1. My aim today is to present a sketch of Jeremy Bentham’s views on punishment, its sources in his writings, and some of its central features. As you will soon see, what I have to say can perhaps be best described as work in progress.
2. What would naturally occur to anyone interested in Bentham’s views on punishment would be to study the book titled Rationale of Punishment, attributed to Jeremy Bentham on the title page as sole author, published in 1830 in London and in English. 2 What more could we want for an account of someone’s theory of punishment than his being the author of a book titled Rationale of Punishment?
3. Very well, let us take a look at this book. Although it appears under Bentham’s name and has been used by countless scholars as Bentham’s work, it was in fact stitched together from various of Bentham’s manuscripts by one of Bentham’s young acolytes, Richard Smith. During the day Smith was a minor official in the government’s bureau of taxation, but in his spare time he was a devoted editor of several of Bentham’s otherwise unpublished writings. Much of the content of the Rationale can be found virtually verbatim in the surviving manuscripts written in various hands and stored in the Bentham archive of University College London. Smith used this material to flesh out the Rationale from the earlier French version of 1811, edited by Etienne Dumont, a diligent and faithful friend of Bentham’s. There is no evidence that Bentham ever looked at the product of Smith’s labors (he did look at Dumont’s), much less that he guided it through the press or otherwise was involved in editing it.
4. For these reasons we might advisedly call this book the Smith–Bentham Rationale. Ideally for the purposes of posterity Smith would have turned over to the archives the complete manuscript copy of the text used by the printer to produce the Rationale. But so far as we know, he did not do this; none of that text is known to survive in any form—or if it is to be found among the surviving manuscripts, there seems to be no record or indication that some of the surviving manuscripts were in fact used by the printer. In any case we have it on good authority that Bentham’s practice was to discard the manuscript text of a book of his once the book had been published. The result is that the surviving manuscripts in the UCL archive have a very uncertain status so far as representing Bentham’s best or final thoughts about punishment (and about many other topics as well).
5. If at this very late date we want to construct a theory of punishment on Bentham’s behalf, what resources do we have for the job? What obstacles do we face in carrying out that task? First, let us note the several texts of varying importance at our disposal for this purpose:
First, there are the published materials:
First on this list is Etienne Dumont’s Theorie des Peines et Recompenses, a two-volume work of 1811, published in Geneva in French, and—as we see on the title page—purporting to be Bentham’s book. The first volume of this work is the source of most of what was later translated and published in English by Smith as the Rationale—as a comparison of the tables of contents of the two volumes will reveal.3
Second, there is Rationale of Punishment, by Richard Smith published in 1830 and reprinted in 1843 in the Works of Jeremy Bentham edited by John Bowring. It is this text in Bowring’s edition that is the most widely used version of the Bentham–Smith book.
Third, there are the Panopticon ‘letters’ of 1787, republished in the Works edited by Bowring.4
Fourth, there is ‘Jeremy Bentham to his fellow-citizens of France, On Death-Punishment’ (1831), a short essay published by Bowring in Bentham’s Works as an appendix to the Smith–Bentham Rationale.5
Fifth, there are large sections of An Introduction to the Principles of Morals and Legislation (IPML) originally written by Bentham prior to 1780 and published in that year and again in 1789 in a revised edition. Much of the most theoretical aspects of Bentham’s views on punishment appear only in this volume.6
Then there are the unpublished materials, among them the following:
‘Law versus arbitrary power, or a hatchet for Dr Paley’s net’, a mss written in 1808–1810; a transcription of this text is on file in the offices of the Bentham Project, but so far remains unpublished.
‘JB to France against the death penalty, on death for the French ex ministers’ (1830), an unedited and unpublished short epistle.
There are also passages in other of his published writings, some available only since his death, e.g. A Comment on the Commentaries, written in the mid 1790s but not published until 1928, and Of Laws in General, written in 1782 but not published until 1970.7
Last but by no means least, there are a thousand or so manuscript folios in the Bentham archives in UCL spread across his career but mostly written in two stages: the earliest surviving mss are from the mid 1770s, the latest from the early 1810s. Many ,but not all, of the folios are in Bentham’s own hand. Only now—literally, as of this week—have they all been transcribed, though much still needs to be done before these transcriptions are complete and accurate. These manuscripts have rarely been consulted except by a few diligent scholars; it is clear from Bentham’s correspondence that they were intended to be the chief source for a book he was writing in the mid 1770s under the title Theory of Punishment. It is the reconstruction of this book that Bentham never finished that is my principal self-appointed occupation.
Anything that claims to be Bentham’s theory of punishment must address and incorporate, where appropriate, all these materials. It is a major task awaiting someone energetic and patient enough to undertake it. Today I will discuss only a small fraction of this large topic.
6. As I noted a moment ago, it is clear from Bentham’s correspondence during the 1770s that he was hard at work on a book he provisionally titled Theory of Punishment—we learn this from a footnote in his Introduction to the Principles of Morals and Legislation of 1789, as well as from his correspondence and other writings of those years.8 This book was never completed as such. Instead, in 1789 IPML was published, a large fraction of which is devoted to the theory of punishment. Four decades later the Smith–Bentham Rationale of Punishment was published, bearing an uncertain (and so far unexplored) relation to Bentham’s intentions in the 1770s.
7. Let us leave aside further discussion of the sources we have from which to reconstruct Bentham’s theory of punishment and look at what we can learn about his views from consulting the tables of contents (ToCs) for his books mentioned above.
The first version of a ToC for Bentham’s ‘Theory’, date unknown, is conjectured to have been written between 1780 and 1785 on two mss sheets (not in Bentham’s hand) in a folio in UCL Box 99 and titled ‘Theory of Punishment’ (Appendix 1). A. T. Milne’s bibliography unfortunately fails to identify the relevant folios for this topic, thus effectively burying the topic until uncovered by some recent grave digging.9 This ToC precedes the Dumont and Smith books by several years, and the topics it proposes are fewer than those in the later French and English versions. This ToC is provisionally dated several years after Bentham had put work on his ‘Theory’ to the side in favor of other projects. There seems to be no evidence that Bentham had in mind a different, earlier ToC contemporaneous with his work on the book in the mid 1770s.
In any case, the title heading this ToC is presumably for a book to be called ‘Theory of Punishment’ The ToC itself is described in a heading on the folio as the ‘Table of Contents of the Work’. This also suggests that at some point Bentham intended this to be the ToC for his book, the ‘Theory of Punishment’, that preoccupied him during the 1770s.
Whether this ToC had any influence on the much later editions of Dumont (1811) and Smith (1839) is hard to tell. Many of the topics in this ToC are identical with topics in the ToCs of their books. The sequence in which the topics are discussed in these books closely resembles the sequence of topics in the Dumont and Smith versions. Three major topics—punishment by way of torture, the death penalty, and the Panopticon—that figure in the Dumont and Smith books, have no place in this early ToC. Thus it is obvious from these ToCs that the book on punishment that Bentham never finished departs in many ways from the versions published by Dumont and by Smith.
Conspicuous by its absence in this ToC of 1780–85 is any material for an Introduction (five out of the six topics identified in the ToC as the Introduction to Theory are struck out by someone); they turn out to be related to material that now appears as the first chapter of IPML. It is more than possible that Bentham took material from his intended ‘Theory of Punishment’ and used it instead to open IPML, causing him to strike out these passages from the ToC under discussion. A major challenge to the reconstruction of Bentham’s ‘Theory’ is to decide what material, if any, is available in the archives that can serve as the basis for constructing an introduction for this book. Whether either Dumont or Smith knew of this ToC and if so, why he/they chose to depart from it, is unknown.
How far in the reconstruction of Bentham’s ‘Theory of Punishment’ should we use this ToC as our guide? I return to this question shortly.
Dumont reports that the mss on which he relied for his version of Bentham’s theory ‘were written [by Bentham] in 1775’. He added that he ‘freely used the rights of an Editor—according to the nature of the text and the occasion, I have translated, commented, abridged, or supplied’. To which he added, ‘this co-operation on my part has had reference to the details only, and ought not to diminish the confidence of the readers; it is not my work that I present to them, it is, as faithfully as the nature of things will permit, the work of Mr Bentham’.
The resulting book by Dumont is clearly the earliest published text in any language purporting to be something that could be called without much controversy the chief printed source for Bentham’s theory of punishment. The ToC for Dumont’s version seems not to have enjoyed any supervision or revision by Bentham. Worse than that, in his correspondence Bentham describes the Theorie as Dumont’s book, ‘not mine’, thus distancing himself from the finished product, and leaving the reader unsure just how to take this authorial disavowal.
Smith’s 1830 English edition is heavily based on Dumont’s 1811 version. The English title, however—Rationale of Punishment—is not Dumont’s; presumably it is Smith’s, perhaps chosen to distinguish his book from Bentham’s unfinished and unpublished ‘Theory of Punishment’ of the mid 1770s. There seem to be no surviving mss in which Bentham or one of his secretaries uses Smith’s title to describe any of the material in mss form.
As noted earlier, a sample of topic-by-topic comparison of the Smith Bentham Rationale with the Dumont–Bentham Theorie reveals considerable divergence between the two. Smith correctly notes in the ‘Advertisement’ at the front of the Rationale that his version is not ‘a literal translation’ of Dumont’s French version, nor is it an assemblage exclusively of Bentham’s own manuscripts on the theory of punishment. Instead, Smith tells us that he ‘availed himself, wherever he could, of the original manuscripts’—manuscripts presumably mostly if not entirely in English.10 As for ‘wherever he could’ not ‘avail himself of the original manuscripts’, he contributed a version of what he thought Bentham would have said, or at least would not have disapproved.
How much material in the Smith–Bentham Rationale is in fact neither Bentham nor Dumont, but contributed by Smith (much as were Dumont’s additions) in the belief that Bentham would have approved, is difficult to say. A reconstruction of Bentham’s own views from surviving mss in his own hand (or that of a faithful amanuensis) depends crucially on being able to make these distinctions. Smith did not include in his legacy any indications of the extent and character of his own contributions.
How closely does the Smith–Bentham Rationale (as reprinted in 1843 by Bowring in the first volume of Bentham’s Works) track the earlier 1830 version by Smith? Apparently it has always been tacitly assumed that the two versions are identical, with differences if any confined to matters of format, pagination, and the like. But so far as I know no one has left any record of a word-for-word comparison of the two English texts (or, while we are at it, of either English text with the French). Such discrepancies between the two that I have noticed in a very small sample of passages are relatively minor. On the strength of such comparisons, one might plausibly assume that the discrepancies in the two texts are correspondingly insignificant. Short of conducting a word-for-word comparison of the two English versions, however, we cannot be sure how close the two texts really are. For my reconstructive purposes, it doesn’t matter how this question is resolved.
In addition to the material already reviewed, rummaging around in obscure papers deposited with the Bentham Project four or five years ago turned up a detailed ToC prepared by J. M. Phillips while he was on the Project staff some forty years ago (probably in the late 1960s). He describes his ToC as ‘[a] catalogue of Punishment mss arranged [in] the order of possible publication, excluding the “Introduction”‘. Little is known today about Phillips and his association with the Project. Yet as his ToC and adjacent typescripts in his papers demonstrate, he had by far the most complete understanding of Bentham’s intentions and of Bentham’s surviving manuscripts devoted to the theory of punishment.
In addition to providing us with a ToC for Bentham’s unfinished book, Phillips accomplished another major task. He identified a hundred (102, to be exact) topics that constituted the book and coordinated with them the relevant manuscript folios from the archives. The result is that for the first time we have a clear picture of what constituted the text of Bentham’s ‘Theory’.
On what principles did Phillips construct the ToC for this book? He divided the surviving mss into four sets, those determined by author and those determined by date. If Smith was the author (whether or not based on Dumont) of a portion of text and not Bentham nor copied for him by an amanuensis, Phillips proposed to exclude the text. (More precise criteria for distinguishing between the two sets of mss are nowhere stated in Phillips’s papers that survive in the Project offices. Clearly, most reliance was placed on deciphering the handwriting in a folio in order to assign the text to Bentham, or to Smith, or to some third writer.) In any case, since the Bentham Project has no interest in publishing material written by anyone other than Bentham, identifying material by Dumont and Smith in order to exclude it is crucial.
Orthogonal to the division by author is a division by date. Anything on punishment not from the 1770s, except for a few folios dated around 1811, Phillips proposed to exclude, on the ground that such manuscripts have at best a dubious relevance to Bentham’s intentions of the mid.1770s.
The folios that remain after these exclusions—call them the Bentham set— number about 575, out of a total of nearly a thousand in the entire UCL archive that are mentioned by A. T. Milne as dealing with punishment.
Phillips’ legacy, valuable as it is, still leaves much to be done. No transcripts by Phillips of any Bentham mss on punishment have been found. It is clear that Phillips thought that his ToC, if faithfully followed, would yield a book as close as humanly possible at this late date to the book on punishment that Bentham in his early manhood wrote but never completed, much less published.
In any case, even though Smith’s ToC departs notably from the other ToCs cited above, in my judgment it convincingly constitutes the best version of the book on punishment Bentham never completed.
Nothing can be seriously considered to be Bentham’s theory of punishment unless it takes into account all these materials—a considerable undertaking. Given the magnitude and diversity of the relevant materials at our disposal, here I can do no more than sketch some of the main and enduring features of his views, as well as trace out a few of the important topics on which his views changed.
8. Let us notice first two very different statements from the mid.1770s by Bentham that sketch the main features of his project. The first and briefer seems never to have been developed further. The second and longer (and possibly a draft of the Introduction, Chapter VI, to his ‘Theory’; see appendix 2) gives a much better sketch of the nature and status of the theory of punishment and hints at the central role of two characteristic features of most of his thinking: his consequentialism and his hedonism, which together constitute his utilitarianism.
9. Next, let us note a matter of style, the form in which Bentham chose to cast his views. To a remarkable extent his typical discussion of a topic proceeds by offering the reader one taxonomy after another, that is, lists of species under a given genus—a method of analysis of which he was inordinately proud but is nothing more than a familiar tactic in classical rhetoric. By way of illustration, consider what we find in IPML chapter xvi, §§ 32.35, on ‘The Division of Offences’. Here Bentham offers the reader the identification and discussion of five kinds of distress one person can cause another, nine ways in which a person can be caused personal injury, two kinds of offenses against reputation, and nineteen kinds of offenses against property. This is the barest sample of Bentham’s love affair with the taxonomy of crimes and punishments. He called it his method of ‘bipartition’ or ‘bifurcation’.11 Actually, as the example above show, division of a genus into two species—suggested by his terms ‘bifurcation’ and ‘bipartition’—was a rarity. Three or often more species under a genus typically emerge from Bentham’s scrutiny of topic after topic.
10. Why does Bentham devote so much attention to distinguishing species under a genus? I think it is because he is convinced that one cannot evaluate a given practice or proposal without considering alternative practices and proposals, in the attempt at making one’s survey of options as complete as possible. Identifying these alternatives is a crucial feature of utilitarian thinking, since the challenge facing the utilitarian thinker is to find out which among the available alternative practices or policies gives (what we might call in today’s jargon) the best cost/benefit ratio.
11. So far as Bentham’s central views relevant to the theory of punishment are concerned, they can be stated in a dozen or so summary paraphrases, all of which can be found in IPML (along with much else) as well as in the manuscripts that survive intended for his ‘Theory of Punishment’:
Punishment may be defined as an evil inflicted on an offender by persons in authority. (More than a century earlier Hobbes had defined punishment in his Leviathan (1651) in much the same way).
Punishment may be justified, if at all, as the least worst use of coercion in the name of securing public safety, in the belief that the gain in security—the satisfactions of living in a secure society— outweighs the loss in punishment, which involves deliberately causing harm to offenders. In short, an increase in public safety achieved through general deterrence and individual reformation of convicted offenders.
‘All punishment is mischief: all punishment in itself is evil,’ Bentham says, and therefore is to be reduced to the minimum in severity (intensity, duration) consistent with securing its ends.12 This is Bentham’s paramount claim, and it marks off his theory from all nonconsequentialist theories of punishment.
Punishment is an evil because, in the nature of the case, it must be (or be perceived to be) painful or worse—and nobody likes being in pain.
Thus punishment is a practice essentially forward-looking (to prevent crime and secure public safety), not a practice essentially backward-looking (to render retribution based on desert). It has three goals: prevention of crime, correction of offenders, and compensation of victims.13 This is a corollary of his utilitarianism.
Punishments must be apportioned to crimes so that the would-be offender is always given an incentive to commit a less serious crime (e.g. rape) rather than a more serious one (e.g. rape and murder).
The punitive sanctions available to magistrates fall into three classes— political, moral, and religious. (Earlier Bentham identified a fourth sanction, the ‘physical’, but he abandoned it no doubt because each of the other three sanctions involves physical coercion or the threat thereof in one way or another). Of these the most important for the legislator is the political sanction—punishment authorized by the criminal law and imposed by a judge on someone found to have violated the law without excuse or justification.
Afflictive punishments, chief of which is the death penalty, are to be avoided wherever possible in favor of less brutal interventions.
The kinds of punishment most attractive to the general public are those that are most analogous to the crime, e.g. the death penalty for murder.
Torture cannot be ruled out categorically; it is justifiable if by torturing one guilty person many who are innocent can be spared injury—a direct application of the principle of utility to the evaluation of a public policy.
As viewed by persons likely to commit crimes, some punishments appear to be more severe than they are, others are just the reverse— they are less severe than they appear to be. Capital punishment is an example of the first, life imprisonment an example of the second.
Prisons should be run as private profit-making institutions.
Cases unmeet for punishment are those that are either groundless, inefficacious, or in unprofitable
Twelve rules define the proportioning of punishments to offenses, eg, according to rule 1, ‘The value of a punishment must be not less in any case than what is sufficient to outweigh that of the profit of the offence’.14 These rules are one of the most important contributions to the utilitarian theory of punishment, and carry that theory well beyond what any of Bentham’s predecessors had to offer. We are left, however, wondering whether there might be other rules that Bentham has overlooked.
Punishments vary from each other in eleven ways, e.g. in their popularity, remissibility, subservience to reformation, frugality, etc. These factors present Bentham at his analytical best; they (like the rules above) constitute an unrivalled contribution to the theory of punishment and endure to our day as well. Again, however, we are left wondering whether there might not be a further two or three such variables that Bentham overlooked.
12. One of the best examples of Bentham’s change of view on important aspects of his theory of punishment concerns the nature of imprisonment. In his early writings, he favored a threefold regimen of ‘Solitude, Darkness, and Hard Labor’—to which he sometimes added ‘Simple Diet’.15 The rationale seems to have been to break the rebellious spirit of the prisoner and thus reduce the likelihood of recalcitrance in prison and recidivism after release. Bentham came to see that his proposal if adopted would be bound to destroy anyone who endured it year after year. His later proposal of imprisonment under the Panopticon scheme (beginning ca. 1810) abandoned darkness and hard diet in favor of gainful labor outside the prison walls. As Semple put it in her excellent book on the Panopticon, ‘all would benefit; the injured party would get compensation, the master would profit from the prisoner’s work, and the prisoner himself would be working for a master of his choice instead of in prison’.16 What could be better designed to please a utilitarian than such a scheme? The miseries of detention and confinement would be reduced to the minimum and would easily be outweighed by the many benefits distributed among the several parties.
13. No doubt the most famous element in Bentham’s theory of punishment is his proposed Panopticon Prison, which more than any of his other projects occupied his mind (and depleted his pocketbook) for years, as he sought without success for support from Parliament to construct the necessary building. Bentham’s scheme was made public as early as 1791 in a series of twenty-one so-called ‘letters’ written four years earlier while travelling in Russia. As Semple pointed out, the fundamental management philosophy of the Panopticon was ‘deterrence and reformation’, and the administrative means to achieve these ends were ‘lenity, severity, and economy’.17 If this seems to verge on the self-contradictory, that is because it does.
14. Bentham’s most famous philosophical principle is the principle of utility, which he says is the ‘principle which approves or disapproves of every action whatsoever, according to the tendency which it appears to have to augment or diminish the happiness if the party whose interest is in question; or ... in other words, to promote or to oppose that happiness’.18 What makes him think that his various punitive proposals can be defended on utilitarian grounds? Or does he fail to address this crucial issue? If we consult the relevant passages in Bentham’s works, what do we find? I shall have to postpone trying to answer these questions for another occasion.
15. To what criticisms are Bentham’s punitive doctrines susceptible? There are at least three. First, his theory completely ignores any rights of due process and equal protection for the accused, the convicted, and the sentenced. In his view, all such practices can be defended, if at all, only on utilitarian grounds. Such rights as we do have are purely legal in source and nature; thus they are subject to review and revision at any time. Here we see emerging his legal positivism. Second, his theory tacitly encourages the authorities to ponder whether in a certain case it might be desirable to arrest and prosecute someone they are not certain is guilty, because of the prospective gain in perceived (even if not actual) public security from learning that an alleged offender had been arrested and prosecuted. Finally, his theory underestimates the corruptibility and sharp practice of managers of prisons conceived as profit-making enterprises.
16. Here I must close by suggesting that Bentham’s writings on punishment, which span half a century (the mid.1770s through the early 1830s), are a rich source worthy of much further study and reflection.
