1. I will not discuss persuasive precedent or overruling.

2. The most adamant defenders of the rule-approach and critics of the analogy-approach are Larry Alexander and Emily Sherwin (e.g., Alexander, Larry, Constrained by Precedent, 63 S. Cal. L. Rev. 3 (1989)Google Scholar; Larry Alexander & Emily Sherwin, Demystifying Legal Reasoning (2008)) and the later Schauer (especially Schauer, Frederick, Why Precedent in Law (and Elsewhere) Is Not Totally (or Even Substantially) About Analogy, 3 Persp. Psychol. Sci. 454 (2008)CrossRefGoogle Scholar; but also Frederick Schauer, Thinking Like a Lawyer (2009)).

3. See, e.g., Edward Levi, An Introduction to Legal Reasoning (1949); Postema, Gerald J., A Similibus ad Simila: Analogical Thinking in the Law, in Common Law Theory 102–133 (Edin, Douglas E. ed., 2007)CrossRefGoogle Scholar; Rigoni, Adam, Common-Law Judicial Reasoning and Analogy, 20 Legal Theory 133 (2014)CrossRefGoogle Scholar; Lloyd Weinreb, Legal Reason: The Use of Analogy in Legal Argument (2005); Hunter, Dan, Reason Is Too Large: Analogy and Precedent in Law, 50 Emory L.J. 1197 (2001)Google Scholar.

4. Grant Lamond has most prominently criticized the rule-approach for its problems with distinguishing. See Lamond, Grant, Do Precedents Create Rules?, 11 Legal Theory 1 (2005)CrossRefGoogle Scholar. Others have followed him, e.g., Rigoni, supra note 3. The American Legal Realists brought special attention to the problems with deriving rules from opinions (for an overview, see, e.g., Leiter, Brian, American Legal Realism, in The Blackwell Guide to the Philosophy of Law and Legal Theory 50–67 (Golding, Martin P. & Edmundson, William A. eds., 2005)Google Scholar). Others who have drawn attention to these problems are, e.g., Rigoni, supra note 3; Schauer, supra note 2, at ch. 9; and Kent Greenawalt, Statutory and Common Law Interpretation (2013), at ch. 8.

5. See especially Schauer, supra note 2, and Alexander & Sherwin, supra note 2, at 70.

6. For example, Rigoni argues that the rule-approach does worse than the analogy-approach with respect to distinguishing and at best as well with respect to the constraining function of precedent. Rigoni, supra note 3. Alexander and Sherwin develop a rule-approach that, they admit, cannot account for distinguishing—but claim that this is a small price to pay given that the alternative cannot show that precedent is constraining. Alexander & Sherwin, supra note 2, at 114.

7. See, e.g., Joseph Raz, The Morality of Freedom (1986).

8. Compare id. at 35ff.

9. Compare id. at 38ff. It is important that the group of excluded reasons can be big or small—as long as some group of otherwise relevant reasons is excluded from consideration, the authoritative reason is protected.

10. The usual way to describe distinguishing is by saying that a judge may distinguish based on a difference between precedent-case and present-case. However, this formulation favors the analogy-approach. Therefore, I choose a neutral formulation.

11. This assumption is not shared by all theorists. Alexander and Sherwin give up on descriptive accuracy in favor of what they perceive to be a normatively better account when they advocate eliminating distinguishing. Alexander & Sherwin, supra note 2, at 114.

12. See, e.g., David Klein & Gregory Mitchell, The Psychology of Judicial Decision Making (2010).

13. See Schauer, supra note 2, at 7, for a discussion of this peculiarity of precedent.

14. See, e.g., Schauer, Frederick, Precedent, 39 Stanford L. Rev. 571 (1987)CrossRefGoogle Scholar for an overview of the values that can only be realized through reasoning by precedent if precedents are constraining.

15. Emily Sherwin and Larry Alexander cite this idea of a balance attainable through precedent but reject its possibility. Alexander & Sherwin, supra note 2, at 66. I am more optimistic.

16. Omychund v. Barker, 26. Eng. Rep. 15, 23 (Ch. 1744). Dworkin introduces this saying in the following way: “Sentimental lawyers cherish an old trope: they say that law works itself pure.” Ronald Dworkin, Law's Empire (1986), at 400ff.

17. Importantly, the kind of predictability that is demanded here is predictability based on promulgated legal sources. Theoretically, judicial decisions could be predictable even without publishing any precedent-opinion, and without the doctrine of binding precedent, if only the personality of the judges was well enough known. But that would not be predictability as a rule-of-law value because it would require those who predict to accept that it is the judges (and their personalities) who do the ruling, not the law. See Waldron, Jeremy, Stare Decisis and the Rule of Law: A Layered Approach, 111 Mich. L. Rev. 1, 13 (2012)Google Scholar.

18. Id. at 10.

19. Two decisions will be inconsistent with one another if they amount to treating the same situation differently. But because of the unlimited number of aspects (including the time and the place of the situation), no two situations are ever the same, strictly speaking. So any combination of decisions can be made consistent with each other, if the decision-maker is willing to treat any aspect of the situations as relevant for the decision. For example, a cruel parent might point out that her decision to reward one child for cleaning her room is not inconsistent with her decision to punish the other for cleaning her room because one of her children has a missing tooth, so it's different.

20. I am here deliberately choosing the vague term “aspect” of a case, which is meant to approximately denote some property/circumstance of, or relation within the case—something the case can be said or has been said to have. That a case has an aspect can be expressed either through precise categorical terms or through vague allusions, and what the aspect is that such a term or allusion refers to might not be clear to the speaker/writer and/or the hearer/reader. If I describe a situation and note “She had a weird je ne sais quoi,” then I have successfully denoted an aspect. Any situation, case, or object has uncountable aspects.

21. As I see it, a reasoning method in which precedents have a constraining effect at the very least must give the existence of the precedent a role in the reasoning process such that it influences its outcome. This might be the case because the conclusion about how to decide the present-case under consideration is different than it would have been if the precedent had not existed. It is also the case if the conclusion is the same but the reasons on which it has been reached are different because the precedent-case excludes reasons that would have otherwise counted and adds reasons that would otherwise have not existed. An adequate account of reasoning by precedent should be able to portray it as such a constraining method. I prefer this description of precedent-constraint to Schauer's formulation according to which a precedent is constraining only if it can make a judge change her mind about how to decide a present-case, Schauer, supra note 2, at 7, which Rigoni interprets as a psychological requirement, Rigoni, supra note 3, at 140. No method of reasoning can force anyone to change their minds: if I am set on believing that the moon is made of green cheese, then the best reasons cannot make me change my mind. Alternatively, one might interpret Schauer's requirement as demanding more than that the precedent-case must influence the decision-making process. He might be demanding that the precedent must have decisive influence, at least in some possible cases. But I think that is the case whenever the existence of the precedent-case changes the collection of reasons that support the decision: adding or subtracting a reason always has the potential to have decisive influence in some cases—those in which it changes the balance of reasons for and reasons against. (We cannot interpret having “decisive influence” as providing a reason that replaces all other reasons. In this case, Schauer's definition of “constraint” would makes nonsense of the practice of distinguishing. For more on this, see the discussion of Lamond's arguments in Section III.A.)

22. For an introduction to the account of argument-schemes that informs this paper, see, e.g., Douglas Walton, Argumentation Schemes for Presumptive Reasoning (1996).

23. For example, it might be argued that even if some other experts agree with S, this is by far not enough to justify accepting that x. Therefore, the third critical question in our example is too lax, and should be replaced by this one: “Do the majority of other trustworthy experts in area P agree with S?”

24. Following other theorists, I will treat rules as taking the form of a conditional. The antecedent of the conditional lists categories of properties a case needs to have for it to fall under the rule. The consequent states what must be done when the rule applies. Compare, e.g., Frederick Schauer, Playing by the Rules (1991), at 23ff.

25. This way of distinguishing between the two accounts results in my grouping the approaches of, e.g., Alexander & Sherwin, supra note 2, together with the approaches of Lamond, supra note 4; Horty and Bench-Capon (Horty, John, Rules and Reasons in the Theory of Precedent, 17 Legal Theory 1 (2011)CrossRefGoogle Scholar; Horty, John & Bench-Capon, Trevor J. M., A Factor-Based Definition of Precedent Constraint, 20 Artificial Intelligence & L. 181 (2012)CrossRefGoogle Scholar); and Rigoni, Adam, An Improved Factor Based Approach to Precedential Constraint, 23 Artificial Intelligence & L. 133 (2015)CrossRefGoogle Scholar. This happens because I count any approach as a rule-approach that works by first establishing rules (whether they are strict or defeasible) and then uses these rules to determine the decision in the precedent-case. It is important to point out that the authors of these approaches disagree in important ways. Alexander and Sherwin would likely reject the idea that Lamond, Horty, and Bench-Capon formulate rule-approaches, and Lamond calls his account a “reasons” account, a name taken up later by Horty and Bench-Capon. The difference is that Lamond, Horty, and Bench-Capon base their approaches on the idea that precedents authoritatively determine the relative strength of reasons (factors) based on which cases are decided, including the reasons (factors) that spoke against the precedent-decision but were not strong enough. However, as I will discuss below, the distinction between these two groups consists, in great part, of the kind of rule they employ. In Lamond, Horty, and Bench-Capon's accounts, the combination of factors that justified the precedent-decision form the antecedent of a defeasible rule, the precedent rule. Nonetheless, this distinction should not be underestimated. Alexander and Sherwin want strict rules, while Lamond and Horty claim that precedent only sets up highly defeasible rules and assigns relative weights to all the reasons of the precedent-case. Below, I will sketch how Alexander and Sherwin's strict rule-approach becomes first Horty and Lamond's defeasible rule-approach and then something even more flexible when amendments are made to deal with the two main objections commonly leveled against rule-approaches. The use of argument-schemes should make it easy to see how, despite important differences, all these accounts are related because they all include the extraction and then employment of a rule that mediates between two cases. This is structurally very different from the analogy-approach, according to which the two cases interact directly in the process of reasoning.

26. Alexander & Sherwin, supra note 2.

27. Id. at 11.

28. This is what Neil MacCormick calls a “legal syllogism,” though he would allow for the syllogism to be defeasible. Neil MacCormick, Rhetoric and the Rule of Law (2009), at ch. 4.

29. Weinreb, supra note 3, seems to be defending such an approach.

30. An even more simple representation of this scheme would have been:

However, this representation would have obscured the role that the doctrine of precedent plays in this reasoning. As the precedent itself does not contain an authoritative rule, the authoritative direction for action must come from somewhere else. It comes from the authoritative doctrine of precedent according to which legal sameness (not mere factual similarity) requires same legal treatment.

31. Alexander & Sherwin, supra note 2, at 68ff. Many analogy-approaches do not provide a viable answer to this. Scott Brewer claims that he is offering an analogy-approach but uses the justifiability of a derived rule to account for relevance. Scott Brewer, Exemplary Reasoning: Semantics, Pragmatics and the Rational Force of Legal Argument by Analogy (1996). Edward Levi, in his classic book An Introduction to Legal Reasoning, endorses an analogy-approach but his project does not require him to explain how it works any further than by saying: “The determination of similarity or difference is the function of each judge.” Levi, supra note 3, at 2. Lloyd Weinreb claims that the recognition of important similarity might feel like some kind of mystical “intuition,” but it is actually the employment of past experience. However, Weinreb does not explain why we should think that intuitions arise only from experiences with legal reasons instead of a judge's first-order moral and political reasons or, even worse, her biases and moods. Weinreb, supra note 3. Earlier in his career, Frederick Schauer explored the idea to use socially and otherwise established categories to explain how the overall relevance of the similarity between two cases can be determined. E.g., Schauer, supra note 14. However, he abandoned this idea and adopted a rule-approach because our language provides such a wealth in categories that it is possible to find a category for almost any two objects. Schauer, supra note 2.

32. See Lamond, supra note 4.

33. See id. at 9.

34. MacDonald v. Ortho Pharmaceutical Corp., 475 N.E.2d 65 (1985).

35. Restatement (Second) of Torts §388 cmt. n (1965).

36. Lamond, supra note 4, at 15–16.

37. Id. at 7–8. Lamond acknowledges that there is the option of adding exceptions and thereby effectively changing the rule. He dismisses this by pointing out that there does not seem to be any additional burden of justification for doing this—judges distinguish as easily as they follow. Changing an existing rule should incur a higher burden of justification. Of course, theoretically, we could deal with this issue by simply striving to list all the exceptions to the rule in the rule right away. However, this would make rule formulation forbiddingly hard, especially rule formulation based on just one case. Most likely, it would result in such long and complex rules that simply reading the rule would take as much time as making the decision without the rule. In addition, it is seriously questionable whether this would lead to any good rules, given how susceptible the human mind is to cognitive biases. (For a discussion of the trappings of rule-making based on cases see Schauer, Frederick, Do Cases Make Bad Law?, 73 U. Chi. L. Rev. 883, 893 (2006)Google Scholar). For a summary of the objections to trying to list all exceptions to a rule in the rule, see Rigoni, supra note 25, at 134–135.

38. Lamond, supra note 4, at 16.

39. Id. at 19.

40. For an overview of nonmonotonic logics, see, e.g., G. Aldo Antonelli, Non-monotonic Logic, in The Stanford Encyclopedia (2012), http://plato.stanford.edu/archives/win2012/entries/logic-nonmonotonic/. John L. Pollock is famous for having given the idea of defeasible reasoning a firm place in epistemology. John L. Pollock, Knowledge and Justification (1974). Douglas Walton brought the idea into argumentation theory and, on its basis, developed his theory of argument-schemes. Walton, supra note 22. Rigoni, supra note 25, provides arguments for choosing a nonmonotonic logic over a monotonic logic when formalizing reasoning by precedent.

41. Horty, supra note 25; Horty & Bench-Capon, supra note 25; Rigoni, supra note 25. Based on Lamond's ideas, Horty and Horty and Bench-Capon develop a method for classifying the aspects of a precedent-case into reasons for and against the precedent-decision and assigning them weights. The reasons (factors) for the decision make up the antecedents of the rules; the reasons (factors) against are those that will later be excluded. A complex ordering relation for reasons is added based on which precedent-constraint is explained. Rigoni, supra note 25, develops this account even further.

I believe that many rule theorists who are not as strict as Alexander and Sherwin will likely be willing to accept Lamond's, Horty's, Bench-Capon's, and Rigoni's account as an advancement and sophistication of their own accounts. Joseph Raz's account, for example, seems to be open to such a development. See, e.g., Joseph Raz, The Authority of Law (1979), at ch. 10.

42. Statutory rules, correctly understood, are probably defeasible too—after all, common law judges tend to re-interpret them when they would lead to absurd outcomes or when two statutes seem to conflict. But I think Lamond's point still stands; precedents are much more defeasible than statutes.

43. Some paragraphs of U.S. Restatements of the common law seem to contain something like such lists—see, e.g., Restatement (Second) of Torts §402A. cmt. n (1965).

44. Horty, supra note 25; Horty & Bench-Capon, supra note 25.

45. This is not meant to diminish the difficulty of interpreting statutes.

46. The difficulties in interpreting opinions have been discussed often and at length. Neil Duxbury, The Nature and Authority of Precedent (2008), at ch. 3, gives an overview over all the problems with interpreting opinions and describes in detail the discussion about the determination of the ratio decidendi that took place in the 1930s to 1960s. Today, theorists of both camps—for and against rule-approaches—take this problem seriously. Examples are Schauer, supra note 2, at ch. 9, and Rigoni, supra note 3. In Horty's, Bench-Capon's, and Rigoni's (2015) approaches, these difficulties are bracketed by the authors, who begin their theorizing only after the aspects of the case are categorized and determined as belonging to the rule's antecedent or the list of defeated reasons. Horty, supra note 25, at 5; Horty & Bench-Capon, supra note 25, at 186; Rigoni, supra note 25, at 146.

47. Or the judge might not have decided whether a reason she considered while deciding her case was necessary to justify her conclusion, blurring the lines between ratio and dicta. This difficulty is discussed at length in Greenawalt, supra note 4, at ch. 8.

48. See, e.g., Stone, Julius, The Ratio of the Ratio Decidendi, 22 Mod. L. Rev. 597 (1959)CrossRefGoogle Scholar.

49. I argue elsewhere that this is not necessarily a failing on the side of the precedent court. Katharina Stevens, Setting Precedents Without Making Norms? (forthcoming).

50. Dillon v. Legg, 441 P. 2d 912 (1968).

51. How difficult this would be was remarked upon, for example, by Judge Burke in the dissent to the same case.

52. Or even sets of well-defined factors that could be used in the kind of defeasible rule–based reasoning that Horty, supra note 25, describes.

53. See Rigoni, supra note 3, at 149.

54. Alexander & Sherwin, supra note 2, at 40ff.

55. Leiter, supra note 4. See also, e.g., Karl Llewelyn, The Bramble Bush: On Our Law and Its Study (1960), at chs. 2–4; Jerome Frank, Law and the Modern Mind (1949), at ch. 14.

56. See supra note 21 for a justification of this understanding of constraint.

57. See Shecaira, Fabio, Analogical Arguments in Ethics and Law: A Defence of Deductivism, 33 Informal Logic 406 (2013)CrossRefGoogle Scholar. It might also be possible to see something of Ronald Dworkin in this account (Dworkin, supra note 16, at ch. 7). It is important to note that this move will not only alienate strict rule theorists like Alexander and Sherwin. Choosing the rule that a precedent represents in this way will be seen as a blatant act of judicial discretion by most exclusive positivists, as for example Raz. See, e.g., Raz's defense of hard positivism in Raz, Joseph, Authority, Law and Morality, 68 The Monist 295 (1985)CrossRefGoogle Scholar. Therefore, authors who would have accepted the idea of integrating defeasible rules might not be on board anymore when first-order reasons are used to determine the content of the defeasible rule. This move, to them, might simply amount to the admittance that precedent is not constraining at all and that the American Legal Realists were right. I disagree, but I think it is important to realize how far the constraining effect of precedent gets reduced by this change.

58. The judge's ability to pick a rule might be further constrained by principles associated with precedent-interpretation. There are, for example, principles about what should count as dicta in an opinion, and what may be included in a precedent rule, or how narrow precedent rules should be. I will not discuss these principles here. I think that even with them in place, allowing judges to choose rules by mixing their own first-order reasons in with the reasons provided in the opinion means reducing the precedent-constraint considerably, depending on how vague the opinion is. And then there is, of course, the American Legal Realist argument that these principles really do not constrain anything because for every principle pointing in one direction, you can find one pointing in the other. See, e.g., Llewelyn, Karl, Remarks on the Theory of Appellate Decision and the Rules and Canons About How Statutes Are to Be Construed, 3 Vand. L. Rev. 395 (1950)Google Scholar.

59. Notice that if the opinion of precedent A provides enough information so that only one rule is consistent with it, then rule a is the best choice simply in virtue of being the only choice.

60. See supra note 2.

61. Richard A. Posner, The Problems of Jurisprudence (1993), at 86ff. See also Posner, Richard, Reasoning by Analogy, 91 Cornell L. Rev. 761 (2006)Google Scholar.

62. Alexander & Sherwin, supra note 2, at 68ff.

63. Proposed, e.g., in Holyoak, Keith J. & Thagard, Paul, Analogical Mapping by Constraint Satisfaction, 13 Cognitive Sci. 295 (1989)CrossRefGoogle Scholar. The by now classic, easily accessible introduction and summary of their theory is Keith J. Holyoak & Paul Thagard, Mental Leaps: Analogy in Creative Thought (1995). Dedre Genter has developed an only slightly different model—see, e.g., Genter, Dedre, Structure Mapping, A Theoretical Framework for Analogy, 7 Cognitive Sci. 155 (1983)CrossRefGoogle Scholar. The multi-constraint model and models like it are still the basis for research on analogy. See, e.g., an overview by Holyoak, Keith J., Analogy and Relational Reasoning, in The Oxford Handbook of Thinking and Reasoning 234–259 (Holyoak, Keith J. & Morrison, R. G. eds., 2012)CrossRefGoogle Scholar.

64. Holyoak & Thagard, supra note 63, at 296; Holyoak & Thagard, supra note 63, at 26.

65. Holyoak & Thagard, supra note 63, at 26/37f.

66. Id. at 29ff.

67. Id. at 25. See also Holyoak & Thagard, supra note 63, at 299ff.

68. Holyoak & Thagard, supra note 63, at 27ff.

69. Id. at 29.

70. Id. at 28f.

71. Id. at 29.

72. Hunter, supra note 3, at 1217ff./1222ff.

73. For an influential and by now classic account of relevance, see Dan Sperber & Deirdre Wilson, Relevance: Communication and Cognition (1986).

74. See Holyoak & Thagard, supra note 63, at 35ff.; Holyoak, Keith, The Pragmatics of Analogical Transfer, in The Psychology of Learning and Motivation 59–87 (Bower, G. H. ed., 1985), at 70ffGoogle Scholar; Holyoak & Thagard, supra note 63, at 297.

75. Holyoak & Thagard, supra note 63, at 30ff. See also Holyoak, supra note 63, at 249f.

76. Holyoak & Thagard, supra note 63, at 30.

77. See id. at 35ff.; Holyoak & Thagard, supra note 63, at 303ff.

78. Of course, this does not yet answer the question how well supported the inference must be for us to accept it. In analogical reasoning in general, this depends on the context in which we perform the mapping. Are we interested in the question whether we should allow kindergarten teachers to say goodbye to their wards once they have been fired? Or are we interested in the question whether kindergarten teachers should be able to sue for custody? Or are we interested in the question whether kindergarten teachers are likely to suffer intense emotional trauma when they witness the accidental death of their wards?

79. Hunter, supra note 3, at 1224ff. This is also what Frederick Schauer believes in Why Precedent in Law (and Elsewhere) Is Not Totally (or Even Substantially) About Analogy, supra note 2.

80. See Schauer, Why Precedent in Law (and Elsewhere) Is Not Totally (or Even Substantially) About Analogy, supra note 2.

81. The account of arguments by analogy I will now give is in harmony with (though not entirely the same as) the account provided by Juthe, Andre, Argument by Analogy, 19 Argumentation 1 (2005)CrossRefGoogle Scholar.

82. In other words, instead of taking the role of an arguer who tries to find an argument that will show that her conclusion is acceptable, the judge takes the role of an argument evaluator who tries to determine whether an argument succeeds at supporting a conclusion not chosen by her but given in the argument. I have argued for this interpretation of judges as interlocutors at length elsewhere. Stevens, Katharina, The Constraining Force of Analogies and the Role of the Judge, in Unpacking Normativity: Conceptual, Normative and Descriptive Issues (Himma, Kenneth Einar, Jovanovic, Miodrag & Spaic, Bojan eds., forthcoming 2018)Google Scholar.

83. It might now be possible to understand why authors like Posner, supra note 61, and Alexander & Sherwin, supra note 2, thought that analogical reasoning is nothing else than inductive or deductive reasoning. Once the mapping has been accomplished, it is possible to simply choose a category for each single mapping that contains both mapped aspects and make a rule. Once the rule has been formulated, both deductive and inductive arguments can be constructed. But to say that such a reconstruction is possible after the analogical mapping is not the same as saying that the analogical mapping never happened.

84. Once this conclusion is established, a further step combines it with the authoritative doctrine of precedent according to which legal sameness requires same treatment and leads to the conclusion that the present-case must be decided according to the precedent's decision. See supra note 30.

85. Dillon v. Legg, 441 P.2d 912 (1968).

86. This is also pointed out, for example, by Horty, supra note 63, at 28.

87. Take for example the famous case of MacPherson v. Buick Motor Co., 217 N.Y. 382, 111 N.E. 1050 (1916). The facts as described in the opinion written by Justice Cardozo included that the wheel of a car manufactured by Buick Motor Company broke because the wheel was defective, something that could have been discovered by a reasonable inspection. However, it has been pointed out that in fact there was good reason to believe that the wheel might have broken for other reasons. See Henderson, Jr., MacPherson v. Buick Motor Company: Simplifying the Facts While Reshaping the Law, in Torts Stories (Rabin, Robert L. & Sugarman, Stephen D. eds., 2003)Google Scholar. But even if it did, that does not change that the case answers a question involving the issue of who has a duty of care for defective products.

88. Levenbook puts great emphasis on this point. See Levenbook, Barbara Baum, The Meaning of a Precedent, 6 Legal Theory 185, 191 (2000)CrossRefGoogle Scholar.

89. This is, I think, one of the core ideas of Lamond, supra note 4.

90. I believe that talking of critical questions that are excluded from consideration in reasoning by precedent is another, possibly illustrative way of expressing Lamond's insight that later judges have to treat precedents as correctly decided: whether the precedent judge got the description or evaluation of the precedent-case right is simply not to be questioned because there are no usable objections to be had in that direction.

91. This is, I think, what Scott Brewer had in mind in Brewer, supra note 31.

92. Evidence that the attempt to come up with arguments against one's own initial position leads to debiasing and more balanced judgment can for example be found in Zenker, Frank, Know They Biases! Bringing Argumentative Virtue to the Classroom, in Ossa Conference Archive (2013), at 191Google Scholar.

93. See, e.g., Holyoak & Thagard, supra note 63, at 101. The claim that source and target interact in the mind of the reasoner, forming new understandings, has also been made about metaphors (which are closely related to analogies) by Max Black, Models and Metaphors (1962), at ch. 3.

94. For discussions of presence, see, e.g., Chaim Perelman & Lucy Olbrechts-Tyteca, The New Rhetoric (1969), at 116ff.; Chaim Perelman, The Realm of Rhetoric (1982), at 35ff.; Christopher Tindale, The Philosophy of Argument and Audience Reception (2015), ch. 10.

95. Lilian Bermejo-Luque formulates this point especially beautifully: “For example, in conveying that time is money, we convey an idea of time being as money in some relevant sense; a sense that, on the other hand, has not been made explicit. What we convey is not mere information, but rather a cognitive proposal of seeing time as money in a certain way, i.e., the way in which the comparison is relevant. As far as the addressee fully understands the analogy …, she is thinking of time as money …. This is why understanding an analogy is, in the end, falling, at least a bit, under its rhetorical spell.” Bermejo-Luque, Lilian, A Unitary Schema for Arguments by Analogy, 32 Informal Logic 1, 8 (2012)CrossRefGoogle Scholar.

96. Rigoni, supra note 3, at 153.

97. Dedre Genter & Arthur B. Markman, Structural Alignment in Comparison: No Difference Without Similarity, 5 Psychol. Sci. 152 (1994). See also Holyoak, supra note 63, at 246.

98. MacDonald v. Ortho Pharmaceutical Corp., 475 N.E.2d 65 (1985).

99. McEwen v. Ortho Pharm. Corp., 528 P.2d 522 (1974).