'The science of conjecture: Evidence and probability before Pascal'
ISR. Interdisciplinary Science Reviews; London; Summer 2002; Scott Campbell;

Volume: 27
Issue: 2
Start Page: 158-160
ISSN: 03080188
Subject Terms: Nonfiction
Probability
History
Personal Names: Franklin, James
Abstract:
"The science of conjecture: evidence and probability before Pascal" by James Franklin is reviewed.

Full Text:
Copyright Institute of Materials Summer 2002

'The science of conjecture: evidence and probability before Pascal', by James Franklin

2001, Baltimore, MD, Johns Hopkins University Press, 512 pp., $59.95/L41.50, ISBN 0 8018 6569 7

This is the intellectual book of the year, and it ought to become one of the great classics of intellectual history. In it James Franklin brilliantly describes the early development and application of the concept of probability, by which is meant not just the sort of probability associated with dice throwing (which he calls `factual probability'), but also what we often refer to as `likelihood" (and which is sometimes termed `logical probability'). That is, the book deals with the early history of non-demonstrative ('inductive') reasoning, where some degree of likelihood can be established from the evidence, but not certainty (in other words, not logical entailment). Such reasoning is often treated cursorily in logic classes, but it is the reasoning we most often have to employ in the sciences, medicine, the law and everyday life. For example, the `big bang' theory is supported by some good arguments, good enough to justify a belief in it, but we cannot be certain that it is true. The same applies to my belief that I will not die by going for a drive in my car today - it is entirely likely that I will survive, but I cannot be certain. As such, non-- demonstrative reasoning is one of the most important areas of study there is.

The general intellectual community knows little about the early history of this topic (and there is still much confusion even about the topic itself). One common view is that ex pressed in Ian Hacking's popular 1975 book, `The emergence of probability. Franklin's view is that this is a pretty good work (and I agree), but in it Hacking claimed that it was only at the end of the Renaissance that anyone began to realise that evidence could exist which objectively supported a conclusion, but did not entail it, and could do so independently of what some learned authority thought. Some historians have subsequently pointed out the mass of evidence against this thesis, but Franklin takes this a step further with a brilliantly comprehensive book length survey of the use of probability/likelihood in ancient, medieval, and Renaissance law, in Greek and Roman times, in Christianity, Judaism, and Islam, in astronomy, history, cognitive science and artificial intelligence, medicine, philosophy, theology, ethics, and in relation to business, contracts, insurance, and so on. I'm not aware of any other work of such high intellectual calibre which has this scope and range.

What makes a work of this type so difficult to put together is the sheer range and quality of knowledge required, and it is hard to think of anyone else who has such a masterful understanding of so many different fields. Franklin currently works as a mathematician at the University of New South Wales, but he is also a first rate philosopher, in my opinion one of the most underrated in the world, and he is incredibly knowledgeable about a huge range of difficult fields in the sciences, law, and the humanities. As someone once said of Clive James, `he's a brilliant bunch of guys'. (I could add that he's also an extremely hard working bunch of guys.)

In releasing this book, the result of many years' work, Franklin has jumped into the very top rank of historians of ideas. And I'm talking about real, serious, heavyweight historians of ideas, the kind who do real history and really know about their subject, as opposed to the `sociologist of science' types who are coming to dominate the history of ideas, who think that the only possible explanations for why historical figures believed or said anything are social ones, and who claim that it is best not to know about the subject you are writing about so that you can more clearly see the social forces at work without being deceived into thinking that there can be good objective reasons for and against the various prevailing viewpoints in the field.

In fact, Franklin's book is not only a great work of history, it is also, as he intended it to be, an antidote to this sort of thinking in general. 'Strong' sociologists of science and postmodernists, as much as one can understand them, all seem to deny the view that there can be objective non-demonstrative reasons for a conclusion, and they quite often use supposed historical studies to support such a view. Franklin says that a `historical survey of methods of evaluating evidence can play a small part in reason's fight back'. But I shouldn't give the impression that the book is about this modern fight - there are only a couple of references to it.

The book is also as well written as any I have seen of this type. In fact, after the first few chapters get things warmed up, it becomes a cracking read. No doubt a reader without the relevant background will wonder what words like 'contingent' mean, but the use of the occasional term of art is unavoidable in such a study. In general, the writing (in an area that often features writing verging on the incomprehensible) is marvellous: direct, clear, and simple but elegant. This is a book that could be - and should be - read with profit by anyone with a little acquaintance with any of the areas covered. As a non-mathematician, I even managed to follow one of the later chapters on mathematical probability; and even if one were to ignore the occasional bits of maths in that chapter, the historical story is still easily understood.

Franklin is quite brilliant at choosing quotes from the authors he studies. Anyone who has ever studied writers from before the seventeenth century will k now that the going can often be hard. But Franklin has cherry picked, with the greatest care, the clearest quotes from his authors. Moreover, because these quotes concern a historical topic that is not widely known, the quotes are not the usual ones. This is a book which throws fresh light upon a whole lot of ancient and medieval authors, and Franklin is concerned to show that much of what is commonly believed about these authors and these times is false. He has the ability - which David Stove also had - to subtly shift one's perceptions about a topic. Of course all sorts of writers of varying degrees of ridiculousness can try to change one's perceptions, and even many of the best philosophers have embarked on dubious ventures to colour the history of ideas with their own particular shade. The difference is (apart from the fact that he clearly isn't engaged on an ego trip) that Franklin provides convincing and well supported research. And it's so good to see - in age when it's de rigueur to follow Feyerabend's lead and insinuate that all our scientific heroes were nothing but conniving rogues and power mad, smooth talking charlatans that Franklin is able to convince the reader that common sense and reason are major factors in our intellectual heritage.

So reading this book will not only teach you about the history of probability, but you'll also pick up a load of other interesting facts about leading intellectual figures. You'll build up a more accurate view, for example, of Aristotle, Aquinas, Ptolemy (who, you may be surprised to know, explicitly advocated the principle of economy), Kepler, the witch finders, William of Ockham, Pascal, and more. There are inevitably some passages which leave you unsure of what is meant, but in general I found the comprehensibility of the historical quotes backed up by Franklin's exposition -- the best I've encountered. Reading Franklin is like reading the best pop science book you can think of, except that it's about history and probability, and it's real research of t he highest importance and intelligence, and the writing is literature of the highest standard, with a dignity, wit, and wisdom only rarely found in pop science.

One of the most interesting facets of the book is its discussion of the role the law played in the early history of probability. This is not a fact that has been widely appreciated, but it is hardly surprising when one considers that, more than anyone else in those times, lawyers and legal theorists were constantly having to evaluate evidence that did not lead to certainty, and they developed various methods of evaluation, and recognised various degrees of strength of evidence. (Another salient point here is that in earlier times, the law offered one of the best careers many intellectually capable people could have.) Franklin also gives due credit to the scholastics, who've borne the brunt of a lot of criticism over the years, much of it deserved; however Franklin argues strongly that the central and positive role they occupied in much intellectual progress is no longer sufficiently appreciated.

There is a lot of talk these days about interdisciplinary studies, and frankly a lot of what gets put about under this name is second rate. This book is a shining example of what a good interdisciplinary study should be - it couldn't be more interdisciplinary, and it couldn't be more brilliant. But it also illustrates one of the problems facing interdisciplinary researchers: good interdisciplinary study is exceptionally hard to do well. First of all, it takes a lot of time and effort to master more than one field. Moreover - and this is not so widely realised - in any case only really smart people are capable of mastering more than one field, and of putting them together in genuinely interesting ways. Most people struggle to master one area, so not surprisingly many of the people involved in interdisciplinary studies have this problem (this is truer in the humanities and social sciences than in the harder sciences, I should add). Franklin is one of those giant brains who come along every so often. Not only is he an incredible scholar, with an enormous breadth of knowledge, but he's really smart. (And he writes beautifully - to get all these things combined happens about as often as peace in the Middle East.) This work isn't just a work of scholarship, it isn't just a survey, it's a work of understanding and synthesis of the highest order, and - if it becomes more widely known, as it thoroughly deserves to - it will enrich and improve our thinking on a large number of subjects.

[Author note]
SCOTT CAMPBELL reviews

[Author note]
Scott Campbell (scott.campbell@knottingham.ac.uk) holds a joint appointment as a lecturer in the Department of Philosophy and the Institute for the Study of Genetics, Biorisks and Society at the University Nottingham. He was born in Tasmania, and has previously taught at the University of New South Wales, St Catherine's College Oxford, and University College London. His current research is focused on a defence of logical probability.



Reproduced with permission of the copyright owner. Further reproduction or distribution is prohibited without permission.