Logic and Uncertainty in the Human Mind

A Tribute to David E. Over

Shira Elqayam, Igor Douven, Jonathan St B. T. Evans, Nicole Cruz

• Publisher: Routledge
• Year: 2020
• Print ISBN: 9781138084254
• eBook ISBN: 9781351620413
• Topic: Psychology
• Subtopic: Cognitive Psychology & Cognition
• Index: Psychology

David E. Over is a leading cognitive scientist and, with his firm grounding in philosophical logic, he also exerts a powerful influence on the psychology of reasoning. He is responsible for not only a large body of empirical work and accompanying theory, but for advancing a major shift in thinking about reasoning, commonly known as the 'new paradigm' in the psychology of human reasoning.

Over's signature mix of philosophical logic and experimental psychology has inspired generations of researchers, psychologists, and philosophers alike over more than a quarter of a century. The chapters in this volume, written by a leading group of contributors including a number who helped shape the psychology of reasoning as we know it today, each take their starting point from the key themes of Over's ground-breaking work. The essays in this collection explore a wide range of central topics—such as rationality, bias, dual processes, and dual systems—as well as contemporary psychological and philosophical theories of conditionals. It concludes with an engaging new chapter, authored by David E. Over himself, which details and analyses the new paradigm psychology of reasoning.

This book is therefore important reading for scholars, researchers, and advanced students in psychology, philosophy, and the cognitive sciences, including those who are not familiar with Over's thought already.

1

The contribution of David E. Over

An intellectual biography

Ken Manktelow and Jonathan St. B. T. Evans

David Earl Over is a true cognitive scientist. With a firm grounding in one discipline, philosophical logic, he has reached way beyond that area to exert a powerful influence on others, most notably the psychology of reasoning. He is now one of the leading figures in this area, responsible for not only a large body of empirical work and accompanying theory, but for advancing a major shift in thinking about reasoning, commonly known as the ‘new paradigm’ psychology of reasoning – a term he coined.

David’s roots in philosophical logic go very deep. Along with the philosophy of mind, it was his primary interest in his first degree, in Philosophy, from Gettysburg College in Pennsylvania USA, not far from his birthplace in Maryland. In fact, the roots go deeper still, to his high school years, when he started reading outside what he thought was a boring curriculum. He failed even then to make sense of the claim, made by some philosophers, that the material conditional could represent the meaning of the natural language conditional; a high school failure he reports never being able to overcome. Following his graduation, in 1968 he set out for the UK, and took a one-year MA in Philosophy at Keele University – he was interested in British philosophical logic and philosophy of mind. He then began a PhD in Philosophy at Birkbeck College, University of London; his research topic became what, according to David, is best described by the title of a classic paper by J. L. Austin (1961): ‘Ifs and cans’. David quotes its first passage as an unimprovable statement of the problem:

Are cans constitutionally iffy? Whenever, that is, we say that we can do something, or could do something, or could have done something, is there an if in the offing – suppressed, it may be, but due nevertheless to appear when we set out our sentence in full or when we give an explanation of its meaning?

(Austin, 1961, p. 153)

David gives an example to illustrate this question, and point up the fascination he felt about the things that can be done with if. Suppose, he says, that by getting the MA at Keele, he could teach part-time in Britain and earn some extra money while doing his PhD. Does the use of ‘could’ here mean that if he had tried to do part-time teaching in Britain he would have succeeded? A British degree, at that time, gave one legal permission to be a teacher in Britain. But suppose he did not want to teach while doing his research, because he was so focused on his PhD; and that the ‘because’ here refers to a chain of causal factors, such as pressure to finish the PhD on time, leading to a decision not to do this teaching. Does the statement that he could still have done it mean that if he had wanted to do it, or had decided to do it, then he would have succeeded in doing it?

These questions are closely related to the traditional philosophical problem of free will and determinism. But they are also essentially connected with questions about modality, and the analysis of indicative and counterfactual conditionals. Much of the work that so influences us today appeared in the late 1960s and 1970s, by Adams, Stalnaker, and Lewis – just around the time of David’s MA and PhD work, and he was much concerned with it. He still is. David’s PhD supervisor was Dorothy Edgington, and he describes himself as having been extremely lucky in this. She gave him help on all the relevant topics, as did Hans Kamp, then a new lecturer at University College London (UCL). Kamp had received his PhD at UCLA, supervised by Montague, in 1968, and so was able to bring David up to speed on the latest research on modal logic, intensional logic in general, and Montague grammar.

The University of London was a superb place for David’s studies. He spent most of his time at Birkbeck and UCL, but would also take the short walk to the London School of Economics (LSE) from time to time, to take in John Bell’s lectures on mathematical logic, and talks and seminars in the Philosophy Department. Karl Popper had retired by then, but Imre Lakatos was a lively and stimulating presence. There was also much research on probability ‘in the air’ in the University of London at the time; Edgington was very interested in Ramsey, there was Dennis Lindley at UCL, and debates about probability at the LSE. These must have included Colin Howson, later to become a major figure in Bayesian reasoning; David does not recall speaking directly with him at this time, but has benefited from discussions with him, and other LSE people, since.

At this point, David had not yet been ‘contaminated’, as Shira Elqayam once put it, by experimental psychology. He was only vaguely aware of the research on the psychology of reasoning then in full swing at UCL, and did not meet Jonathan Evans, a contemporaneous PhD student himself. He did, though, have one brief meeting with Peter Wason. However, his interests in philosophical logic and the philosophy of mind were preparing him to be receptive to the psychology of reasoning – and for the advent of a cognitive science approach to the topic. There is a prescient passage at the end of Austin’s ‘Ifs and cans’ that appears to prefigure this development, and perhaps it became a kind of subliminal stimulus to David’s thinking:

Is it not possible that the next century may see the birth, through the joint labours of philosophers, grammarians, and numerous other students of language, of a true and comprehensive science of language?

(Austin, 1961, p. 180, original emphasis.)

Austin does not mention psychology here, but David was certainly looking for such a new subject, dissatisfied as he was with the narrowness, as he saw it, of mainstream British philosophy during the time of his doctoral work. He was therefore primed for psychological contamination.

This duly came about in the early 1980s, when he was working at what was then Sunderland Polytechnic (to become the University of Sunderland) as a lecturer in philosophy. In 1980, Ken Manktelow joined the staff there, in another department in another building, as a lecturer in psychology. Ken’s PhD, awarded less than a year earlier and examined by Wason, was in the psychology of reasoning. The psychology staff were housed in a Victorian building some way off the main campus; David’s office was in a 1970s block central to the campus, the Forster Building, which also housed a lot of the teaching rooms and, crucially for our story, a common room. Historical note for younger readers: common rooms. These have largely disappeared from the newer British university campuses, apparently on the grounds that, since you cannot quantify any work that goes on in them, no work is therefore going on, and so they are a waste of money. That this is a demonstrable falsehood is attested by this book and this story.

So here we are in the Forster Building; there are psychology and philosophy classes taking place in this room or that, and, in between, refreshments are had in the Forster Common Room. Some time in, probably, 1981, Manktelow and Over happen to be there at the same time, and are introduced. Neither of them can remember exactly when this happened, or who the introduction agent was, but like other polytechnics in the UK at that time, Sunderland’s ‘mission’ and open departmental structure encouraged interdisciplinary courses, and Ken and David taught on one of these. Someone must have realized that they might have interests in common, and so the match was made.

Then that powerful but indefinable accident happened: they hit it off as people. Not only did their academic interests complement each other, but they found lots to talk about which had nothing much to do with work, each could readily make the other laugh, and pretty soon they were meeting as friends, not just as work colleagues. They both had baby daughters at the time too, another piece of common ground. They decided that their complementarity of academic interests could be put to good use, and devised a final-year course on language and thought, which uniquely had a joint Ph/Ps – philosophy and psychology – module code, meaning that it could be taken by students calling either subject home. This joint development would have been much more difficult at that time, if not impossible, at a traditional British university. There was a nice intersection, and no clear boundary, between the two focal areas, language and thought, and the field was wide open for a philosophy–psychology intersection as well. Ken and David realized that the whole thing could be wrapped up in a parcel labelled inference; they could examine explicit inference, the kind that goes on in reasoning, and implicit inference, the kind that makes natural language work, and they could thrill students with the different ways in which these fundamental aspects of human mentality could be examined. They were perhaps a little optimistic in their expectations about how these wonders would be received by students, especially when open arguments took place (they took some of the classes jointly). The students may have found it difficult to take notes during these episodes.

Helped again by circumstance – the rapid growth in the psychology of reasoning and the philosophical study of conditionals and linguistic inference taking place in the 1980s – this course was never delivered the same way twice, and after a couple of runs, it dawned on David and Ken that, since there was no core text that they could recommend to their students, they should write one. By this time, they had begun to develop their academic relationship into research ideas as well as teaching, and in 1987 published their first paper together, called ‘Reasoning and rationality’. Rationality was a topic that emerged naturally from the concerns of their joint course, and the case for a review of it was enhanced by the fact that there was at that point very little work directly on it in the psychology of reasoning; it was, and is, an age-old philosophical topic, of course. A cross-disciplinary perspective was, they felt, timely.

The writing of this paper was given an additional and powerful stimulus: some time in 1986, David appeared with a copy of Leda Cosmides’ PhD thesis (Cosmides, 1985), which had been delivered the year before (this was an early example of the payoffs from his talent for networking). Cosmides’ thesis, the bulk of which was published in a gigantic paper in Cognition – it took up an entire issue – three years later, introduced her social contract theory (Cosmides, 1989). This was the idea that there was a basic ‘Darwinian algorithm’, as she termed it, consisting of a belief expressed in conditional form, that If you take a benefit then you pay a cost, together with a cheater-detection device, whereby people are sensitized to the possible presence of those who take the benefit without paying the cost. Arrestingly, for David and Ken, the test-bed for her evolutionarily inspired theory was the Wason selection task. This paradigm, having been a main focus of Ken’s own thesis, figured prominently in their classes on the Language and Thought course. Cosmides provided extensive investigations of the previously well-documented ‘thematic facilitation effect’, the idea that the selection task was more likely to be solved (logically) when couched in realistic terms, something that Ken’s own work had already called into question. So for him, the appearance of this thesis was a ‘this changes everything’ moment.

David, however, was notably more sceptical, and indeed voiced criticism of Cosmides’ work from the off. This is where his philosophical knowledge, in this case of deontic reasoning, and its relation to deontic logic and decision making, really began to pay dividends when applied to research in psychology. He was able to point out the logical flaws and missing distinctions in the arguments of Cosmides, and in the earlier work of Cheng and Holyoak (1985) on pragmatic reasoning schemas. The 1987 Manktelow and Over paper put their account of Cosmides’ work (its first citation in the literature) together with a critique of mental logic and mental models theory and an emphasis on the importance of dual processes, what we now more usually call dual systems, referring to Evans’ (1984) heuristic-analytic theory. These threads, as you will see elsewhere in this book, have formed continuous strands in David’s psychological thinking to the present day; they were there from the beginning.

David and Ken went on to design their own research on deontic reasoning, rejecting the schema theories of Cosmides and Cheng & Holyoak, but retaining the idea, inherent in them, that central to people’s thinking about rules of obligation and permission was their assessment of expected utility: the probable benefits and costs associated with adherence to and transgression of such rules. In other words, that deontic reasoning could be conceived of as a form of decision making. This was their first foray into ‘new paradigm’ experiments on reasoning, although it was not thought of or referred to in those terms at the time (see Manktelow & Over, 1991). The new paradigm, as it now is, has probability at its heart, and a wholesale recasting of the selection task in Bayesian probabilistic terms appeared in a major statement by Oaksford and Chater (1994). Decision making is also inherently probabilistic, of course, and the combination of probability and utility, i.e. expected utility, in deontic forms of the Wason selection task was demonstrated in the work of Ed Sutherland, a PhD student supervised by Ken and David (Manktelow et al., 1995). Oaksford and Chater found that they had to combine probability with utility to account for performance on the deontic selection task, and Ken and David went on to further development of their decision-theoretic view of deontic conditionals, with the help of Dinos Hadjichristidis (Over et al., 2004).

All these threads in David’s work appeared in their first book, called Inference and Understanding (Manktelow & Over, 1990). This was the text mentioned above as filling a hole required by the Language and Thought course they were teaching. Another foundational element of David’s version of the new paradigm psychology of reasoning makes its debut here as well: the Ramsey test. This appears in a section on conditional logic and addresses how we may decide the ‘truth’ of a subjunctive, or counterfactual, conditional, such as If Oswald had not killed Kennedy then someone else would have. As you will see later in this chapter and in this book, the Ramsey test, and its refinements, have occupied a central role in David’s later work on reasoning.

While the book was being written, a very important interpersonal thing happened. The First International Conference on Thinking was convened at Aberdeen University in 1988. Ken persuaded David to accompany him – it would be David’s first exposure to a psychology conference. He was not impressed. He was exasperated by the 20 minutes talk/5 minutes discussion format; apparently, philosophy conferences allot far more time to discussion. So exasperated, in fact, that he left early. However, the conference had been jointly convened by Jonathan Evans, Ken’s former PhD supervisor, so he took the opportunity to introduce David to Jonathan. And the rest really is history, as you can see amply portrayed throughout this book.

Four years later, in 1992, Manktelow and Over were invited,...

Table of contents

• Cover
• Praise Page
• Half Title
• Title Page
• Copyright Page
• Table of Contents
• List of Contributors
• 1 The contribution of David E. Over: An intellectual biography
• 2 Satisficing, meta-reasoning, and the rationality of further deliberation
• 3 Deduction from uncertain premises?
• 4 Compounds of conditionals, uncertainty, and indeterminacy
• 5 The suppositional conditional is not (just) the probability conditional
• 6 Probabilistic entailment and iterated conditionals
• 7 Two systems for thinking with a community: Outsourcing versus collaboration
• 8 Integrating Causal Bayes Nets and inferentialism in conditional inference
• 9 Objecting to uncertain conditional sentences: A cross-linguistic study
• 10 Children’s comprehension of conditional requests
• 11 Delusional rationality
• 12 Relevance and conditionals: A synopsis of open pragmatic and semantic issues
• 13 Correlation detection with and without the theories of conditionals: A model update of Hattori and Oaksford (2007)
• 14 Working memory, autonomy, and dual process theories: A roadmap
• 15 The development of the new paradigm in the psychology of reasoning
• Index