Argument and Argumentation

Terminological Clarifications

• Definition of an Argument:

  • An argument can be defined as a complex symbolic structure in which specific parts, known as premises, provide support for another part, known as the conclusion.

  • Alternatively, an argument can be viewed as a complex speech act comprising:

  • One or more acts of premising (which assert propositions in favor of the conclusion).

  • An act of concluding.

  • A stated or implicit marker (such as "hence" or "therefore") that indicates that the conclusion follows from the premises (Hitchcock 2007).

  • The relation of support between premises and conclusion can be cashed out in several distinct ways:

  • The premises may guarantee the truth of the conclusion.

  • The premises may make the truth of the conclusion more probable.

  • The premises may imply the conclusion.

  • The premises may make the conclusion more acceptable or assertible.

• Distinction Between Argument and Argumentation:

  • For theoretical purposes, arguments can be analyzed as freestanding entities abstracted from their contexts of use in actual human activities.

  • Argumentation refers to the instances of exchange of arguments within human communicative practices and activities where arguments occur as communicative actions.

  • Argumentation is defined as the communicative activity of producing and exchanging reasons in order to support claims or defend/challenge positions, especially in situations of doubt or disagreement (Lewiński & Mohammed 2016).

• Argumentation as a Special Dialogue:

  • Argumentation is best conceived as a kind of dialogue, which can occur across various setups (Lewiński & Aakhus 2014):

  • Intrapersonal dialogue (arguing with oneself).

  • Monological formats such as long speeches or written texts (articles or books) aimed at a silent audience.

  • Group exchanges rather than strictly dyadic interactions.

  • Argumentation is a special kind of dialogue; most everyday dialogues (such as asking for the time or sharing vacation details) are not instances of argumentation.

  • Argumentation only occurs when, upon making a claim, a person receives a request for further support in the form of reasons, or estimates that further justification is required (Jackson & Jacobs 1980; Jackson 2019).

  • These situations initiate dialogues of "giving and asking for reasons" (Brandom 1994; Bermejo Luque 2011).

  • Epistemically, because most knowledge is acquired from others, argumentation serves as a critical mechanism to filter incoming information rather than accepting claims uncritically (Sperber, Clément, et al. 2010).

• Relationship Between Reasoning and Argumentation:

  • Reasoning is defined as the process of reaching conclusions on the basis of careful, reflective consideration of available information (i.e., an examination of reasons).

  • Traditional/Widespread View: Reasoning and argumentation are related through their shared concern with reasons, but remain fundamentally distinct phenomena:

  • Reasoning belongs to the internal, mental realm of thinking, where an individual infers new information from available data.

  • Argumentation belongs to the public realm of exchanging reasons expressed in language or symbolic media intended for an audience.

  • Alternative/Internalization View: Reasoning and argumentation are two sides of the same coin; what is known as reasoning is by and large the internalization of public practices of argumentation (MacKenzie 1989; Mercier & Sperber 2017; Mercier 2018).

Types of Arguments

• Broad Classification:

  • Arguments come in multiple kinds, differentiated by their inferential guarantees and structural characteristics:

  • Deductive Arguments: Arguments where the truth of the premises is supposed to guarantee the truth of the conclusion.

  • Ampliative Arguments: Arguments where the conclusion goes beyond what is logically contained in the premises (a distinction introduced by C.S. Peirce). Major classes include:

    • Inductive Arguments: Arguments where premise truth makes conclusion truth more likely without ensuring absolute certainty.

    • Abductive Arguments: Arguments that infer explanatory hypotheses from observed facts.

  • Analogical Arguments: Arguments featuring across diverse philosophical traditions that infer shared properties based on similarities between domains.

  • Fallacious Arguments: Arguments that appear legitimate and persuasive but are in fact logically defective.

Deduction

• Validity and Soundness:

  • A deductively valid argument is one where the truth of the premises necessitates the truth of the conclusion; the conclusion cannot be false if the premises are true.

  • A sound argument is a valid deductive argument whose premises are also factually true.

  • Classic Syllogistic Example:

  • Premise 1: All humans are living beings.

  • Premise 2: All living beings are mortal.

  • Conclusion: Therefore, all humans are mortal.

  • In terms of possible-worlds semantics, in a deductively valid argument, the conclusion holds in all possible worlds where the premises hold, allowing absolute certainty.

• Property of Monotonicity:

  • Deductive arguments typically exhibit monotonicity: if premises $A$ and $B$ deductively imply conclusion $C$, then adding any arbitrary premise $D$ will not invalidate the argument.

  • Formal notation: If $(A \text{ and } B) \rightarrow C$ is deductively valid, then $(A, B, \text{ and } D) \rightarrow C$ is equally deductively valid.

  • Exceptions: In relevant logic, monotonicity does not hold because arbitrary added premises can violate necessary relevance criteria between premises and conclusions.

• Logical Systems and Mathematical Roots:

  • Deductive arguments are analyzed in classical propositional and predicate logic, as well as subclassical systems (e.g., intuitionistic and relevant logics).

  • These formal systems capture the relation of necessary truth-preservation, originally developed to model mathematical proofs in the work of Frege, Russell, Hilbert, Gentzen, and foundational Greek mathematics (Euclid's Elements; Netz 1999; Dutilh Novaes 2020a).

  • Mathematical argumentation involves broader elements beyond pure deduction (Aberdein & Dove 2013).

• Theoretical Challenges to Necessary Truth-Preservation:

  • Logical Paradoxes: Paradoxes such as the Liar Paradox and Curry's Paradox suggest that deductive validity and necessary truth preservation can diverge (Beall 2009; Field 2008).

  • Contingent Logical Truths: Theories of contingent logical truths challenge strict necessary truth preservation (Kaplan 1989; Nelson & Zalta 2012).

  • Warrant/Assertibility Preservation: Alternative accounts propose that deductive transitions preserve warrant or assertibility rather than objective truth (Restall 2004).

  • Paraconsistent Logic: Preservationist approaches to paraconsistent logic posit that deductive relations preserve the coherence (or incoherence) of premise sets (Schotch, Brown, & Jennings 2009).

• Philosophical Justification and the "Scandal of Deduction":

  • Justifying deduction involves resolving the modal nature of necessity (metaphysical, logical, linguistic, epistemic; Shapiro 2005) and constructing non-circular foundations (Dummett 1978).

  • The Scandal of Deduction: Because deductive arguments are non-ampliative, their conclusions contain no information not already present in the premises, raising the problem that deduction may not be genuinely informative (Sequoiah-Grayson 2008).

• Historical Role and Modern Critiques:

  • Aristotle formulated the first comprehensive theory of deductive argumentation in Prior Analytics and scientific demonstration in Posterior Analytics.

  • Spinoza utilized the more geometrico ("geometrical style") in his Ethics, exemplifying the philosophical ideal of mathematical certainty.

  • Deductivism has faced criticism for over-concentrating argumentation studies on deductive validity (Bermejo-Luque 2020).

  • Psychology of Reasoning: Research shows everyday human reasoning ("in the wild") relies on Bayesian probabilistic principles rather than formal deductive rules (Oaksford & Chater 2018; Elqayam 2018).

  • Artificial Intelligence: Computer scientists rely on non-monotonic reasoning (Reiter 1980) and defeasible reasoning (Pollock 1987), positioning deduction as an exception rather than the general rule in human reasoning practices (Dutilh Novaes 2020a), though some philosophers hold that deductive reasoning remains widespread (Shapiro 2014; Williamson 2018).

Induction

• Definition and Core Mechanics:

  • Inductive arguments draw conclusions about future instances and general principles based on observations of past instances and regularities.

  • Example: Observing that the sun has risen in the east every day leads to the conclusion that it will rise in the east tomorrow, and to the general principle "the sun always rises in the east."

  • Inductive arguments rely on statistical frequencies to generalize beyond the observed sample cases to unobserved cases.

• Cogency and Support Degrees:

  • A good inductive argument is termed cogent when its premises provide strong, though non-maximal, support for the conclusion.

  • Unlike deduction, the degree of support in an inductive argument is never absolute; the conclusion can be false even if the premises are true.

  • In possible-worlds terms, an inductive conclusion holds in a significant proportion of possible worlds where the premises hold.

• Historical Trajectory:

  • Aristotle conceptualized induction (epagoge) as a progression from particulars to universals in dialectic and scientific contexts.

  • Deductivist models dominated Aristotelian traditions until the early modern Scientific Revolution, when observation and experimentation were formalized as scientific pillars by Francis Bacon (1561–1626).

• Uniformity Principle and Cognitive Role:

  • Induction relies on the Uniformity Principle: the assumption that the natural world operates with statistical regularity.

  • Generalizing from observed frequencies is considered the most fundamental principle of human cognition (Clark 2016).

• Critiques of Induction:

  • Hume's Problem of Induction: David Hume established in A Treatise of Human Nature that justifying induction requires an argument validating the Uniformity Principle (Howson 2000).

  • The justification cannot be deductive, because a change in the course of nature is conceivable.

  • The justification cannot be probable/inductive, because probable arguments already presuppose the Uniformity Principle, creating a circular argument.

  • Therefore, the Uniformity Principle cannot be rationally justified.

  • Harman's Critique: Gilbert Harman (1965) argued that enumerative induction is either not always warranted or is merely an uninteresting special case of Inference to the Best Explanation (IBE), denying induction status as an autonomous inference form.

• Modern Responses and Social Domain Applications:

  • John D. Norton formulated the material theory of induction (Norton 2003) as a contemporary response to Hume.

  • Hume's problem reemerges in modern debates over using group statistical frequencies to make inferences about individuals, such as implicit bias research and predictive law enforcement algorithms (Jorgensen Bolinger 2021).

Abduction

• Definition and Applications:

  • An abductive argument is one where, from observing a few relevant facts, a conclusion is drawn regarding what best explains those facts.

  • Abduction is widespread in science, legal reasoning, medical diagnosis, and explainable artificial intelligence (Josephson & Josephson 1994).

  • Legal Example: A prosecutor summarizing circumstantial evidence to conclude that the defendant's guilt is the most plausible explanation.

• Comparison with Deduction and Induction:

  • Abduction is non-truth-preserving (alternative unexpected causes could explain the observed evidence).

  • Unlike induction (which focuses on forward-looking prediction from sample frequencies), abduction is primarily backward-looking, seeking to explain events that have already occurred.

  • Abduction connects seemingly independent events through causal or explanatory links, leveraging human cognitive tendencies to seek causal explanations (Keil 2006).

• Historical Conceptual Evolution:

  • While causal explanation dates to Aristotle's scientific demonstration, C.S. Peirce introduced "abduction" as a distinct third inferential class alongside deduction and induction.

  • For Peirce, abduction pertained to generating explanatory hypotheses within contexts of discovery.

  • Modern accounts place abduction within contexts of justification, framing it as Inference to the Best Explanation (IBE) (Lipton 1971 [2003]), though some scholars maintain a distinction between abduction and IBE (Campos 2011).

• Philosophical and Evaluative Challenges:

  • Humans exhibit a cognitive tendency to over-explain or impose overly simple causal patterns on complex phenomena (Lombrozo 2007; cf. Sober 2015 on parsimony).

  • Bas van Fraassen (1989) articulated a major critique of IBE, questioning whether explanatory superiority guarantees truth, as no a priori argument links explanatory quality directly to truth.

  • Defenders rely on empirical arguments demonstrating abduction's practical reliability, placing it on similar pragmatic ground as induction.

Analogy

• Structure and General Schema:

  • Arguments by analogy operate on the principle that if two items are similar in certain known respects, they are likely similar in other, unobserved respects.

  • Widely used across legal reasoning (precedent) and moral philosophy.

  • Moral Argument Example: If an advanced alien species farming humans for food is morally wrong, then by analogy, human farming of non-human animals for food is also wrong.

  • General Schema (where $S$ represents the source domain and $T$ represents the target domain):

  1. $S$ is similar to $T$ in certain (known) respects.

  2. $S$ has some further feature $Q$.

  3. Therefore, $T$ also has the feature $Q$, or some feature $Q^*$ similar to $Q$.

• Evaluation and Weakness:

  • The formal schema does not distinguish cogent from weak analogies; debates typically center on Premise 1 regarding whether $S$ and $T$ are sufficiently similar in respects relevant to feature $Q$

• Historical Traditions:

  • Greek Tradition: Plato used analogies extensively (e.g., in Gorgias, comparing rhetoric to pastry-baking and philosophy to medicine; Irani 2017); Aristotle analyzed analogy in Prior Analytics and Topics.

  • Chinese Tradition: Analogy served as a primary argumentation mechanism, especially among Mohist thinkers (Mohist canons).

  • Latin Medieval Tradition: Deeply analyzed in logic, theology, and metaphysics.

• Modern Philosophical Examples and Thought Experiments:

  • Judith Jarvis Thomson (1971): The Violinist Argument regarding the moral permissibility of abortion.

  • John Searle: The Chinese Room Argument regarding whether digital computers possess genuine understanding.

  • Achilles' Heel: Both famous arguments face critiques challenging whether the similarity between the source domain $S$ and target domain $T$ is strong enough to justify extrapolating properties (e.g., unplugging from a violinist vs. abortion; understanding Chinese symbols in a room vs. artificial intelligence).

Fallacies

• Definition and Historical Background:

  • Fallacies are arguments that appear legitimate and sound but are in fact logically defective.

  • Formal study originated in Aristotelian logic, particularly in Sophistical Refutations.

  • Competent argumentation requires both producing good arguments and identifying/avoiding fallacious ones.

• Key Fallacy Types:

  • Fallacy of Equivocation: Exploiting the ambiguity of a term or phrase used multiple times in an argument to draw an unwarranted conclusion.

  • Begging the Question (Petitio Principii): Structuring an argument where a premise and the conclusion express the exact same proposition in different words.

  • Appeal to Authority: Supporting a claim by citing an authority figure rather than providing substantive supporting reasons.

  • Ad Hominem Fallacy: Attacking personal characteristics or circumstances of an arguer rather than addressing their actual argument.

  • Fallacy of Faulty Analogy: Executing an analogical argument where the source and target domains lack sufficient relevant similarity.

• Theoretical and Philosophical Questions:

  • Determining precise criteria distinguishing fallacies from valid arguments (e.g., Bayesian accounts of fallacies).

  • Investigating cognitive causes of fallacious thinking.

  • Evaluating whether focusing on fallacies is an effective theoretical approach to argumentation (Massey 1981).