Why and How to Learn Why: Analysis-based Generalization of Procedures

Max Wertheimer, in his classic Productive Thinking, linked understanding to transfer: Understanding is important because it provides the ability to generalize the solution of one problem to apply to another. Recent work in human and machine learning has led to the development of a new class of generalization mechanism, called here analysis-based generalization, which can be used to provide a concrete account of the linkage Wertheimer suggested: these mechanisms all, in different ways, use understanding of examples in the generalization process. In this paper I review this class of mechanism, and describe a method for causal attribution that can produce the analyses of examples that the generalization methods require, in the domain of simple procedures in human-computer interaction. This causal analysis method is linked with analysis-based generalization to form EXPL, an implemented model which is a concrete, though limited, instontiation of Wertheimer s scheme. EXPL constructs an understanding of an example procedure and generalizes it on the basis of that understanding. Results of an empirical study suggest that some of EXPL's attribution heuristics are used by people, and that while a subclass of analysis-based methods, called superstitious methods, seem to provide a more plausible account of people's generalization under the conditions of the study than a contrasting class of rationalistic methods, at least some participants appear to use methods from both classes. The results also show that explanation-based methods, which rely on comprehensive domain theories, must be used in conjunction with a means for extending the domain theory. If thus enhanced, explanation-based methods are able to mimic the effects of other analysis-based methods, and can provide a good account of the data, though combinations of other methods must also be considered. Finally, I return to Wertheimer s ideas to argue that none of the current analysis-based generalization methods fully captures Wertheimer s notion of understanding. Proper choice among different possible analyses of an example is crucial for Wertheimer, but I argue that this problem may be beyond the reach of learning systems.     

Ethics of Immigration: The Issue of Convicted Criminals

In this paper, I explore and probe Joseph Carens’ remarks, in his recent book The Ethics of Immigration, on the immigration status of foreign convicted criminals who have served their sentence, and who wish either to immigrate into our country or who are already here. Carens rejects deportation when it is not called for by considerations of national security, and agrees that considerations of public order can justify barring convicted foreign criminals from entering the country. I broadly agree with his arguments against deportation: my remarks in this respect are clarificatory and exploratory as much as anything else. But (I argue) both his argument for open borders and his scepticism with respect to radical cosmopolitanism are in tension with his claim that past criminal convictions can act as a bar to entry.     

Reasoning From Uncertain Premises

Previous studies have shown that 1 participants are reluctant to accept a conclusion as certainly true when it is derived from a valid conditional argument that includes a doubtful premise, and 2 participants typically link the degree of uncertainty found in a given premise set to its conclusion. Two experiments were designed to further investigate these phenomena. Ninety adult participants in Experiment 1 were first asked to judge the validity of three conditional arguments Modus Ponens, Denial of the Antecedent, and Affirmation of the Consequent. They were then required to evaluate conclusion uncertainty as a function of two degrees of asserted uncertainty in the major conditional premise (If p then it is very probable that q and if p then it is not very probable that q) of the arguments from the first task that were otherwise unchanged. Results revealed an effect for asserted-uncertainty in two of the three argument forms. Marginal support was found for the hypothesis that perceived argument validity would be a predictor of performance. Experiment 2 investigated the way 40 adult participants combined two sources of asserted uncertainty, one in the major premise and another in the minor premise, when they had to score the uncertainty of the conclusion. The two most prominent kinds of responses were to choose the same likelihood as the weaker of the two expressed in the premises, or a lower one. However, the within-subject consistency was poor. Theoretical implications are discussed.     

Death,Betrayal, and a Guardian Angel

A familiar Epicurean argument for the conclusion that death (i.e., being dead) is not bad for those who die goes like this. The dead cannot experience anything, including being dead and its effects. But something is bad for an individual only if that person can experience it or its effects. Therefore, death is not bad for those who die. In this article, I consider several alleged counterexamples to this argument's second premise, along with some responses to them. The responses are not entirely without merit, as we will see. However, I contend that even if none of the cases cited are straightforward counterexamples to the Epicurean premise, they can be used to challenge it indirectly. I conclude that this familiar Epicurean argument is unsound.     

Between reasoning

In two experiments we investigated three-term reasoning with spatial relational assertions using the preposition between as compared to projective prepositions (such as to the left of). For each kind of assertion we distinguish the referent expression (i.e., the grammatical subject) from the relatum expression (i.e., the internal argument of the preposition; e.g., [The hedgehog]_{referent_expression} is to the left of [the frog]_{relatum_expression}; [the snake]_{referent_expression} is between [the donkey and the deer]_{relatum_expression}). Previous research has shown that integrating premises with projective prepositions is easier (a) when the relatum expression of the second premise denotes an element already given by the first premise (relatum = given), and (b) when the term denoting a given element precedes the term denoting a new element (given–new). Experiment 1 extended this finding to second premises with the preposition between. In Experiment 2, between figured in the first premise. In this case, participants built an initial preferred model already from the first premise, although such a premise is indeterminate with respect to the array that it describes. Since there is no need left for integrating the second premise, this premise is instead used to verify the initial model and to modify it when necessary. A further investigation of conclusion evaluation times showed that conclusions were evaluated faster when they first mentioned the element that was included most recently into the mental model of the premises. The use of premises with between permitted the separation of recency of model inclusion from recency of appearance of an element in a premise.     

Deduction from if-then personality signatures

Personality signatures are sets of if-then rules describing how a given person would feel or act in a specific situation. These rules can be used as the major premise of a deductive argument, but they are mostly processed for social cognition purposes; and this common usage is likely to leak into the way they are processed in a deductive reasoning context. It is hypothesised that agreement with a Modus Ponens argument featuring a personality signature as its major premise is affected by the reasoner's own propensity to display this personality signature. To test this prediction, Modus Ponens arguments were constructed from conditionally phrased items extracted from available personality scales. This allowed recording of (a) agreement with the conclusion of these arguments, and (b) the reasoner's propensity to display the personality signature, using as a proxy this reasoner's score on the personality scale without the items used in the argument. Three experiments (N = 256, N = 318, N = 298) applied this procedure to Fairness, Responsive Joy, and Self-Control. These experiments yielded very comparable effects, establishing that a reasoner's propensity to display a given personality signature determines this reasoner's agreement with the conclusion of a Modus Ponens argument featuring the personality signature.     

Variations in Mental Ability According to Month,Season, and Period of Birth

(Werner, Heinz, Ed. On Expressive Language. Worcester, Mass.: Clark Univ. Press, 1955. Pp. 81.). Reviewed by Michael Wertheimer.     

Der Rabe und der Bayesianist

The Raven and the Bayesian. As an essential benefit of their probabilistic account of confirmation, Bayesians state that it provides a twofold solution to the ravens paradox. It is supposed to show that (i) the paradox’s conclusion is tenable because a white shoe only negligibly confirms the hypothesis that all ravens are black, and (ii) the paradox’s first premise is false anyway because a black raven can speak against the hypothesis. I argue that both proposals are not only unable to solve the paradox, but also point to severe difficulties with Bayesianism. The former does not make the conclusion acceptable, and it entails the bizarre consequence that a great amount of non-black non-ravens substantially confirms the ravens hypothesis. The latter does not go far enough because there is a variant of the first premise which follows from Bayesianism and implies a weaker, but nevertheless untenable, variant of the conclusion.

---

Der Rabe und der Bayesianist

     

Rotten,Vile, and Depraved! Depictions of Criminality in Superhero Cartoons

The literature argues that media depictions of crime present messages that conform to and promote the dominant ideology about the causes of crime and the nature of criminality. Most research focuses on television news and adult programs, but little research examines messages about criminality present in children's shows. To fill this gap, a content analysis of children's cartoons was conducted, using episodes of Batman: The Animated Series, Spider-Man, and Justice League Unlimited. Three central themes emerged. First, much criminal activity centers on greed. Second, criminals are aware of right and wrong but pursue crime to meet their own self-interests. Third, criminals are different from law-abiding people. Even if they are flawed or evil, criminals are rational and therefore culpable actors. These shows provide messages about criminality consistent with the dominant ideology.     

Access,phenomenology and sorites

The non‐transitivity of the relation looks the same as has been used to argue that the relation has the same phenomenal character as is non‐transitive—a result that jeopardizes certain theories of consciousness. In this paper, I argue against this conclusion while granting the premise by dissociating lookings and phenomenology; an idea that some might find counter‐intuitive. However, such an intuition is left unsupported once phenomenology and cognitive access are distinguished from each other; a distinction that is conceptually and empirically grounded.     

Max Wertheimer's university career in Germany

Summary This paper presents archival and related material bearing on the German university career of Max Wertheimer. It is shown that many factors interacted to shape that career. What drew Wertheimer to psychological science was its potential philosophical relevance, as expounded by his mentors Christian von Ehrenfels and Carl Stumpf. Once his career had begun, however, he faced the challenge of maintaining intellectual independence in a system where patronage counted, and also the problem of producing systematic philosophical work as well as empirical research in psychology at appropriate points in his career. These difficulties, compounded by Wertheimer's highly personal style of thinking and writing, appear to have been at least as inhibiting to Wertheimer's career as was Anti-Semitism. Thanks to unusual circumstances created in part by Kurt Riezler at the University of Frankfurt, Wertheimer became full professor there in 1929 — only to be forced into exile in 1933, on obviously anti-Semitic grounds.Portions of this paper were originally presented in an invited commentary at the symposium documented in this issue. Grateful acknowledgements are due to Professor Viktor Sarris and the other symposium participants for their assistance. Research for the paper was funded in part by NSF grnat No. SES-8511230.     

Deontological evidentialism and <Emphasis Type="Italic">ought</Emphasis> implies <Emphasis Type="Italic">can</Emphasis>

Deontological evidentialism is the claim that S ought to form or maintain S’s beliefs in accordance with S’s evidence. A promising argument for this view turns on the premise that consideration c is a normative reason for S to form or maintain a belief that p only if c is evidence that p is true. In this paper, I discuss the surprising relation between a recently influential argument for this key premise and the principle that ought implies can. I argue that anyone who antecedently accepts or rejects this principle already has a reason to resist either this argument’s premises or its role in support of deontological evidentialism.     

Knowledge Grounded on Pure Reasoning

In this paper, I deal with epistemological issues that stem from the hypothesis that reasoning is not only a means of transmitting knowledge from premise‐beliefs to conclusion‐beliefs but also a primary source of knowledge in its own right. The idea is that one can gain new knowledge on the basis of suppositional reasoning. After making some preliminary distinctions, I argue that there are no good reasons to think that purported examples of knowledge grounded on pure reasoning are just examples of premise‐based inferences in disguise. Next, I establish what kinds of true propositions can to a first approximation be known on the basis of pure reasoning. Finally, I argue that beliefs that are competently formed on the basis of suppositional reasoning satisfy both externalist and internalist criteria of justification.     

Mental simulation and argument

When a quantified premise such as: Everyone loves anyone who loves someone, occurs with a premise such as: Anne loves Beth, it follows immediately that everyone loves Anne. It also follows that Carol loves Diane, where these two individuals are in the domain of discourse. According to the theory of mental models, this inference requires the quantified premise to be used again to update a model of specific individuals. The paper reports four experiments examining such iterative inferences. Experiment 1 confirmed that they are harder than immediate inferences. Experiment 2 extended the finding to negative inferences, i.e., granted that Anne does not love Beth, it follows from the quantified premise that Carol does not love Diane. Experiment 3 established that intermediate steps referring to specific individuals are accepted more readily than intermediate steps referring to quantified variables. Experiment 4 showed that the participants' written justifications corroborated the model theory.     

Max Wertheimer on seen motion: Theory and evidence

Summary Max Wertheimer, the chief founder of an experimentally based Gestalt psychology, conducted his pioneering studies in motion perception on new theoretical grounds. Since the influence of this approach may be greater in today's cognitive psychology than it has ever been during the half-century of introspectionism and radical behaviorism, it is appropriate to review the actual roots of Wertheimer's (1912) seminal publication and his continuing research on apparent and real motion perception in the light of past and recent work. Illustrative examples, especially of Wertheimer's early research, are provided in this paper. The implications of his experimentation and biopsychological theorizing are still of major interest for present psychological inquiry. Nevertheless, the need for more future systematic comparative research on motion perception must be emphasized. The Epilogue of this paper examines why important parts of Wertheimer's experimental contributions to psychology may have been underrated or neglected by many contemporary psychologists.Some of the issues dealt with here are treated at length in a three-part laudatio (Sarris, 1987a, b, 1988a). When this article was prepared for print, its original title, Max Wertheimer on Seen Motion: The Significance of Sound Experimentation for Present and Future Research, as presented at the Wertheimer Memorial Symposium, was changed to the present title. Under the present title, a brief version of this paper was presented at the APA Convention, New Orleans, Louisiana, USA, August 11–15, 1989.I am grateful for the comments and cooperative help provided by Carl F. Graumann, Michael W. von Grünau, Eckart Scheerer, and Michael Wertheimer.     

Kantian constructivism and the Reinhold–Sidgwick objection

In this paper, I give a reconstruction of the so‐called Reinhold–Sidgwick objection and show that Korsgaard‐style Kantian constructivists are committed to two key premises of the underlying argument. According to the Reinhold–Sidgwick objection, the Kantian conception of autonomy entails the absurd conclusion that no one is ever morally responsible for a morally wrong action. My reconstruction of the underlying argument reveals that the objection depends on a third premise, which says that freedom is a necessary condition for moral responsibility. After mapping the common replies to the objection, I demonstrate that none of these replies is available to Kantian constructivists. But they need not be committed to the absurd conclusion that no one is ever morally blameworthy. Kantian constructivists who want to resist the Reinhold–Sidgwick objection are well advised to subject the third premise of the underlying argument to critical scrutiny.     

The Concept of Acquiescence

Suppose a police car gives chase to some violent criminals, putting innocent bystanders at risk. The criminals have not threatened the police in any way; so we would not normally say that the police have been coerced into chasing. Nor are the police merely responding to natural circumstances, so they are not acting under necessity, in the usual sense. The case is different from one in which an ambulance speeds to hospital, putting innocent bystanders at risk, because the reason for the police speeding has to do with the unreasonable behaviour of others. Yet that unreasonable behaviour does not constitute threatening behaviour. How then should we describe what happened?     

Rational self-doubt and the failure of closure

Closure for justification is the claim that thinkers are justified in believing the logical consequences of their justified beliefs, at least when those consequences are competently deduced. Many have found this principle to be very plausible. Even more attractive is the special case of Closure known as Single-Premise Closure. In this paper, I present a challenge to Single-Premise Closure. The challenge is based on the phenomenon of rational self-doubt—it can be rational to be less than fully confident in one’s beliefs and patterns of reasoning. In rough outline, the argument is as follows: Consider a thinker who deduces a conclusion from a justified initial premise via an incredibly long sequence of simple competent deductions. Surely, such a thinker should suspect that he has made a mistake somewhere. And surely, given this, he should not believe the conclusion of the deduction even though he has a justified belief in the initial premise.     

When white collar criminals turn to fatal violence: The impact of narcissism and psychopathy

The present article takes as its point of departure Perri's view, according to whom the impact of personality traits such as psychopathy and narcissism has unduly been omitted in research on white collar criminality. However, this article argues that although such factors have an important role to play in some cases when white collar criminals turn violent in order to prevent the detection of their felonies, they constitute neither the necessary nor the sufficient conditions for white collar criminals to commit murders. This means on the one hand that, although potentially increasing the risk for violence, these traits are not by themselves sufficient as the triggers of violent acts and can function so only under certain social‐contextual circumstances. However, such personality traits are not by themselves necessary, as there can be empirical cases of fraud detection murder committed by white collar criminals lacking them. This position is supported by both a logical analysis of Perri's main arguments and by empirical evidence, consisting of a few real‐life cases of Scandinavian white collar criminals who have “turned red.”