applying the discussion structures, Popa shows how the dialectical progression of the scenario-based academic dispute proceeds between Searle (S) and Dennett (D):
Discussion move per stage
Confrontation
1 S: -/p 2 D: ?/(-/p)
STP DT
Digital computers cannot understand How do you know?
Opening
3 S: !/s(p⇢(I⇢E))
4 D: +/s((p⇢(I⇢E))
PRO
ACC
Do you agree that if they could understand then in the scenario Searle-in-the-room would understand Chinese?
Yes Argumentation 5 S: +/(p⇢(I⇢E)), ㄱ(I⇢E)) 6 D: ?/ㄱ(I⇢E) ... ARG DT
Well, obviously in the scenario Searle-in-the-room would not understand Chinese!
Is it that obvious that he wouldn’t? ...
Fig. 4 Popa’s discussion structures (p. 88).
In my opinion, what remains to be investigated is the nature of the arguments put forward in scenario-based argumentation, and whether they rest upon a systematically identifiable form of reasoning and of arguing by employing specific schemes. This issue will be dealt with in Sections 3 and 4.
2.3 Reasoning and arguing
As we have seen, one can make different uses of thought experiments, from simply illustrating a claim in more vivid terms, to arguing in defence of its acceptability or unacceptability. Since the debate regarding the epistemic value of scenario-based thought experiments is grounded in the consideration of their heuristic or argumentative use, I believe it is fundamental to highlight the distinctive characteristics of the two phenomena of reasoning and of arguing. As is the case for thought experiments and imaginary scenarios, also argumentation and reasoning are tightly linked activities, and it is has proven hard to
clearly distinguish one from the other. Also, these two terms have been characterized differently by different scholars.
Informal logic perspectives have offered several distinctions between the two phenomena. Walton (1990, p. 401), for example, generally intends reasoning as a sort of abstract structure that can be individual and static, but also interactive and dynamic. This abstract structure is employed in various pragmatic contexts, among which the context of argumentation. Therefore, reasoning can be used in argument, that is intended as a social and interactive activity aimed at persuasion ( ibidem). Reasoning, so intended, is the process of inference that takes place before the actual argumentative activity, that links certain propositions with others - on the basis of warrants or rules - in order to arrive at a conclusion (p. 402). In other words, the reasoner rationally directs her prior knowledge, her beliefs or her intentions towards a conclusion, that can be another belief - or some other propositional attitude - or it can be an action.
Reasoning is the making or granting of assumptions called premises(starting points) and the process of moving toward conclusions (end points) from these assumptions by means of warrants. A warrant is a rule or frame that allows the move from one point to the next point in the sequence of reasoning. The term 'warrant', used instead of the more familiar (but narrower) term 'rule', is appropriate because of the existence of frame-based, and other kinds of non-rule-based reasoning (Walton, 1990, p. 403. Emphasis in original).
The presence of non-rule based reasoning entails that logic should not be restricted at evaluating the validity of abstract forms of reasoning, but it should involve the consideration of the how reasoning is in fact used within different pragmatic contexts (p. 402). In the context of argumentative exchanges, therefore, reasoning should be intended as being employed in discourse, and not as being identifiable with discourse. It is also important to underline that the logical study of reasoning is usually concerned with assessing the validity or truth of the sequential inference processes from premises to conclusions; natural reasoning, instead, should be intended as a more general thought process, as a combination between an abstract rule - or warrant - of reasoning and a semantic relation between concepts (Konstantinidou & Macagno, 2012, p. 1075). The use of reasoning in the specific context of argumentation, on the other hand, usually proceeds from the opponent’s premises to one’s own conclusions, hence serving a probative function in providing arguments to support a claim (Walton, 1990, p. 404). In this pragmatic context, reasoning moves forward and is guided by the specific goal of successfully justifying the conclusion, so that the validity of the reasoning process no longer rests on its logical truth, but on its adequacy with respect to achieving its purpose (p. 404).
Since reasoning can have different uses and purposes, Walton (p. 405) identifies four different categories through which reasoning can be defined, each category comprising of opposing pairs. According to the first category reasoning can be ‘monolectical’, when an individual reasons alone, or ‘dialectical’, when two or more participants reason interactively by deriving steps for their reasoning from those of the other(s). The second category sees reasoning as being ‘alethic’, when based on the truth value of propositions, or ‘epistemic’, when the reasoner draws upon premises that are known to be true. Thirdly, we have ‘static’ reasoning when the premises involved are fixed and cannot undergo change, and ‘dynamic’ reasoning, where premises can change at every subsequent level, also depending on changing circumstances. Finally, reasoning can be categorized as ‘practical’, when it is aimed at providing a line of conduct for the agent in a particular situation, or as ‘theoretical’, when the agent seeks evidence for or against the truth of a proposition.
As mentioned above, logical approaches tend to see reasoning as forms of arguments, in the sense that arguments themselves are understood as a set of statements combined in a premise-conclusion structure and that follow from one another, providing support for the truth of the conclusion (Copi, 1986, cited in Walton, 1990, pp. 408-409). Informal logicians, instead, draw a more clear distinction between the two activities and intend arguments as the reasons provided in support of a certain conclusion. Hence, what is highlighted is the purpose that such reasons serve in the communicative exchange and the dialectical and interactive aspect of the process. Along these lines, according to Walton,
Argument is a social and verbal means of trying to resolve, or at least to contend with, a conflict or difference that has arisen or exists between two (or more) parties. An argument necessarily involves a claim that is advanced by at least one of the parties. In an asymmetrical case, one party puts forward a claim, and the other party questions it. In a symmetrical case, each party has a claim that clashes with the other party's claim (Walton, 1990, p. 411).
In this perspective, arguments are not simply reasons, as they represent different ways of resolving a conflict of opinion. According to Walton, most of the times arguments occur in dialogue, and the type of reasoning associated with dialogical argumentation is dialectical reasoning (pp. 411-412). By ‘dialectical’, then, Walton intends reasoning that is not solitary, and is therefore at the basis of argumentative exchanges because these involve more than one party. I believe that the two notions of dialectical and dialogical are crucial to mark the distinction between reasoning and argumentation, and they can be very useful to identify how these are used in scenario based thought experiments. However, I believe that Walton’s definition of the terms does not account for certain important aspects of the phenomena, that
are, instead, taken into consideration by Baumtrog (2018) and by the pragma-dialectical theory of argumentation.
According to Walton, argumentation by definition occurs in a dialogical context, and as such involves dialectical reasoning, hence taking ‘dialectical’ and ‘dialogical’ to be almost synonyms. Instead, according to Baumtrog (2018) the notion of ‘dialogical’ involves the presence of another party to which the communicative exchange is directed to, whereas 1 ‘dialectical’ involves the presence of conflicting points of view in the exchange. From this perspective, the notion of ‘dialectical’ is what marks the difference between reasoning and arguing. Similarly to Walton, Baumtrog (2018, p. 90) allows for the possibility of reasoning being monological but also dialogical (or ‘monolectical’ and ‘dialectical’; Walton, 1990, p. 405), because it is possible for more than one individual to reason together in order to develop or illustrate a conclusion. An example could be two or more academics reasoning together in the design of a constructive thought experiment (Brown, 2011, pp. 35-40), therefore drawing onto each other’s inferences in order to arrive at a hypothesis or to clarify aspects of a theory through the consideration of an imaginary scenario. This can be an example of employing reasoning processes to design thought experiments with heuristic purposes. The important specification suggested is that the process of reasoning does not involve another party’s disagreement or doubt, as it is identified with the line of inference from premises to conclusion (Baumtrog, 2018, p. 88).
Conflict, doubt or criticism is what distinguishes argumentation from reasoning, and what makes it ‘dialectical’. Therefore, prerogative of argumentation is not being interactive - because reasoning too can be dialogical - but being somehow ‘conflictual’ ( ibidem). What counts as an argument, in this perspective, is not a premise-conclusion structured set of propositions, because these could also correspond to forms of reasoning used to deduce a hypothesis. Instead, “where reasons are used to support a conclusion which is in opposition to another claim or which is produced as a result of doubt, ‘argumentation’ can be used to capture the conflictual character” (p. 89). Along these lines, as previously suggested by Walton’s distinction between reasoning and argumentation, also these distinctions can be applied to the two uses that are made of scenario-based thought experiments. In fact, following Baumtrog’s notions, one or more thought experimenters may reason together in order to deduce a hypothesis - as is done, for example, in mathematical or philosophical proofs - hence constructing heuristic thought experiments. Instead, when providing a dialectical defence of a standpoint or advancing criticisms against a standpoint through a
1 By this, Baumtrog (2018) does not mean that a dialogical or a dialectical exchange must occur between more
that one person. In fact, he argues for the possibility of a dialectical and quasi-dialogical argumentative exchange also with one self.
scenario-based thought experiment, the experimenter is making an argumentative use of this tool.
Finally, the pragma-dialectical approach suggests a definition of argumentation that also aims at overcoming the typical logic-based view of arguments as sets of propositions organized in a premise-conclusion combination. In fact, argumentation is intended as part of a rule-governed critical discussion between two or more parties that is aimed at resolving a difference of opinion and that occurs in context (van Eemeren & Grootendorst, 1984, p. 18). Surely, the relation between premises and conclusions remains crucial, but many other aspects are taken into account, such as goal-directedness, situational contexts, types of speech acts involved and persuasive intentions. In this perspective, verbal expressions serve as arguments only in certain contexts, namely in those communicative processes that aim at resolving a conflict of opinion (van Eemeren et al., 2001, p. 17). Therefore, also in the pragma-dialectical theory, argumentation is understood as a dialectical activity (as well as externalized, socialized and functionalized; van Eemeren & Grootendorst, 1984, pp. 1-17). The dialectical aspect of argumentation makes it possible to view this activity as a discursive situation between two or more parties, aimed at establishing the (un)acceptability of a claim against the reaction of an antagonist (p. 17). This characterization of argumentation also involves an interpretation of fallacies in a way that departs from the formal logic theories. These regard fallacies as errors of reasoning, whereas pragma-dialectics construes fallacies as discussion moves that prevent the exchange to achieve its goal of resolving the dispute at issue (van Eemeren et al., 2001, p. 21).
The pragma-dialectical theory does not excessively focus on defining reasoning, as its aim is to provide new crucial concepts with regards to argumentative exchanges. For this reason, the metatheoretical principle of ‘externalization’ implies that the argumentation theorist should not focus on the internal reasoning processes that underlie the arguments presented, but rather on the quality of the arguments themselves and on the commitments made by the parties involved (van Eemeren & Grootendorst, 2004, p. 54). It follows that reasoning is implicitly understood as something that occurs before the activity of argumentation, as an internalised activity that guides the production of arguments. Despite the lack of an explicit distinction between the two processes at issue, the pragma-dialectical approach allows argumentation analysts to better understand the difference between heuristic and argumentative uses of thought experiments. In fact, by reconstructing thought experiments on the basis of the ideal model of a critical discussion, destructive thought experiments have been shown to manifest typical argumentative features and as being based on criticism against another party’s theoretical claim. 2
3 Reasoning analogically in scenario-based thought experiments
In this Section I will use some insights from the previous one in order to identify what kind of reasoning is involved when designing destructive and constructive thought experiments. In
light of the debate surrounding the epistemic status of thought experiments, I will maintain Brown’s distinction (2011) to show that, in both cases, forms of analogical reasoning are at play. In order to clarify this, I will suggest some modifications to the reasoning schemes identified by Häggqvist (2009) and Popa (2017) as underlying the cognitive process of constructing a thought experiment. Finally, I will identify a problematic aspect of such schemes, a problem that also is grounded in the analogical nature of the reasoning involved.
3.1 Reasoning schemes: the missing premise problem
As shown in Section 2, reasoning can be employed in a variety of contexts and with a variety of purposes. It can be explanatory, descriptive, investigative or meditative; it can be aimed at problem solving, at understanding or at reason giving. In this last regard, it may be used as a basis for developing arguments with a justificatory purpose. The differences between the activities of reasoning and of argumentation previously outlined (see Section 2.3) are helpful to avoid a common confusion of the two notions with one another, but they also elucidate how much the two phenomena are related. This link between reasoning and argumentation becomes explicit, for example, though the dialectical tool of argumentation schemes. This is because argument schemes can be intended as combining the two concepts, being patterns of argument that represent the logical and semantic relation between premises and conclusion (Konstantinidou & Macagno, 2012, p. 1070; Baumtrog, 2018, p. 88). These are intended as the modes of reasoning that underlie speakers’ argumentation to move from premises to conclusions.
The patterns of arguments were labeled as argumentation schemes, abstract structures merging together an inferential relation, or rule, systems of reasoning and logical rules. Argumentation schemes do not simply correspond to the warrant leading from the premise to the conclusion [...]. They are stereotyped forms of reasoning structures proceeding from a specific rule of inference (such as modus ponens), where both the premises and the conclusion are represented as abstract propositions (Konstantinidou & Macagno, 2012, p. 1074).
Following the aforementioned ideas that an exchange becomes argumentative when it involves doubt, criticism or justification of a point of view - aimed at resolving conflict of opinion - argument schemes can be intended as moving on from representing forms of reasoning to having an argumentative function when they are accompanied by a matching set of critical questions, to assess the validity of the argument scheme employed (Walton, 1990, p. 407; van Eemeren & Grootendorst, 1992, p. 86). The identification of argument schemes in
scenario based thought experiments will be dealt in more depth in Section 4. Right now, I will focus on identifying what kind of reasoning schemes underlie the development of such experiments, and show that relations of analogies are fundamentally at play.
It is important to stress that, despite the fact that arguing and reasoning are two fundamentally different activities, reasoning is not at all excluded from argumentative processes. Therefore, it is not a matter of confusion to talk about the reasoning underlying both constructive and destructive thought experiments - both heuristic and argumentative ones. In other words, it is one thing to talk about the nature of the reasoning that underlies the scenario based experiment; another is to talk about the use that one makes of such an experiment, once it has been constructed through the activity of reasoning. Having clarified this, reasoning schemes can be identified to make explicit this process.
As outlined in the previous Section, Häggqvist (2009) and Popa (2017) are two among the scholars that have, indeed, identified such reasoning schemes at the basis of the development of thought experiments. These schemes render explicit the intentions of the experimenter that applies a theoretical claim under discussion onto an imaginary scenario, hence considering the consequences of such application and drawing conclusions. Both authors focus on the cognitive process at the basis of destructive thought experiments: the reasoning takes the form of modus tollens, hence the schemes translate into abstract sentences what happens when a theoretical claim is rejected. The rejection of the claim depends directly on the entailments that follow from its application onto a fictitious scenario. If the claim were acceptable, a certain event should follow within the scenario; because such event does not follow, the claim is deemed unacceptable.
(a) ◇ C (b) 1. ¬ p
T ⊃( C⬜➝ W)
C⬜➝ ¬W 1.1 ¬ (I ➝ E)
________________ 1.1’ p ➝ (I ➝ E)
¬ T
Figure 5 Reasoning schemes by Häggqvist (a) (2009, p. 63) and Popa (b) (2017, p. 83).
The fact that the schemes presented in Figure 5 represent the reasoning behind destructive thought experiments does not mean that a scheme for constructive thought experiments cannot be identified. In my opinion, this can be done by transforming them into the form of
modus ponens. In this manner, they reflect a reasoning process that intends a thesis to hold, if a certain event follows in the scenario; since such event does, indeed, follow, the claim must be deemed acceptable.
(a) ◇ C (b) 1. p
T ⊃( C⬜➝ W)
C⬜➝ W 1.1 (I ➝ E)
________________ 1.1’ p ➝ (I ➝ E)
T
Figure 6 Modified version of the reasoning schemes underlying constructive thought experiments.
If we consider all four of the schemes outlined above, I believe that they can be seen as representing a form of reasoning by analogy. What makes it possible to identify the analogical nature of this form of reasoning is the use of the imaginary scenario. This is because the original theoretical claim is transposed onto the fictitious setting, and on the basis of the fact that something follows - or fails to follow - in the scenario, the claim is accepted or rejected. As a matter of fact, analogies operate in the same manner: they are meant to capture parallel relations that hold across different kinds of situations (Gentner & Maravilla, 2018, p. 186). What happens in scenario based thought experiments, then, is precisely that a complex target claim - the thesis under discussion - is mapped onto a more familiar scenario, in which its implications can be more easily taken under consideration. What is crucial, therefore, is that the elements of the scenario are believed to be related with each other in a way analogous to how the elements of the real world are related to each other. As a consequence, the application of a hypothesis in a fictitious scenario can be compared to the application of the same hypothesis in reality.
So far, the schemes presented above seem to reflect unproblematically the cognitive process at play. However, they all seem to fail to identify explicitly a further relation of analogy, and one on which the validity of the whole reasoning relies upon. The problem that should be noted is that they do not address the actual comparability of the imaginary scenario depicted in the thought experiment and the real world situation, meant to be parallel to it. This comparability is assumed from the start as the grounds from which the whole reasoning process takes off, and yet it is only left implicit. The problem of not rendering the analogy between the scenario and the real situation explicit, is that the transfer of the conclusions in the scenario onto the conclusions in reality remains apparently unjustified. This is why I believe that this warrant should find a place in the schemes presented above. Also, it should be included because the experimenter, in the process of designing a scenario based thought experiment, is guided by the idea that the acceptability of her standpoint relies precisely upon this relation of analogy, otherwise the reasoning process would be instantiated by a different kind of pattern.
The same criticism can apply to the three forms of argument identified by Häggqvist (2009, pp. 65-67) as connected to the performance of thought experiments. Fig. 7 through Fig. 9, represent possible defences of the claim under discussion - therefore, potential responses to destructive thought experiments - given its implications when applied to the scenario. Despite the fact that Häggqvist focuses of thought experiments aimed at the rejection of a claim, I believe that by operating slight modifications to the schemes presented below, these could apply also to constructive thought experiments. Unfortunately, for reasons of brevity it is not possible to deal with this matter in the present study.
(1) T ◇ C
T⊃ ( C◻ ➝W) __________________ ◇C & ( C◻ ➝W)
Figure 7 Häggqvist’s first argument scheme (2009, p. 66).
In this first argument scheme, the responder is suggesting that the claim (T) should not be rejected because the event (W) under consideration does, in fact, follow in the scenario (C). And, if (T) were the case, also (C◻ ➝W) would be the case.
(2) T ◇ C ( C◻ ➝ㄱW) __________________ ㄱ (T⊃ ( C◻ ➝W) )
Figure 8 Häggqvist’s second argument scheme (2009, p. 66).
In this second pattern, it is argued that the event (W) does not follow in the scenario - in line with the basic reasoning pattern outlined in Figure 1 and Figure 5. However, it is argued that it is not the case that if the claim (T) were to hold, then the event (W) should follow in the scenario (C); therefore, (T) should not be rejected.
T⊃ ( C◻ ➝W) ( C◻ ➝ㄱW)
__________________ ㄱ ◇C
Figure 9 Häggqvist’s third argument scheme (2009, p. 67).
In this last scheme, the scenario itself is argued to be impossible. Therefore, the experimenter’s claim of what should follow from (T) in the scenario, andthe fact that (W) does not actually follow in it, are both acceptable premises.
In all three of the argument schemes identified by Häggqvist, the problem of the missing premise applies in the same way as it applies to his basic cognitive scheme. Indeed, the comparability between the fictitious situation depicted in the thought experiment and the situation in the real world is never explicitly assessed, but it is assumed from the start. More precisely, the original theoretical claim and its parallel claim, applied in the scenario, are not explicitly presented as analogous. Also in case (3), the impossibility of the scenario does not seem to be supported clearly. In other words, the scenario is not deemed as unacceptable because it is not adequately comparable to the situation at hand, it is claimed as impossible without a clear explanation as to why. This relation of analogy is the fundamental assumption on which these schemes all implicitly rest upon, and yet it finds no place within them.
3.2 Reasoning quasi-dialectically in destructive thought experiments
In Section 2.3 I have shown that, according to Baumtrog (2018) and to the pragma-dialectical theory (van Eemeren & Grootendorst, 1984; 2004; van Eemeren et al., 2001), what marks the difference between reasoning and arguing is the element of conflicting opinions that characterizes the latter. Argumentation, therefore, is dialectical in nature, and can be dialogical - hence involving another party - or quasi-dialogical - when it occurs within oneself (Baumtrog, 2018). As for reasoning, it may be monological - when involving a single individual - or dialogical - when more that one individual reason together, in order to develop or illustrate a conclusion. Therefore, the fundamental element that distinguishes the activity 3 of arguing from the activity of reasoning is the fact that the former involves doubt, criticism and conflict, whereas the latter does not.
Monological or dialogical reasoning, in Baumtrog’s terms, can be identified as underlying the process of designing a constructive thought experiment. If we take Thomson’s violinist thought experiment as an example, the author can be said to resort to a form of reasoning involving only herself, or as interacting dialogically with other scholars in identifying and clarifying aspects of her central claim through the use of a scenario-based experiment. Here, no doubt seems to be involved, no criticisms are anticipated and no conflict of opinions has yet occurred. Surely, criticism and need for justification may arise, but they are not involved at the reasoning stage. I believe that, with regards to destructive thought experiments, instead, the same cannot be exactly stated, as some form of conflictuality is involved already at the level of reasoning that underlies the further argumentative use made of the scenario-based argumentation developed.
Considering again Baumtrog’s perspective (2018, p. 88), the dialectical nature of argumentation stems from the use of reasons to support a certain hypothesis in opposition to another claim or opinion, os as a result of mere doubt of a claim under consideration. Therefore, if we imagine doubt ot criticisms arising at the level of reasoning, this reasoning will be considered argumentative in character. However, I believe that this element of conflict can also be identified in certain reasoning processes, that cannot be characterized as argumentative. In my opinion, this type of reasoning takes place when a scholar is developing a destructive thought experiment. As seen in Section 2, ‘destructive’ thought experiments aim at rejecting a theoretical claim put forward by another party (Brown, 2011, p. 33), and this is why they make an argumentative use of thought experiments, rather than a heuristic one. To a certain extent, however, I believe that some dialectical features can be identified also in the reasoning process that guides the experimenter’s development of the scenario, may it be a monological or a dialogical form of reasoning - although in dialogical reasoning it is more likely that the switch to an argumentative exchange be more immediate. In the development of a destructive thought experiment the intention is a conflictual and dialectical one, because the reasoning process that guides the search and invention of arguments is already characterized by the intention of rejecting a certain claim. This is why, in my opinion, this form of reasoning can be intended as ‘quasi-dialectical’.
What I mean by quasi-dialectical reasoning, therefore, is similar to what Baumtrog means by quasi-dialogical argumentation (2018, p. 90). The latter occurs when a single individual is in the act of imagining challenging (or responding to challenge of) another party’s point of view. According to Baumtrog, to consider reasoning as dialectical is to already intend it as involving some level of argumentation (p. 89). This is because it would be a process of reasoning developed in response to (anticipated) doubt from - or against - another party. However, I believe that on some level, reasoning can be intended dialectically when employed in the design of a destructive thought experiment. Of course, analogical
reasoning can be identified also in constructive thought experiments, that have a heuristic nature. In fact, these kind of thought experiments envision scenarios in order to arrive at some conclusion or to illustrate and elucidate certain aspects of a claim that are problematic, by mapping them onto more familiar and understandable situations. I would now like to move on to the consideration of how analogies operate in scenario-based argumentation, and investigate whether schemes based on analogy can be identified also on the argumentative level.
4 The role of analogy in scenario-based argumentation
In this Section, I will focus on a critical analysis and evaluation of two important accounts of arguments based on relations of analogy, in their applicability to scenario-based thought experiments. Firstly, I will present van Laar’s scheme for arguments from parallel reasoning
(2014). This argumentative pattern will be considered as an appropriate translation of the reasoning schemes analysed in Section 3.1. Secondly, I will focus on Walton’s scheme for arguments from analogy (2014), that, also, will be critically assessed. The first scheme will be applied to an example of destructive thought experiment, the second oone to an example of a constructive thought experiment.
4.1 Argument schemes: the missing premise problem in arguments from parallel reasoning
As mentioned throughout this study so far, Baumtrog (2018) suggests a distinction between argumentation and reasoning on the basis of the fact that the former involves some degree of conflictuality that the latter lacks. When reasons are provided in support of a claim that is in conflict with another hypothesis - or that is a result of doubt - we have an argumentative exchange. This element of conflictuality also seems to mark the difference between the heuristic and the argumentative use of thought experiments, and, therefore, between constructive and destructive ones. In my opinion, both at the reasoning and at the argumentative level we can identify relations of analogy between the theoretical claim under discussion - as well as its implications in the real world - and its implementation within an imaginary scenario - as well as its implications in the scenario itself. Indeed, when scenario-based thought experiments are used argumentatively, the force of the argument relies on the acceptance, from the listener’s part, of the facts that entailments in the scenario correspond to entailments in the real-life situation.
As we have seen above, it is possible to identify reasoning schemes that guide the cognitive process of constructing such a thought experiment and what distinguishes this type of schemes from argument schemes, used for justificatory purposes, is the presence of matching critical questions. Intuitively, it seems that if the reporter were to pose critical questions to the reasoning process discussed in Section 3.1, they would address the legitimacy of considering the implications of a theory in the scenario - and, therefore, its (un)acceptability - as comparable to the implications of the theory in the real world. In other words, what justifies the parallel between accepting or rejecting the claim in the scenario and doing so in the real-life situation, especially if the scenario depicts highly improbable states of affairs? This sort of questions call for answers that rest upon the presence of analogical relations between certain elements referred to in the arguments. Once the discursive exchange has reached this level of critical examination, becoming more of an interactive and dialectical process, we are no longer in the domain of reasoning schemes but, rather, of argumentation schemes.
The reason why I believe that the relations of analogy that are at play at the reasoning level are also at play at the argumentative level, is that during the argumentative exchange which makes use of scenario-based argumentation, the reviewer is justifying her claim based on the implications of accepting - or rejecting - such a claim in a parallel, fictitious situation. As in the reasoning process seen above, the reviewer aims at establishing analogical relations between the implications of accepting the claim within the scenario, and the implications of accepting it in the real world. In destructive thought experiments, for example, if accepting the theory under discussion in the scenario entails absurd or false implications, then it is argued that it will yield absurd or false implications also in real life, and should, therefore, be rejected. Considering theChinese roomthought experiment (see Section 2.1) we can interpret ‘Searle-in-the-room’ as the analogue, actual digital computers as the target subject and not-understanding as the assigned predicate. The relations that hold between the elements of the scenario that make us assign the predicate of not understanding to ‘Searle-in-the-room’ are considered as corresponding to the relations that hold between elements of the actual, real-world functioning of computers. Therefore, similarly to what Juthe claims (2005, p. 4), the target subject and the analogue are comparable with respect to the assigned predicate only if the elements that determine such predicate in the analogue correspond to the elements of the target subject. From this, it follows that the entailments in one situation correspond to entailments in the other situation, and for this to be argued for it is not necessary that the target subject and the analogue actually share properties, structures or logical form (p. 9). On this basis Juthe operates an important distinction between arguments from similarity and arguments from analogy: in the former, two objects are claimed to be analogous with respect to their properties, whereas in the latter, the objects are considered analogous on the basis of the properties holding between their elements (p. 7).
To consider these relations as comparable is guided by the forms of reasoning identified previously in this study. I believe that an argument scheme that well translates them onto the argumentative level is van Laar’s argument from parallel reasoning (2014). Arguments from parallel reasoning are not ones in which the arguer defends a relation of similarity between objects or properties of objects. Instead, in arguments from parallel reasoning the protagonist holds that the conclusion of her argument has to be considered as valid because of the validity of another argument, that the antagonist has previously deemed as acceptable. Mutatis mutandis, an argument is deemed as invalid, due to the failure of another argument, that the antagonist has previously rejected (p. 94). As a consequence, these arguments are said to be ‘meta-dialogical’, being arguments about arguments (p. 93). At a close look, this seems to be what characterizes scenario-based argumentation in destructive thought experiments. The fundamental - and problematic - aspect is that, in this argumentation type, the standpoint at issue does not regard the similarity between two cases
