In this part second of my article, I discuss Bentham’s treatment of punishment c.1780. The fundamental moral and psychological assumptions he made in developing his thinking about punishment were the same as those he made in his thinking about which acts to make offences. These were discussed in Part I.1 However, the material I now consider draws heavily on a source rarely mentioned therein, namely The Rationale of Punishment (hereafter referred to as RP).2 The manuscripts underlying RP were largely written just before An Introduction to the Principles of Morals and Legislation (hereafter IPML),3 as well as before most of the manuscripts on the penal code that IPML was originally intended to introduce. Both RP and IPML eventually appeared in print, but the penal code did not. I discuss the history of RP in Section 1 of this part of my article. As before, the central topic here is how Bentham applied the principle of utility to the problem at hand, in this case choosing the best types of punishment to use for each offence.
One distinctive feature of this part of Bentham’s thinking is the complex approach that he takes in RP to applying his fundamental assumptions to the design of punishments. We have seen that in IPML he describes a two-step process for applying these assumptions to the problem of deciding whether to make a type of action into an offence: first, determining whether the action is mischievous, all things considered; second, determining whether it is profitable to punish it. IPML suggests that these types of analysis would be carried out at a ‘universal’ level in the following penal code.4 In the case of RP, the corresponding task is to determine which types of punishment would be imposed on offenders who commit a given type of offence.
RP in effect describes a three-step process for determining this. First, it isolates twelve properties that are desirable in any type of punishment. I discuss Bentham’s treatment of these properties in Section 2. The next step involves ‘examinations’ of the various types of punishment to determine which ones have the best combinations of properties. I discuss the examination process, and one sample of it, in Section 3. The third analytical step is not carried out in RP itself, but we can take it to suggest that it would occur in a penal code of some sort. It would there be determined what the proper punishment for a given type of offence would be. Beyond that, there would be some guidance about an implied fourth step: the imposition of a punishment by a judge on an individual offender. Both books thus envision a penal code in which one or two important further steps in the process of applying the principle of utility are taken.5
I discuss in Section 4 how Bentham needed information about the crime-reducing consequences of punishments to carry out the punishment application process. He especially needed information about the causal process of ‘general prevention’ or ‘example’. Yet Bentham, like Cesare Beccaria before him, was writing in an era where he faced a curious theoretical obstacle: both began developing consequentialist thinking about punishment at a time when there was virtually no evidence about the consequences of various types of punishment – or, for that matter, of criminal actions, for instance forgery. Therefore what both men often did in thinking about processes such as general prevention was to employ their general background beliefs about the psychology of potential criminals. We will see that Bentham’s psychological assumptions led him to some surprising and troubling conclusions, ones that I will put in context.
In Section 5 I argue that Bentham always wanted better information about the consequences of acts and policies, as well as feasible methods for their evaluation; I will discuss some steps that he took to develop them. Finally, in the Appendix, I consider the extent to which the manuscripts achieve the goals for the code that are implicit in IPML, or use ideas about punishment expounded in RP.6
1. The Rationale of Punishment, An Introduction to the Principles of Morals and Legislation and the penal code
In this section, I discuss the history of the publication of RP and its relationship with IPML and the penal code. H.L.A. Hart aptly said of IPML that it, ‘like so many of Bentham’s works … had a difficult and delayed birth’.7 In some respects, RP had a far more difficult and delayed birth. In brief, Bentham probably began writing the manuscripts on which it is based in 1776, but he seems to have set them aside in late 1778.8 RP was finally published in 1830.9 In the intervening years the manuscripts had been edited, translated and published in French, and the French version was partly retranslated back into English in 1830.
Here is more detail on that lengthy process. Sometime before August 1808, Bentham gave the manuscripts underlying RP to his Swiss disciple, Etienne Dumont.10 Dumont edited and translated them into French. The first publication of Dumont’s recension, Théorie des Peines et des Récompenses, occurred in 1811.11 (This work also consisted of a volume that eventually became The Rationale of Reward.) Two further French recensions appeared in 1818 and 1832. In 1813 a long and positive account of the part of Dumont’s recension that dealt with punishments appeared in the influential Edinburgh Review.12 There are parts of RP that were obviously written after 1778 – not by Bentham himself, but by his two editors, Dumont and, much later, Richard Smith. Dumont and Smith must have written two chapters of RP by using Bentham’s other works. The first chapter contains criticisms of the British penal colony at Botany Bay in Australia, which was established in 1787–8 (490–7).13 The second is a defence of Bentham’s panopticon prison plan (498–503). Bentham published the original panopticon plan in 1791, although it was written in 1786.14
In the ‘Advertisement’ for RP there is an account of its history and editing, largely written by Dumont for the 1811 recension. He describes the manuscripts he was given as ‘very incomplete’, ‘sometimes consisting of fragments and simple notes’ (388–9). He adds that Bentham refused to review the translation, but simply informed Dumont that he had not changed his views on any of the basic principles governing punishment found in the manuscripts (389). When RP was finally published in 1830, Smith, its editor, partly drew on Dumont’s translation. Smith thus translated some of Dumont’s work back into English. However, he also stated that RP was not a literal translation of Dumont’s book, since, according to Smith, he [Smith] had ‘availed himself, whenever he could, of the original manuscripts’ (388).
Despite this assertion James McHugh, who edited a contemporary edition of RP, states that most of the surviving manuscripts of it in the University College London Library are in Smith’s hand, not Bentham’s.15 It appears that Bentham, who was 82 in 1830, had no significant role in producing RP.16 Legitimate questions remain about how much of RP consists of Bentham’s own words, and we can hope that a future edition of RP in The Collected Works of Jeremy Bentham will clarify this aspect of its complicated history.17
Yet even in the form we have it, RP is an important source for an investigation of Bentham’s thinking about punishment c.1780, together with his penal code. There is a close temporal connection of RP to IPML and the penal code, with much of RP being written just before IPML and many of the penal code manuscripts. The manuscripts on which RP was based were largely written between 1776 and 1778, while most of the penal code manuscripts were written from 1778 to 1780 at least, and possibly to 1782 or even 1785. IPML was to be the introduction to the code – most of it was also written from 1778 to 1780, and this material was actually printed in 1780.
Hence, these three great sources of Bentham’s thinking about penal law were mainly written in the period from 1776 to about 1782. It is important to keep in mind that none of the three works was completed by Bentham, although he did revise IPML and see it into publication in 1789. Even so, that work refers to three sections of Chapter 17 that were not published, as well as to parts of the penal code it was to introduce. As I mentioned in Part I of this article, Bentham wrote several other works on penal law c.1780.18
This was ‘a moment of acute crisis’ in the administration of English criminal law.19 The common British practice of the transportation of felons to America, which had started in 1718, ended when the American Revolution began in 1775. At first, it was thought that transportation could resume when the rebellion was over. In the meantime English prisons, not then commonly used as a punishment for felonies, became increasingly crowded, and naval hulks were used – on a temporary basis at first – as prisons. Meanwhile, a movement to reform the often appalling practices in prisons picked up steam with the publication in 1777 of John Howard’s influential work, The State of the Prisons in England and Wales. A major piece of reforming legislation, the Penitentiary Act, based largely on Howard’s thinking, was passed in 1779. However, the legislated national prisons or ‘penitentiaries’ were never built, and in 1787 Parliament decided to resume transportation, this time to Australia. The only work on penal law that Bentham published between 1776 and 1787 was A View of the Hard-Labour Bill of 1778, a set of comments and suggestions on the bill that became the Penitentiary Act.20 But Bentham was then busy first with what became RP, then with IPML and the penal code manuscripts.21 He was 30 years old in 1778. Janet Semple, an expert on Bentham’s panopticon prison plan, states that he ‘was in his prime, physically, intellectually, and creatively’.22
There are clear connections in content between RP and IPML. This is not surprising, given that both works discuss punishment, but there are even duplications in content. In three instances RP and IPML have similar chapters covering the same topic, the RP chapters clearly being earlier versions of the chapters in IPML.23 These chapters in IPML are the only ones explicitly devoted to punishment. In addition, two chapters originally intended for RP ended up in IPML,24 while certain gaps in the arguments of IPML are filled in by passages in RP.25 Finally, there is an explicit reference to RP at one point, asserting that it supplements IPML’s treatment of the ends of punishment. Other passing references to RP are also made.26
Despite these obvious connections, RP will strike the casual reader as very different from IPML. Of course, IPML contains a lengthy ‘division of offences’ and RP only discusses punishment, so this difference is to be expected. Yet there are some significant differences in how they treat punishment. IPML begins with an exposition and defence of the principle of utility, while RP rarely mentions the principle, and often only in passing.27 A second difference is that RP has very little psychological and philosophical material related to pleasure, pain, action, intention, motive and character. IPML contains twelve chapters devoted to these subjects and the chapters on punishment often refer to them.28 Both differences contribute to our sense that RP’s treatment of punishment is far less philosophically and psychologically grounded than is IPML’s.
However, in thinking about their relationship it is important to keep in mind that some differences are only terminological. So while RP does not open with a discussion of the principle of utility, for example, it does open with the statement that ‘punishment, whatever shape it may assume, is an evil’ (390). Chapter 13 in IPML, which begins its treatment of punishment, famously asserts that ‘all punishment is mischief: all punishment in itself is evil’.29 ‘Mischief’, the central term in IPML, is often used in RP (for instance 396, 397, 427, 430). However, in contrast to IPML, RP contains no analytical discussion of the nature of mischief. Utilitarian reasoning is thus implicitly deployed throughout RP. In fact there is one important explicit discussion of the principle of utility that seems to be an early version of IPML, Chapter 2 (411–13).
The best way to summarise the difference in Bentham’s treatment of punishment in RP and IPML is as follows. Both apply the principle of utility and certain psychological assumptions to the design of a system of penal law. As we will see, RP goes further than IPML in doing so, since RP evaluates various types of punishments, for example the death penalty, which IPML does not. However, IPML does discuss the principle of utility and certain concepts in psychology and ethics, such as mischief and intention, in greater depth. IPML was written just after RP and is markedly more sophisticated philosophically. Both works envision a penal code of some sort to carry the process of application forward, although Bentham only worked out his ideas about the structure of a penal code as he wrote IPML and the code itself.
In Sections 2–4 below I discuss the distinctive ideas about punishment that are most fully developed in RP. I do not consider Bentham’s basic psychological and moral assumptions about deterrence, even though they were first developed in the RP manuscripts (397–402). I have elsewhere examined these assumptions, which are more fully presented in IPML.30
2. The structure of The Rationale of Punishment and Book I
In this section, I discuss the structure of RP, along with the parts of Book I of most interest in understanding how Bentham proposed to carry out his project of applying the principle of utility. The process begins in Book I with his list of the twelve desirable properties that punishments can have. I discuss two of them, ‘characteristicalness’ and ‘exemplarity’, in detail.
The structure and main topics of The Rationale of Punishment
RP consists of six books. It has two main topics: a classification of the types of punishment and an evaluation of the best types of punishment. Its classification is an analysis of all the types or kinds of punishment that exist or could exist. In exploring this issue, Bentham first addresses the nature of punishment. He distinguishes it from several legal and social processes that resemble it, such as private vengeance, self-defence, legally-mandated compensation and taxation (390–5). He then introduces the classification or ‘division’ of all the types of punishment. The first and fundamental distinction is between the types that affect a person directly, which he calls ‘corporal punishments’, and the types that affect a person indirectly, for instance by depriving her of property or diminishing her ‘condition’ or reputation.31 Bentham calls these punishments ‘forfeitures’ (395–6). He subdivides these categories in turn. Corporal punishments, for example, include the categories of ‘simple afflictive punishments’, such as whipping, and ‘restrictive’ punishments, such as imprisonment; forfeitures include fines and the loss of an office or rank. There are also complex punishments that combine some of the simpler types. In the later chapters ‘examining’ each type of punishment, further subdivisions are mentioned.
The logical structure of this classification is similar in many respects to the classification or ‘division’ of offences in IPML Chapter 16. Indeed RP notes that a proper classification of punishments must have some similarities to a proper classification of offences, since both produce mischief (395–6).32
The second main topic of RP concerns which types of punishment are the best morally. Here Bentham is in effect applying the principle of utility to the evaluation of types of punishment, although, as already mentioned, he often does not frame the discussion that way. Again, his approach in RP envisions three basic steps. The first step, carried out in Book I, is to determine the properties of punishments desirable from a utilitarian perspective; the second consists of the ‘examinations’ which determine the types of punishment that have the best combination of properties; and the third is to determine which type of punishment will be imposed on a given type of offence.
There are several complications in the three-step process that should be mentioned. In the first place, the desirable properties can all be present in different degrees, so Bentham often addresses the question of the degree to which a desirable property is present. Second, there are different species of a given type of punishment, and some have more desirable properties than others. Bentham thus sometimes investigates which species is most desirable. For example, in examining imprisonment, especially the species found in England, Bentham says that in general any term of imprisonment will deprive an inmate of, for example, the liberty of going about the surrounding area when and how she pleases. In contrast, some prisons also deprive the inmate of the opportunity of seeing her friends or spouse and the chance to be kept clean and free of vermin. A prison can be operated without the latter conditions and Bentham describes such conditions as abuses – that is, as undesirable. The three species of prisons that Bentham later proposes are intended to be clean and properly heated, with visitation rights and so on (404).33 This brings out the fact that, third, he does not limit himself to considering whether desirable properties are present in a type or species of punishment. Bentham also considers whether undesirable properties such as filthiness are present. Finally, he examines certain combinations of types of punishment, an approach that he endorses in IPML.34
A central, although implied, conclusion in RP is that for many kinds of offences and offenders, the best type of punishment is a combination of some form of laborious punishment – that is, compulsory work – and confinement in a prison or penitentiary designed in the way that Bentham envisages.35 This prefigures his later, detailed panopticon plan. However, I will show that Bentham also concludes that other kinds of punishment have many desirable features, and it would be the third step of the application process – that is, a penal code – that would establish what kind or kinds of punishment would be imposed on a person who committed a certain kind of offence. Throughout this part of my article I discuss what we find in the manuscripts of the code that bears on this important question.
Bentham’s conception of the fundamental end or goal of punishment as such, and the causal pathways by which various types of punishments achieve this end, will be familiar to readers of IPML. He states that the end of punishment is the reduction in the number of offences. Of course, we need to add that achieving this end will promote the most happiness only if there is a utilitarian penal code in place, and that the punishments are designed to achieve goals such as ‘frugality’ or the minimisation of pain.36 Offence reduction can occur via three basic causal pathways: ‘prevention’, or what we now call ‘deterrence’; ‘disablement’, or what we now call incapacitation; and ‘reform’. Offences are ‘prevented’ – that is, offenders are deterred – when the infliction of punishment induces a fear in a potential offender that he will be punished if he offends, so that he refrains from offending.37
Bentham subdivides this pathway into ‘particular prevention’ and ‘general prevention’. Particular prevention occurs when the offender who is punished refrains from reoffending out of fear of being punished; general prevention occurs when someone else refrains from offending out of fear of being punished (396).38 What Bentham calls ‘general prevention’ he also refers to as ‘example’ in IPML and RP (392).39 He further states that example is the most important property of a type of punishment (396).40 A disabling punishment such as imprisonment takes away a person’s ability to offend. Bentham briefly characterises reform as the result of certain punishments ‘taking away the desire of offending’ (396, cf. 404, 425ff.) We will consider his thinking about this pathway below. IPML mentions the three basic pathways, but treats reform briefly.41
The twelve desirable properties
One chapter of Book I discusses twelve desirable properties of punishments (402–6). Three other chapters in that book expand on the significance of three of the properties (407–13). Chapter 15 of IPML closely follows RP’s discussion of the twelve.42 We can formulate Bentham’s thinking this way: given the principle of utility, a type of punishment’s having any of the twelve properties to a large degree is desirable, or has value, ceteris paribus.43
The twelve properties are as follows: (1) variability (explained below); (2) equability (the fact that a given amount of a type of punishment has similar effects on all offenders); (3) commensurability (the property of a set of types of punishment that yields the result that all potential offenders rank the punishments for all offences in the same way); (4) characteristicalness (see below); (5) exemplarity (see below); (6) frugality (see below); (7) subserviency to the reformation of the offender (see below); (8) efficacy in disabling the offender; (9) subserviency to [providing] compensation; (10) popularity (that is, acceptance by the populace); (11) remissibility (the possibility of remitting some or all of the punishment once it begins); and (12) simplicity of description (a property that enables potential offenders to know what they will undergo if they choose to offend).44
Some of these properties are more abstract. One example is ‘variability’, which is the possibility that a type of punishment can be imposed in varying degrees of severity. The more interesting and important properties are less abstract and more closely connected to the psychological processes that Bentham focuses on. In this section, I carefully discuss characteristicalness and exemplarity. I will later discuss subserviency to the reformation of the offender.
Characteristicalness, Analogy and Retaliation
Characteristical punishments have some analogy to the offence being punished (cf. 403–4).45 In RP there is a short chapter which discusses various kinds of characteristical punishments designed to make different analogies (407–9). One kind would be directed at the bodily member that was used to commit the crime. A punishment for forgery could thus involve transfixing the offender’s hand with an iron device ‘fashioned like a pen’ (408). Another principle of analogy would use a punishment similar to the means employed by the criminal: subjecting an arsonist to a fiery punishment or poisoning a poisoner (407–8). Bentham states that he intends only to illustrate the concept of an analogous punishment but is not recommending any (407). Since, according to RP, there are eleven other properties that a punishment may have or lack, there may be reasons for not using analogous ones (406). Bentham states that characteristical punishments are often popular – that is, they often have the property of popularity (405).46 Conversely, he acknowledges that some characteristical punishments may be ‘hateful’, and are hence unpopular. This might be true, he says, when poisoning produces ‘convulsions and distortions’ (408, cf. 483).47
Characteristicalness is an important property in Bentham’s thinking in RP. He states there that ‘for great crimes’, characteristicalness is one of the two most important properties a punishment can have (406, cf. 449).48 Its importance rests on psychological claims traceable back to Beccaria. One of Beccaria’s claims is that the association of ideas is a fundamental feature of human psychology; another is that an effective system of criminal law can create a firm association in the minds of potential offenders of the idea of a crime and its punishment. If it does this, potential offenders, when considering whether to commit a crime, will realise that there is a punishment for it; this will in turn reduce their inclination to commit the offence. Beccaria uses these two claims to argue for punishments that promptly follow crimes – this, he says, strengthens the association of ‘crime’ and ‘punishment’ in the minds of potential offenders.49 He also argues that the phenomenon of the association of ideas favours punishments that conform as closely as possible to the nature of the crime.50
Bentham argues in a similar fashion. He asserts that if there is an analogy between a crime and its punishment, this makes it more likely that a potential offender will think of the punishment for an offence when she considers committing it. While he states that similarity is one source of analogy, and his examples involve various similarities, he adds that contrast provides another source of analogy. (He gives as an example the idea of a giant, which has an analogy to the idea of a dwarf (407).) Bentham concludes that a characteristical punishment for a given type of offence has three psychological advantages: it is more easily learned, comes to mind more readily if a person considers committing the offence and is more easily remembered (403).51
The closest analogy to a criminal act is retaliation – that is, harming or injuring the offender in the same way that she harmed her victim and to the same extent that she was harmed (409–10). This ancient idea, Bentham grants, is generally popular. But he argues, as William Blackstone did, that it is subject to many objections.52 For one thing, with crimes ‘of a public nature’, such as treason, retaliation is impossible because the offence has no specific individual victim; there is consequently no way to make the offender suffer as his victim has (410).53 Bentham asserts that even when retaliation is possible, it tends to ‘err on the side of excessive severity’. ‘Its radical defect’, he says, is ‘its inflexibility. The law ought so to apportion the punishment as to meet the several circumstances of aggravation or extenuation that may be found in the offence’ (411).
I will explain his reasoning. An aggravating circumstance of an offence is one that calls for increased punishment.54 In IPML, for example, Bentham calls a ‘degree of deliberation’ an aggravating circumstance, roughly corresponding to what is now called ‘premeditation’. He says that if a person plans to commit an offence and does so, this will typically call for an increase in the punishment for it.55 If an offender commits an offence unintentionally, Bentham says, this is often an extenuating circumstance calling for a reduction in punishment.56 Bentham’s criticism is that retaliatory punishments for three instances of one type of offence – producing a ‘simple corporal injury’, say – would be inflexible. This is because they would inflict the same loss that the victims suffered on three offenders, even if one of them carefully planned the crime, one committed it intentionally but on the spur of the moment, and one committed it unintentionally.57
One characteristical punishment that Bentham regards as appropriate is castration for rape (411). The penal code manuscripts incorporate some characteristical punishments, including, in fact, the ‘iron pen’ for forgery.58 Two others are the punishment for treason, namely being shot out of a cannon towards the country which the offender assisted, and ‘scorching’ by fire for aggravated theft, when ‘advantage is taken of the calamity of a fire’.59 There are also retaliatory punishments in the manuscripts. A notable example is the punishment of incendiarism (that is, arson) by burning the offender ‘over a slow fire’.60 The Appendix provides more information on characteristical punishments in the manuscripts.
While there is no question that Bentham favoured some characteristical punishments c.1780, it is not clear how long he continued to hold this view. He played virtually no role in Dumont’s translation of the manuscripts underlying RP, so the passages in RP that tentatively endorse them might not have represented his opinions by 1811. It is interesting to note that when Bentham defended his panopticon plan in 1802, he mentioned five desirable properties of punishment or penal justice in general (not twelve), and he did not include characteristicalness among them.61 So it is possible that his focus on his prison plan brought about a significant change in his thinking. However, we need to remember that the panopticon is not a penal code, so it might rather be the case that Bentham then favoured the use of the panopticon for some generally more serious crimes, and characteristical punishments for some others.
Hart rightly described a few of Bentham’s characteristical punishments in RP as ‘grim and sometimes grotesque’, adding that they are ‘often repellent to modern taste’.62 He went on to make the valid point that the crimes to which they were attached were then often punishable by death in England.63 One clear example of this is forgery, which was a capital offence in 1780.64 Another is Bentham’s proposed punishment for treason, which actually seems more lenient than the horrifying medley of steps required by the English law of his time.65 However, Hart failed to mention an important feature of some of Bentham’s characteristical punishments, which we will now consider.
Exemplarity
Exemplarity plays an important role in RP and leads Bentham to some startling conclusions. It also calls for close attention.
The word ‘exemplarity’ would seem to mean the property of being or serving as an example. And, as mentioned above, when Bentham says that a punishment serves as an example he means that it produces some ‘general prevention’ or general deterrence (for instance 392, 396).66 However, Bentham’s definition of ‘exemplarity’ narrows the meaning considerably. In RP he defines it to mean that a punishment having this property appears to be more painful than it really is. In his terminology, a punishment has exemplarity when its ‘apparent’ painfulness is greater than its ‘real’ painfulness (404).67 The more painful a punishment seems to be to observers, without really being painful, or without being nearly as painful, the more exemplarity it has. We can also say that the more misleadingly painful a punishment is, the more exemplarity it has. Bentham’s definition narrows the meaning of exemplarity in part because it entails that only an observed punishment can have this property, whereas someone can be deterred by a punishment she does not observe. This could happen if, for example, she hears or reads about a punishment of a given offender, or if she was taught at school, say, that forgers are hanged.
Bentham’s concept of exemplarity is designed to apply to public punishments. In fact many punishments in England in 1780 were carried out in public: the pillory and hanging are well-known examples. It is probably less well known that prisons, which detained many sorts of prisoners, were usually open to the public during the day. When Howard began visiting English prisons around 1773, they were largely open to visitors. Among the reasons for this openness were that debtors and people awaiting trial, who were often housed with convicts, were entitled by law to have visitors; gaolers were known to scrimp on food to prisoners, or to deny it altogether, so private charity could make up for this; and gaolers, who ran prisons as profit-making enterprises, often sold beer inside. Allowing visitation was consequently an opportunity to increase their profits.68 Unrestricted access also meant, though, that the visitors could include prostitutes and fellow criminals. Howard, who documented these circumstances, did not explicitly address the issue of restricting public access to prisons in the section of his book devoted to proposed improvements,69 but the Penitentiary Act of 1779, which he helped to write, did restrict it.70
Highly restricted access to English prisons became the rule later,71 but it was a development that Bentham did not completely accept. The Penitentiary House, one of the three kinds of prison that Bentham outlines in RP, envisions public access (431). The original panopticon plan proposed that the doors of these prisons would generally be ‘thrown wide open to the body of the curious at large – the great open committee of the tribunal of the world’.72
As noted above, Bentham states that ‘example’, in the sense of general prevention or general deterrence, is the most important property of a type of punishment (396).73 The utilitarian reason for this is clear: the punishment of one person can affect the behaviour of many more people than the offender, and this can significantly increase the beneficial offence-reducing effects of it (396).74 Bentham also states that exemplarity is one of the two most important properties that punishments for ‘great crimes’ can have (406). The utilitarian reason for this is less clear. From what I have said thus far, it seems that we can only say that, if punishments are going to be public, their exemplarity would be desirable, given the principle of utility. But we have not seen why the principle would favour public punishments. RP states a reason for making punishments public and observable, based on another important psychological claim that Bentham drew from Beccaria. This statement occurs in a passage in the chapter on transportation, which discusses the effect of the suffering of English convicts in the penal colony in Australia on potential offenders in England. It asserts that it barely has any effect ‘upon that class of people who are most likely to commit offences, who neither read nor reflect, and whose feelings are capable of being excited not by the description, but by the exhibition of sufferings’ (492).75 This passage was clearly written after most of RP, but another one occurring in a discussion of banishment and probably written in the late 1770s, is similar (434–5).76 Beccaria criticised the practice in his native Lombardy of deporting criminals to work as galley slaves in Venice.77 He based this criticism on another general psychological assumption. Beccaria asserted that ‘experience has shown that the common crowd does not adopt stable principles of conduct’, so that, in order to preserve society, the law must engender motives ‘that have a direct impact on the senses’.78 Thus both Beccaria and Bentham asserted that the rational limitations of most potential offenders generally call for public punishments.79
What is distinctive about Bentham is that he uses his concept of exemplarity to argue for the desirability of punishments that look more painful than they are. His argument for this conclusion explicitly mentions the principle of utility, as well as two other desirable properties of punishment. He states that the principle of utility speaks in favour of imposing ‘real’ pain on an offender insofar as it is needed to reform him or to compel him to provide compensation. He goes on to argue that, insofar as punishment is meant to serve as an example, the mere appearance of its being painful may suffice. Observers of a punishment can form a belief that it is painful even if it is not, or is not as painful as it appears. RP and IPML both emphasise that it is the ‘apparent’ painfulness of punishment ‘that acts upon the mind’ of observers (404).80 The principle of utility therefore can favour deceiving observers about how painful a punishment is when its deterrent effects on them can be achieved by deception, since the evil or mischief of it is reduced, while these beneficial effects are produced anyway. Yet Bentham does not argue that punishments should not be painful at all. He argues that a degree of real pain is needed to reform the criminal, and to compel him to provide compensation, but that apparent pain may serve to provide a (possibly misleading) example. Summarising these ideas, perhaps the most famous passage in RP reads as follows:
The real punishment ought to be as small, and the apparent punishment as great as possible. If hanging a man in effigy would produce the same salutary impression of terror [that is, fear] upon the minds of the people, it would be folly or cruelty ever to hang a man in person (398).81
RP contains several passages that describe deceptive punishments. Two characteristical punishments are so described. The ‘iron pen’, mentioned above as the punishment for forgery, was to be inserted through the offender’s hand, but the visible part of it would be thicker than the part that pierced his hand, giving the impression that it was thick throughout (408). The iron pen is also mentioned as a punishment for forgery in the penal code manuscripts, though without elaboration.82 RP states that an offender who committed calumny or disseminated ‘false reports’ could likewise have his tongue – the bodily member employed in his offence – pierced by a spike. The visible part of it would be thicker than the part that pierced the tongue, also giving the impression that it was thick throughout (408). The ‘tongue spike’ is also mentioned in the manuscripts as a punishment for perjury, again without elaboration.83
Here is the place where Hart’s remarks about characteristical punishments need reconsideration. He failed to note that Bentham endorsed some characteristical punishments that would seem to observers to be more painful than they really were. The element of ‘real’ leniency in them shows that his approach to punishment was not barbaric, but rather a rational effort to balance the competing considerations favoured by the principle of utility. Or, to express the point differently, even some of the more repellent examples manifest Bentham’s commitment to ‘frugality’, another desirable property of punishment. In RP frugality is the property of punishments by which they produce no more pain than is needed to achieve the aims of punishment (404).84
The idea that a punishment should be misleadingly painful has clear drawbacks, even given Bentham’s psychological assumptions. There is the very real risk that the deception would be revealed by an offender who had experienced it, or by an official who imposed it, thereby reducing the effectiveness of the punishment on the observers of it. Another problem with deceptive punishments would arise when Bentham adopted the idea that a utilitarian penal code should include the reasons for the kind of punishment to be imposed on a convicted offender. It would obviously be self-defeating publicly to announce the use of a punishment designed to deceive those who were observing it.85
Public punishment in general
This is an appropriate place to consider the general idea of punishing people in public. I have discussed Bentham’s endorsement of some characteristical punishments, for instance the iron pen, that would also have exemplarity, and would thus be carried out in public. I also mentioned a proposed public characteristical punishment that would not have exemplarity, namely the burning of an arsonist. RP discusses a third type of public punishment, namely the ‘ignominious’ (458–63). This is a type of forfeiture intended to diminish the offender’s reputation or to increase his disrepute. Bentham says that all punishments, especially public, corporal ones such as whipping, are ignominious and have these effects on the moral sanction, that is, on a person’s reputation. But he adds that there are some punishments, inflicting no ‘organical’ pain as whipping does, that do nothing more than diminish a person’s reputation. These he describes as ‘simple ignominious’ punishments. Two examples he mentions are the pillory and ‘ignominious dress’ (460–1). Two examples of the latter mentioned in the code are ‘the adulterer’s coat’ and the distinctive coats for each category of culpable insolvents.86 These would not have exemplarity, nor characteristicalness, but they would be public. Bentham’s ‘examination’ of simple ignominious punishments in Book III is generally favourable (458–67).
Although several types of punishment like pillorying did take place in public in England c.1780, and most prisons admitted the public, it is a much-discussed historical fact that some common public English punishments, notably whipping and the pillory, were declining in frequency in England by the late eighteenth century; eventually practically all were abolished. That is, insofar as whipping and hanging continued to be used at all, they were inflicted in prison and out of public view,87 and access to prisons was becoming more restricted. Bentham c.1780 proposed a number of new public punishments, falling into the three categories: characteristical and having exemplarity; characteristical and lacking exemplarity; and lacking characteristicalness and exemplarity. He also endorsed public access to prisons of his design. Bentham thus seems to have been somewhat unusual in his day in looking with favour on expanding the range of public punishments.
But it must be noted that Bentham acknowledged that all punishments observable by the public have drawbacks from a utilitarian perspective. It is difficult to deny that observing a punishment can have a powerful effect on observers. It should be added, however, that this effect may not be the one that the authorities desire. In fact, one of the reasons why public punishments such as the pillory had started to be phased out in England in the late eighteenth century was that the crowds who observed them sometimes showed sympathy for the offender (cf. 466 n).88 On the other hand, crowds might also treat pilloried offenders brutally, occasionally killing them with projectiles such as stones.89 Bentham states that the severity of the pillory consequently ‘depends upon the caprice of a crowd of butchers’ (417). He acknowledged, we saw, that some characteristical punishments, such as poisonings, may be ‘hateful’ and unpopular (408). He also asserted that the ‘infamy’ of offenders punished in publicly accessible prisons could preclude their reintegration into society (431, cf. 416, 441). His solution in this case was to mask the prisoners and disguise any of their distinctive bodily features when the public was present, declaring ‘The business is … for the sake of general prevention, to render the offence infamous, and, at the same time, for sake of reformation, to spare the shame of the offender as much as possible’ (431). This, too, Bentham regarded as an instance of exemplarity, where the real pain of the offenders is reduced and the apparent pain of their punishment is increased (431).90 Lastly, Bentham argued that ‘complex afflictive punishments’ that leave observable marks on convicts are generally undesirable because they too preclude their reintegration into society (416).91
3. Laborious punishment examined
The eighteen chapters of RP in Books II–III often carry forward the form of moral assessment that began in Book I. They are efforts to complete the second step of Bentham’s application of the principle of utility to the design of punishments. In theory, he says he will analyse how well different types of punishment, for example imprisonment and fines, instantiate the twelve desirable properties.92 Another way to describe Bentham’s approach here is to note that he is no longer discussing various abstract goals; he is rather addressing the question of how to achieve these possibly conflicting goals. (We have seen above that Bentham appreciated that general prevention and reformation of the offender could be in conflict.) He therefore presents ‘examinations’ of the beneficial and detrimental features of various types of punishment. I mentioned that the discussions of the properties of punishments in Book I of RP largely parallel Chapter 15 of IPML. It is a striking fact about IPML that it does not consider the beneficial and detrimental features of different types of punishment in the way that RP does. These examinations thus take Bentham’s project of applied ethics, as it pertains to punishment, considerably further than does IPML.
Ten of the eighteen chapters in these two books contain reasonably complete examinations of a type of punishment; two contain less complete ones. Bentham also examines the non-legal ‘moral sanction’ in this fashion.93 The ten types of punishment considered are the following: simple afflictive (that is, milder corporal) punishments, complex afflictive punishments (that leave permanent marks), imprisonment, restrictive punishments (either confining a person to a specific area or banishment), laborious punishments (that is, compulsory work or ‘hard labour’), capital punishment, the moral sanction, simple ignominious punishments, simple restrictive punishments (prohibiting the practice of a profession) and pecuniary punishments or fines.
In this section, I will focus on the examination of ‘laborious punishment’ (437–41, esp. 439–41). There is another important chapter in RP that examines capital punishment and reaches the same radical conclusion that Beccaria did: namely that it should be abolished, even for crimes such as murder. I will not discuss this examination in detail because Hugo Bedau has written an excellent discussion of it, but I will consider its central argument in Section 4.94 The importance of the examination I discuss derives in part from the fact that we can see how Bentham’s thinking was moving towards the panopticon plan of 1791. We can also see how Bentham was responding to the thinking of English prison reformers such as Howard and Jonas Hanway.95
Bentham’s approach in examining any type of punishment is often simply to consider the extent to which each of the properties discussed in Book I of RP is instantiated in that type. However, he sometimes takes this form of examination further because, as I said, when he finds that one species of a type of punishment falls short of instantiating fully a certain desirable property, he considers whether another species of that type instantiates it more fully. In examining imprisonment, especially the various species then found in England, Bentham, like Hanway and Howard, emphasised that it could involve ‘promiscuous association of prisoners’, such as insolvent debtors mingling with experienced burglars (427). Doing this could mean that the prisoners are not reformed; on the contrary, they might be corrupted (427–9). His main response is to propose three types of prison intended to house different types of prisoners. These are: the ‘Black Prison’, designed to house the worst criminals, who would be confined there for life; the Penitentiary House, which would house some less dangerous offenders, who would eventually be released; and the House of Safe Custody, a prison mainly for debtors (429–31).
Imprisonment is a type of punishment that Bentham treats in some detail (420–31). As already mentioned, the appalling conditions in English prisons had become a matter of widespread concern after 1777, when Howard methodically described them in The State of the Prisons. He had visited virtually every prison in England and Wales at his own expense, risking his health and even his life in institutions known for their rampant ‘gaol fever’.96 Bentham cites Howard’s book four times in RP (422, 426, 430, 432),97 and he clearly accepted many of his observations on the defects of English prisons and his proposals for reform.98 He met Howard in 1778 and wrote a vivid letter describing him, his career and their meeting. Bentham wrote: ‘He is, I believe, take him for all in all, one of the most extraordinary men this age can show.’99
Laborious punishment is one of Bentham’s terms in RP for what he calls ‘penal labour’ in IPML.100 The bill that Bentham commented on in 1778 speaks of ‘hard labour’. Compulsory work for the benefit of the state, such as building roads or dredging harbours, was an innovation in eighteenth-century English penal practice,101 although other countries, for example Venice, had used punishments such as galley slavery for centuries. Howard and Hanway both argued for requiring, or at least permitting, some sort of work in prison.102 Bentham’s examination of laborious punishment reaches the conclusion that it is the best single form of punishment – although, in practice, it would always be combined with imprisonment, since labourers must be housed and fed (439–41). This implies the further conclusion of RP as a whole, mentioned above: for many kinds of offences and offenders the best type of punishment is some form of laborious punishment that is performed in the Penitentiary House (425–7, 429–31).103 The idea of a prison separating prisoners into cells in which they would mainly spend their time doing profitable work is central in Bentham’s later panopticon plan.104
Bentham’s examination of laborious punishment discusses eight desirable properties, and finds that it has seven of them to an acceptable degree. I will discuss the five most important properties, central to his assessment of this type of punishment.
Convertibility to profit
Bentham clearly thinks that the main advantage of laborious punishment is that it is ‘convertible to profit’. This property is not actually among the twelve desirable properties listed in Book I, but it is closely related to ‘frugality’ and ‘subserviency to compensation’(404–5).105 The utilitarian thinking behind the desirability of laborious punishment is that a person’s labour can produce things that can be useful in some way, which Bentham speaks of as yielding a ‘profit’ in a broad sense, not necessarily a monetary profit (438).106 Insofar as any punishment of an offender yields beneficial results for someone, this serves to offset the mischief that her offence caused in the first place, as well as the mischief of the punishment itself (404–5).107 Laborious punishment could be beneficial in various ways: it could benefit the public if it involves the building of infrastructure such as roads; it could benefit the public if it reduced the costs of administering the prison by, say, requiring the prisoners to clean it; it could benefit the public or private contractors if the prisoners do work that generates a monetary profit; it could benefit the prisoners themselves if they receive some pay for their work; and, finally, it could generate income that could be transferred as compensation to the victim. In RP, Bentham focuses on the possibility of compensating victims (396, 406). He conceives of the work as unremunerated, though he grants that ‘labour obtained by the force of fear is never equal to that which is obtained by the hope of reward’. (Such an assertion leads Bentham to suggest that the gradual emancipation of slaves ‘would be a noble and beneficial measure’ (441).) In A View, however, he approves of the idea of allowing prisoners to profit from their work.108
Subserviency to reformation
RP contains some serious discussions of this property of punishment (404, 425–9). This contrasts with the brief and somewhat dismissive treatment that it receives in IPML.109 Bentham claims in IPML that all punishments are subservient to reformation since they tend to make offenders averse to reoffending.110 Here Bentham is conceiving of the process of reformation as being what he calls in RP ‘particular prevention’, or specific deterrence (396). That is, he is thinking of all punishments as making offenders more fearful of undergoing them again, and hence less likely to reoffend. In RP, however, Bentham in effect distinguishes between reformation and particular prevention.
All punishment has a certain tendency to deter from the commission of offences; but if the delinquent, after he has been punished, is only deterred by fear from the repetition of his offence, he is not reformed. Reformation implies a change of character and moral dispositions (404).
We might express his thinking this way: there is a difference between a process in which punishment causes a convict to refrain from reoffending out of fear of undergoing it again, and a process in which punishment causes a convict to feel guilt for his wrongdoing and he resolves to refrain from wrongdoing for moral reasons.111 Punishment leads to the same lawful behaviour in both cases, but the agent’s reasons for behaving so differ. I believe Beccaria never makes this important distinction, while Bentham perhaps found suggestions of it in Hanway and Howard.112 Bentham’s claim that laborious punishment has the property of subserviency to reformation means that he believes that it tends to act on prisoners in the second way (440, referring to 425).113
In explaining how laborious punishment tends to reform prisoners, Bentham appeals again to psychology. He says that work tends to engross the attention of prisoners and keeps them from corrupting thoughts of past or future criminal activity. In this way, laborious punishment is like the separation of different classes of prisoners: it serves the minimal moral goal of preventing corruption. However, Bentham continues, such punishment is more positively reformative. The force of habit, a pervasive tendency in human nature, will eventually make most forms of work tolerable, despite its being compelled. He says, ‘few occupations are so irksome that habit will not in time make them sit tolerably easy’ (440). Furthermore, on release from a penitentiary, prisoners will be able to choose their work (440).
Bentham seems to focus on producing what we would now call ‘good work habits’. IPML contains some passages which suggest why he does so. They run as follows. Most crimes are property crimes committed by poor people seeking money or things of value that will enable them to buy necessities or other things they want (cf. 427). The motives that move them to steal are thus usually ‘indolence and pecuniary interest’, that is, a desire for money or things of monetary value and a desire to avoid the pain of working for them.114 The conclusion he draws in RP is that a regime of laborious punishment will extirpate indolence and thereby remove the main source of motivation to commit property crimes. It will not extirpate a prisoner’s pecuniary interest, though, since that is a pervasive form of human motivation,115 and Bentham always envisioned released and reformed prisoners working for income.
This is not an entirely convincing argument. If a prisoner always had a general desire for monetary gain and she acquired in prison a habit of working hard for it, she would not necessarily be morally reformed. She might simply be willing to work hard for some criminal enterprise when released. It seems that the best way to describe the effect of laborious punishment that Bentham describes is simply producing people with good work habits. But it is not clear that it would produce people who would only want to work diligently in morally permissible ways or, as Bentham says, to pursue ‘lawful gains’ (499).
Although IPML treats the subject of reform briefly, it does contain a sophisticated treatment of moral psychology.116 This points in another direction – namely, that if we want prisoners to reform morally, we at least want them to develop ‘tutelary’ motives that would restrain them from earning money in impermissible ways. IPML states that benevolence is the most effective tutelary motive.117 Furthermore, it states that ‘good-will’ or benevolence is the only ‘purely social’ motive, in contrast to ‘self-regarding’ motives such as self-preservation and pecuniary interest and ‘semi-social’ motives such as the love of reputation and the motive of religion.118 This suggests that if punishment produced benevolence in prisoners, they would refrain from offending for a moral reason, as Bentham understands this, in contrast to refraining from a ‘self-regarding’ reason, for example the desire to avoid suffering. However, if we suppose that reform involves increasing a person’s benevolence, it is hard to see how the sorts of laborious punishment Bentham presumably envisioned – such as weaving or assembling furniture – would make prisoners more benevolent.
This is not to say that there is no benefit to society or to the prisoner herself from compelling or encouraging her to develop good work habits. The point is that producing such changes in prisoners would not itself be a moral reformation, as IPML suggests we understand that idea.119 Here is a case in which the more sophisticated moral psychology of IPML might have improved the detailed consideration of punishments in RP. Not only were RP and IPML never completed, but Bentham never integrated the insights of IPML into the analytical framework of RP.
Exemplarity
Bentham says that laborious punishment does not have exemplarity to any great degree, but it does not entirely lack it either. He adds that since offenders performing such work would be imprisoned, they could be required to wear distinctive uniforms: this would convey the idea that they are being compelled to work and suggest that they find it unpleasant (440). He must be assuming that the prison would be accessible to observers, and that the prisoners are not finding the work as unpleasant as observers take it to be, otherwise exemplarity as Bentham defines it would not be instantiated. However, in this instance, there would not be exemplarity in the trompe l’oeil sense that occurs in his tongue spike example.
Characteristicalness
Bentham claims that laborious punishment ‘is not altogether destitute of analogy, at least of the verbal kind, to that class of crimes which are the most frequent’, by which he means property crimes (440). He gives two examples of analogous punishments, in which the behaviour punished would not actually constitute property crimes: slothful offenders, who are compelled to work, and vagabonds, who are confined to a particular spot. He adds: ‘[t]he more opposite the restraint thus imposed is to the natural inclination of the patient, the more effectually will he be deterred from indulging his vicious propensities’ (441).
There is a tension here in Bentham’s reasoning. Analogies of punishments to crimes usually represent similarities of one to the other, for instance when a corporal injury inflicted on the offender is similar to the corporal injury inflicted on the victim (408). Yet here Bentham is claiming that forced labour is analogous to the offence of choosing not to work. In the penal code manuscripts, we find that hard labour may be used as part of the punishment for robbery and theft; here, too, the punishment is not analogous to the crime.120 This difficulty also arises for the combination of laborious punishment and imprisonment, which Bentham implied was the best form of punishment for many kinds of offences. Such a combination of punishments is also not analogous to crimes such as theft and robbery. I noted that when Bentham defended his panopticon plan in 1802, he mentioned five desirable properties of punishment in general and did not include characteristicalness among them.121 It may be that by then he had realised that his favoured form of punishment for many serious crimes was not characteristical for most of them.
Popularity
Bentham acknowledged one important drawback of laborious punishment, namely that it was unpopular in England. In this instance there is not merely a deficient degree of a desirable property, but also some degree of an undesirable property. The objection often made to laborious punishment, he says, is that in England, where the liberty of citizens is protected, punishment through slavery is thought to be intolerable (411, 441).122 In RP, Bentham makes several brief replies. In his examination he makes two.123 First, laborious punishment is not slavery (441). He does not elaborate on this claim – perhaps the difference is supposed to be that such punishment can be limited in time, whereas slavery is a lifelong condition. Second, in any case, the issue is not whether the word ‘slavery’ applies to the practice. The examination process is what will show how desirable all things considered laborious punishment are. And if this process shows that laborious punishment is desirable, all things considered, then the fact a certain word may be applied to it is no reason to reject its use.
If then upon examination it is found not to be possessed, in a requisite degree, of the properties to be wished for in a mode of punishment that, and not the name it happens to be called by, is a reason for its rejection: if it does possess them, it is not any name that can be given to it that can change its nature. (441)
Here we see an early example of Bentham’s response to non-utilitarian forms of reasoning and feeling. The chapter in RP on popularity takes his thinking further (411–13). There is considerably more discussion in IPML, where Bentham criticises the principle of sympathy and antipathy, and reflects upon the significance of popularity.124 The related topic of antipathetic pain plays an important role in Bentham’s manuscripts on paederasty.125 However, Bentham treats ‘popularity’ as one of the twelve desirable properties of a kind of punishment, so it is a force that must be reckoned with in the examination he undertakes here. We might then put Bentham’s response this way: whether laborious punishment is called ‘slavery’ is not relevant to the examination, but whether people disapprove of laborious punishment is relevant.
Summary
Bentham concludes that laborious punishment is the best single form of punishment for many offences, and that the combination of it with a certain kind of prison has still more desirability. He states that laborious punishment does not possess some of the seven desirable properties he considers in a marked degree, for example exemplarity. But Bentham in effect does state that it has only one undesirable property in England, namely that it is unpopular. He might well have perceived this as a relatively minor problem, and that disseminating his utilitarian examination would dissipate the resistance to laborious punishment. I have argued that there are weaknesses in Bentham’s discussion of the characteristicalness, exemplarity and the subserviency to reformation of laborious punishment. This means that even using his own criteria, the case for the all-things-considered desirability of laborious punishment is weaker than he acknowledges. But, as I said, he concludes that the utilitarian case for it is strong.
We should remember that RP implies that there is to be a third step in applying the principle of utility to the design of punishments. This step would take place within a penal code. The material in the code would explain which offences were to be punished by hard labour; it would probably also give some guidance to judges about how much of it should be imposed on different kinds of offenders.
4. Measuring or estimating general prevention
In this section I discuss Bentham’s treatment of what he describes as the most important beneficial consequence of punishment – namely example, or general prevention. Despite its importance in Bentham’s thinking, this is a somewhat elusive concept in RP. Bentham surely believed that the measurement, or at least the estimation, of the general preventive effects of a type of punishment was of great importance for his project of applying the principle of utility to the design of penal law. However, he had no data about these consequences c.1780 when he wrote RP, nor when he wrote IPML or the penal code manuscripts. I now discuss how he dealt with this situation.
Bentham’s goal was to assess how desirable the different types of punishment are, all things considered. Bedau in effect suggested that Bentham provided no ‘weighting’ of the twelve desirable properties, so that it is not clear how much ‘all-things-considered desirability’ any type of punishment possesses.126 For example, consider this relatively simple question about the weights of the twelve properties: if a type of punishment has some characteristicalness but no exemplarity, and tends to corrupt all offenders to some extent, does it have any all-things-considered desirability? Since corrupting offenders is an undesirable characteristic, and having no exemplarity means that the type has no desirability in this respect, and, finally, having characteristicalness gives it some desirability, the question is asking about the net or all-things-considered desirability, given the three partial or pro tanto evaluations. If characteristicalness is very important, and thus a property with great weight, it might be all-things-considered desirable. If corruption is very undesirable, on the other hand, it might not.
In RP, Bentham gives this brief assessment:
There is no one lot of punishment which unites all of these desirable qualities [that is, properties]; but, according to the nature of the offences, one set of qualities are more important than another.
For great crimes, it is desirable that punishments should be exemplary and analogous. For lesser crimes, the punishments should be inflicted with a greater attention to their frugality, and their tendency to moral reformation. As to crimes against property, those punishments which are convertible to profit are to be preferred, since they may be rendered subservient to compensation for the party injured (406).
These remarks are only sketchy weightings of the properties for the three categories of crime he mentions. To see this, suppose we grant that exemplarity and characteristicalness are the most important properties of offences in the first category consisting of ‘great crimes’. This statement does not tell us, for example, how a certain degree of exemplarity compares in weight with smaller degrees of other properties. Consider the following comparison of two types of punishment for a crime in the first category. The first type has a moderate degree of exemplarity, but none of the other twelve properties. A second type i) has no exemplarity, but ii) is very popular; iii) is subservient to compensation to a high degree; iv) is subservient to reformation to a high degree; and v) has none of the other eight properties. Which punishment is supposed to be more desirable, all things considered? Bentham’s statement does not tell us. This is a problem with Bentham’s general form of evaluation of the types of punishments – that is, the examinations. It arises even if we grant that desirability is itself a property that cannot be quantified, but is instantiated in amounts describable by terms such as ‘very desirable’ or ‘somewhat desirable’.
Now suppose that Bentham had given relatively complete weightings of the twelve properties. The problem that his examinations would still have is that he provides no measures or estimates of the extent to which the desirable properties, or desirable combinations of them, produce the three desirable consequences of punishment that utilitarianism emphasises: prevention (both particular and general), reform and disablement (396–7).
As I noted, Bentham takes general prevention or example to be the principal goal of all punishments (396).127 I will now discuss his treatment in RP of this important process. It is striking that he does not include a property of ‘subserviency to general prevention’ parallel to, say, ‘subserviency to reformation’. It might be thought that the property of exemplarity coincides with the property of serving as an example or subserviency to general prevention, but we saw that this is not correct. Exemplarity, as Bentham defines it, is the misleading painfulness of a punishment; it is not the serving as an ‘example’, in the sense that Bentham uses that term. It is plausible to suppose, then, that Bentham conceived of the role of some of the properties he did discuss, such as characteristicalness, as promoting general prevention. And since he followed Beccaria in assuming that the observation of punishments enhances their general preventive effects, Bentham devised his distinctive concept of exemplarity to facilitate rational thought about punitive pain, so that it is both an effective general preventative and frugal.
However, we need to recognise that Bentham simply had no data in 1780 with which to assess how much general prevention would be produced by a type of punishment having exemplarity and characteristicalness, and so forth. Beccaria was in the same position in 1764. England in 1780 kept no national statistics on the incidence of crimes. Such statistics on crime date to 1805, having been first compiled in 1810.128 Consider therefore the question of whether a punishment such as the iron pen would produce a greater reduction of the incidence of forgery than the then-common punishment for it, namely hanging. To answer the question an investigator would ideally need evidence about the incidence of forgery after similar time periods during which each type of punishment was imposed, and when everything else about the society was the same. But Bentham had no data about the incidence of forgery in England in 1780, let alone a reasonable estimate of what it would have been if the iron pen were the punishment for it, rather than hanging.
In thinking about general prevention, then, Bentham and Beccaria often used their background beliefs about the psychology of potential criminals to design punishments. We saw this sort of approach in Bentham’s discussions of exemplarity, characteristicalness and the reforming power of laborious punishment. It is also apparent in the reasoning he employs to reach the radical conclusion that laborious punishment in prison would be a better general preventative than capital punishment for those offences, such as homicide, in which they were the two options he was considering.
It is instructive to look at Bentham’s reasoning for this striking claim. I should add that he conceded that if it were thought necessary to use capital punishment for the sake of deterrence, it should be reserved for the most atrocious crimes such as murder. However, he indicates that he believes that his reasoning shows that it is not necessary even for such crimes (450). We should keep in mind that when Bentham was writing, England was said to have a ‘Bloody Code’ in which capital punishment was used not only for violent crimes but also for many property crimes, for instance forgery.129 His reasoning about the general preventive effects of capital punishment can be summarised as follows.
Capital punishment lacks the property of equability; that is, being put to death does not produce the same amount of pain, or loss of pleasure, for all the criminals subjected to it. With the ‘higher class’ of criminals who commit the most serious crimes, death will be a great loss to some of them, and little or no loss to others. For some it will even be a benefit. Bentham takes this inequability to constitute a reason against using capital punishment for such offenders. These facts, and some others, thus constitute reasons in favour of punishing them with hard labour in prison. Bentham describes such offenders as indolent and averse to labour, often unhappy, accustomed to risking death and intemperate to such an extent that they exhibit ‘brutal and uncalculating courage’ (445–6, 450).130 For these people, Bentham argues, the ‘contemplation’ of perpetual imprisonment with hard labour ‘would produce a deeper impression’ than ‘even death itself’ (450).131 Thus Bentham, like Beccaria, believed that the most dangerous criminals did not fear death as much as other people did. Bentham believed that they feared being compelled to work for a long time much more.132
Now a long term of hard labour in prison, which was Bentham’s main alternative to capital punishment for serious crimes, was basically untried in England c.1780. Therefore his claims about the psychology of the higher class of criminals might well have seemed speculative to his contemporaries, many of whom believed that capital punishment was an effective general preventative for numerous crimes. In any case, his argument does not seem to apply to criminals who commit non-violent crimes. Of course, given the lack of data, the position of supporters of capital punishment was also speculative; it was probably based on different background assumptions about offenders’ psychology. However, it is fair to say that Bentham did not have better evidence than they did about which of the two types of punishment had greater general preventive power.
The same problem of a paucity of information existed for the other focus of Bentham’s project – that is, determining which acts should be offences. In order to evaluate correctly whether a type of act is mischievous, all things considered, Bentham would need information about all of its effects, both short- and long-term. This would include information about the size of these effects, since sometimes many small, deleterious, long-term effects might make a type of act that is slightly mischievous in the short term into a very mischievous act, all things considered. On the other hand, a type of act might be very beneficial in the short term, and this could outweigh some long-term detrimental effects. In such a case, the act might not be mischievous, all things considered, or only barely so. Bentham was aware of these contrasting possibilities.133
In spite of this fundamental obstacle, Bentham’s investigations sometimes yielded impressive results. One in particular is his discussion of criminalising paederasty, discussed in Part I. We could summarise his use of evidence for the conclusion that it is not mischievous this way: given his basic background assumption about consenting acts between adults, the act itself is beneficial to the participants.134 Given his other background beliefs, he argues that people in the community will not be alarmed by such a consensual act nor put in danger by it. In addressing speculative claims about, for example, its effects on population, he used what evidence he had about ancient Greece and its population to rebut it. He cited two leading authorities, one on population in general and the other on population in the ancient world: Adam Smith and David Hume.135 (It was insightful of Bentham to draw on those two sources in addressing one of the questions he was trying to answer.) Bentham reached a remarkably radical conclusion, using his moral theory, background psychological assumptions and what little historical data and demographic information was available. Even though he had little data on the long-term effects of paederasty, his psychological assumptions are plausible, while his reasoning about its lack of mischievousness was, and still is, convincing, as far it goes.
But in other cases Bentham’s evidence, such as it was, sometimes led him astray. In the same entry in the penal code that discusses paederasty he goes on to assert that masturbation is ‘incontestably pernicious’. He was probably utilising a leading authority, the physician S.A.D. Tissot.136 Turning to RP, we find that when Bentham endorses the reforming potential of short spells of solitary confinement, he states that he is relying on the ‘best authorities’ – meaning Hanway and Howard, who had visited prisons, and talked to gaolers (426–7). In January 1778 Bentham followed their example and visited one of the hulks.137 I therefore agree with Gary Marx’s assertion that Bentham ‘does the best he can with what he has’, meaning, ‘with what evidence he has’.138
V. Collecting data, evaluating consequences
Bentham did not only use the evidence he had. He made efforts to expand the evidence he had, and he attempted to develop the intellectual tools needed to evaluate that information, given the principle of utility. These were yet another part of his programme of applying the principle of utility to penal law. They are the topics of this last section. I will widen my temporal focus here and look at developments in Bentham’s thinking up until 1810.
Bentham was aware of the need for data on criminals, crimes and punishments. This is evident in A View of 1778, where he applauds the draft proposal to collect information about prisoners in the penitentiaries that the bill would establish, and where he also mentions his own prior sketch of a national collection of data on all imprisoned criminals.139 Twenty years later Bentham drafted a bill that included a proposal to collect and publish national data on crimes.140 In 1802 and 1809 Bentham cited information about the earlier experiences in Pennsylvania and Tuscany to argue that eliminating the use of capital punishment would reduce the amount of serious crime, or at least not increase it.141
In England the availability of national crime statistics altered the quality of parliamentary debate on criminal legislation, notably in Samuel Romilly’s well-known speech of 1810. Romilly, who had known Bentham for some twenty-five years, initiated the bill that required compiling these statistics.142 In 1808 Romilly began to lead the renewed effort to replace the capital punishment imposed on many property crimes in the Bloody Code. He started with a statute imposing death on pickpockets.143 In his 1810 speech Romilly apparently drew on Bentham’s thinking.144 He used the newly available statistics on property crimes to argue forcefully for the replacement of two other such statutes.145 In a speech guided by the idea that the goal of punishment is crime reduction, Romilly deployed one of the arguments against capital punishment found in RP – namely that increased severity might paradoxically increase crime, because the moral scruples of prosecutors, jurors and witnesses would lead them to refuse to play their role in the conviction of offenders (449).146 It would follow that decreasing the severity of punishment, all else being equal, would decrease the incidence of the targeted crimes.147
Romilly argued that the statistics suggested that the elimination of capital punishment for pickpockets had had the welcome effect of increasing the number of criminal prosecutions of the crime. However, his hypothesis that moral scruples had played a significant role in victims’ previous decisions not to prosecute offenders was not borne out by later inquiries.148 But the discussion of Bentham’s ‘prevention’ or deterrence had become more sophisticated some thirty years after he wrote the manuscripts underlying RP, and Romilly’s use of statistics to illuminate the process of deterrence was probably partly due to his influence. It is said that the first scientific studies of deterrence were published in the 1950s.149
Bentham also gave thought to the central problem of measuring the value that the consequences of actions and policies have. He never presented a complete quantitative treatment of an issue in penal law, but he did discuss some of the basic issues in constructing one. His important early manuscript, ‘Value’, was written in the late 1770s. It discusses using a person’s choices in two kinds of monetary transactions – paying to have a pleasant experience and paying to avoid a painful one – to give a legislator a reasonably useful proxy or indirect measure of her pleasures and pains.150 RP occasionally touches on such proxy measures of pleasures and pains (468–9: perhaps referring to ‘Value’, 470, 518). IPML does not, although it does clarify the central concepts of a quantity of pleasure and pain, and a quantity of intrinsic value, as well as the nature of mischief.151 Finally, in 1798, Bentham sketched a utilitarian cost/benefit analysis of a possible policing policy – namely establishing a ‘National Watch’ or patrol of the roads outside of towns and cities. This would be established mainly to reduce the number of highway robberies. Here the benefits and costs of the policy are assumed to be measured with units of money.152
Bentham was in no position then to offer a complete evaluation: not only did he lack data on the (probable) crime-reducing effect of such a proposal, but he also had no data on how much taxpayers would pay to avoid highway robberies, or on how much it would cost to operate such a system. In summary, he gave a rough idea of the kinds of benefits and costs a utilitarian evaluation would recognise, but he did not specify how large they would be, nor did he claim to have that information. In this way, the sketch is at the same level of abstraction as his list of the twelve desirable properties of punishment in RP, although it is addressing an issue in enforcement policy, not punishment.
The first rigorous cost/benefit studies of penal policies by economists, which estimate the monetary values of all the relevant factors, have been dated to the 1980s.153 These can be seen as carrying out a programme that Bentham sketched some two centuries before.154 He once wrote that the ‘eyes of statesmen’ were needed to discern, and presumably also to measure and evaluate, the long-term consequences of acts and laws.155 We might now say that what is needed are the skill sets of social scientists, but this is not an amendment to which Bentham would have objected.
Appendix: An overview of the penal code manuscripts
I discuss here a significant portion of all the penal code manuscripts written c.1780. These are in boxes 71 and 72 of the Bentham Papers in the University College London Library, and consist of about 1,300 pages.156 I assess the extent to which the manuscripts reached the goals of Bentham’s project, as they were characterised in RP and IPML. Some of them were hardly reached, but there are a number of very interesting manuscripts that would reward scholarly attention.
The manuscripts have 103 separate ‘titles’ or headings. I will call the set of manuscripts under one title an ‘entry’. There are a number of duplicate or almost duplicate titles, so the number of distinct offences covered is about ninety, although some mention more specific offences falling under the title offence. For example, there is an entry titled ‘Of Simple Personal Injuries’ and another titled ‘Personal Injuries Simple’, both of which cover the same offence.157 Eight of the entries, totalling more than 200 pages, are related to this type of offence, which is called ‘simple corporal injuries’ in IPML.158 Bentham clearly reworked this material many times, and a presumably final version was published in 1843 as Specimen of a Penal Code.159 Some of the entries on this offence are apparently early versions of some of the chapters of IPML on action, intention and motive, using the example of a simple personal injury to illustrate the discussion.160
There are entries that are more than 60 pages in length, while there are others that occupy only part of a page. Chapter 16 of IPML lists many more than 90 possible offences, with 184 mentioned just for Class 1. About 60 of the approximately 90 offences treated in the manuscripts fall into Class 1; like IPML, the penal code delves more deeply into this class of offences. The titles of many, but not all, of the other entries correspond to general categories of offences in the other four classes. As we would expect they go into less detail, because Bentham conceived of the penal code as ‘universal’.161 Several entries are analytical, by which I mean that they are entirely or mainly devoted to characterising the offence, for example by describing the various ways it can be performed, or more specific offences falling under a general title.162 Very few of the entries conform fully to the description of their format contained in Bentham’s letter to Tribolet of 30 March 1779.163 I think that he arrived at this conception of the entries after he had written many of them. It is fair to say that he was not close to completing the code in the format he describes in that letter.
I will briefly run through the main topics that IPML suggested the penal code would address, as well as the occurrence of ideas in RP that I considered in this part of my article.164
Definitions and expositions
Most entries on a single offence contain a careful definition of it. This is often supplemented by an ‘exposition’ that explicates the central terms in it, such as ‘cause’. The definitions of some offences are notably general in their formulation. For example, ‘Simple Personal Injuries’ includes not only a ‘positive’ form, where an act causes a temporary bodily pain, but also a ‘negative’ form, where an individual refrains from helping another person and a pain occurs that could have been avoided.165
Justifications, excuses, extenuations and aggravations
Two important entries discuss general concepts applicable to all offences. One concerns justifications, and covers in more detail almost all of the categories where ‘punishment is groundless’ in IPML.166 The second, ‘extenuations’, goes beyond IPML and would reward scholarly investigation.167 While it discusses the relevant categories ‘where punishment must be inefficacious’ in IPML, unintentional actions, for instance, it also covers further cases. Bentham argues in IPML that if an act is one where punishment must be inefficacious, it should be entirely excluded; in such cases, we now say that the agent should have a legal excuse.168 But in the entry, he also includes under ‘extenuations’ cases where there should be punishment, but less of it. We now often speak of these cases as calling for the mitigation of punishment.169
While the entry assumes that the same set of categories sometimes call for an excuse and sometimes only for mitigation, this cannot be the correct position for Bentham to take. His ‘rationale of excuses’, to use Hart’s Benthamic phrase, is that when certain psychological states exist an agent cannot be deterred, so that any punishment of her would be wasted.170 But when mitigations should be recognised some punishment is useful, and presumably can deter potential offenders, so there must be some other rationale for reducing the amount of punishment. One of the new ideas that we find in the entry is this: when mitigating circumstances existed, the agent exhibited a less ‘depraved’ disposition than the typical offender. This would mean that his offence caused less mischief than is typical.171 Finally, the code discusses aggravating circumstances. These are the various sorts of situations that call for more severe punishments. The idea of such circumstances is only mentioned in IPML; an extensive discussion of them, including the reasons for them, occurs in the entry for ‘Simple Personal Injuries’, which corresponds to Specimen.172
Mischief
Very few entries have complete accounts of all the mischief a type of act usually causes.173 This is surprising, given that IPML Chapter 16 suggests that entries would explain what sorts of mischief each offence causes.174 But there are interesting accounts of the mischief, or lack thereof, in certain special cases, especially where an issue of aggravation or extenuation is being discussed.175
Profitability of punishment
The profitability of punishing a type of act is almost never discussed systematically,176 but the issue comes up in a few entries.177
Reasons
The reasons for some features of an entry – for example, why an act is an offence, or why it is punished in a certain way – is presented in about one-third of the entries, but only about half of those discuss the reasons for most of their main features.178
Punishment
Many, but not all, of the entries treating one offence specify the punishments for it. The entry on homicide, notably, does not.179 Often there are several punishments specified for a given offence and judges are given discretion to choose some or one of them.180 Some thirty offences provide for a characteristical punishment as either a mandatory or an optional part of it, or as its entirety.181 Hard labour is mandated or an option for about fifteen of the more serious offences. Imprisonment is mentioned for at least that many offences, but the entries rarely describe the prison itself.182 Several unusual punishments are named but not described, so it is difficult to determine how often Bentham had deceptive punishments in mind.183 As mentioned above, there are several ‘simple ignominious’ punishments.184 Finally, as to capital punishment, I count six offences where Bentham at least allows it. I mentioned above that being shot out of a cannon was part of the punishment for treason.185 The other examples are described in a note.186 The discrepancy between Bentham’s argument against capital punishment in RP and his occasional acceptance of it in the penal code is another indication that the code, RP and IPML were never completed by Bentham, let alone brought into complete consistency, despite the fact that they were all mainly written c.1780.187
Declarations and conflicts of interest
Research ethics statement
Not applicable to this article.
Consent for publication statement
Not applicable to this article.
Conflicts of interest statement
The author declares no conflicts of interest with this article. All efforts to sufficiently anonymise the author during peer review of this article have been made. The author declares no further conflicts with this article.
Notes
- Sverdlik, ‘Bentham’s Project’, 4–11. ⮭
- Bentham, Rationale of Punishment. There is a modern edition: Bentham, Rationale, ed. McHugh. I usually refer to pages of RP in Works of Bentham, vol. 1, ed. Bowring. References to pages in this edition appear in the text and notes in parentheses. ⮭
- Bentham, IPML. ⮭
- Sverdlik, ‘Bentham’s Project’, 12–3. ⮭
- RP occasionally refers to a penal code, although this was probably a generic idea of it (517, 519). ⮭
- There is very little scholarship on RP. The two important discussions of Bentham’s philosophy of punishment by H.L.A. Hart, namely ‘Bentham and Beccaria’ and ‘Bentham’s Principle’, rarely mention it. Two helpful studies of portions of it are Bedau, ‘Bentham’s Utilitarian Critique’, 1041–8 and Semple, Bentham’s Prison, 24–41. ⮭
- Hart, ‘Bentham’s Principle’, lxxix. For the birth of IPML, see Sverdlik, ‘Bentham’s Project’, 3, 22 and nn. 7 and 90. ⮭
- In a letter of April/May 1778 Bentham states that he began work on it a year and a half earlier; in September 1778 he reports that he is taking more pleasure in writing the penal code. It is likely he put it aside when he began working on the penal code. He may have written some passages later: in one, he describes the behaviour of convicts hanged at Newgate Prison (447). Executions did not take place there until late 1783 or early 1784. Bentham’s title for the manuscripts that underlie RP was then The Theory of Punishment (or Punishments): see Correspondence, ii. 100, 169. ⮭
- Bentham, Rationale. ⮭
- In August 1808 Dumont showed Bentham’s manuscripts to Samuel Romilly, an important figure in the reform of English criminal law (see below). Romilly was impressed and urged Dumont to complete his translation of them. Radzinowicz, History, 322. ⮭
- Bentham, Théorie. ⮭
- [Brougham], Review of Théorie des Peines. This was originally published anonymously, but Brougham later acknowledged his authorship. Brougham, Contributions. ⮭
- This chapter discusses the punishment of ‘transportation’. Dumont stated that in producing this chapter he relied on Bentham’s work, Letters to Lord Pelham (389). The first letter was published in 1802, see Letters to Lord Pelham, 269–311. However, RP also features a discussion of ‘banishment’, presumably written in the late 1770s since it focuses on earlier European practices (432–5). ⮭
- Panopticon. ⮭
- Rationale, ed. McHugh, 13. ⮭
- This was Hugo Bedau’s conclusion: ‘Bentham’s Theory’, 1. His judgement seems to be supported by the recently published penultimate volume of Bentham’s correspondence, covering July 1828 to June 1832. Correspondence, xiii. This contains no letter mentioning RP, but Smith and Bentham may have communicated in person. ⮭
- Bedau suggests that a significant part of RP might not contain Bentham’s own words. ‘Bentham’s Theory’, 2, 6. This is clearly true of the added chapters, and of some other parts, but it might not be true of the bulk of it. ⮭
- Of Indirect Means; Of the Limits; Of the Promulgation of the Laws; Promulgation of the Reasons of the Laws; A View; Place and Time; Defence of Usury. Specimen of a Penal Code contains one entry of the penal code. ⮭
- Ignatieff, Just Measure of Pain, 79. For a short current survey of English penal practice in that era, see Emsley, Crime and Society, 258–83. For the history of imprisonment in England up to 1791, see Webb and Webb, English Prisons, 1–65. Ignatieff, Just Measure, 80–113 gives his account of imprisonment from 1750–1820; for background, 11–79. ⮭
- A View. See the helpful discussion in Semple, Bentham’s Prison, 42–59. ⮭
- Bentham implicitly describes A View as applying the theoretical ideas of RP to the proposed penitentiary system in the bill: A View, 3. ⮭
- Semple, Bentham’s Prison, 44. ⮭
- Book I, Ch. 4 of RP (397–8) is an earlier version of Chapter 13 of IPML; Book I, Ch. 6 of RP (399–402) is an earlier version of Chapter 14 of IPML; Book I, Ch. 7 of RP (402–6) is an earlier version of Chapter 15 of IPML. ⮭
- In a letter Bentham describes chapters of his book on punishment dealing with ‘pains and pleasures’ and ‘circumstances influencing sensibility’. Correspondence, vol. 2, 127. These probably became Chapters 5 and 6 of IPML. ⮭
- Notably, in IPML Chapter 16 Bentham states that it is not even possible to fashion a retaliatory punishment for three of the five classes of offence described in that chapter. IPML, 275–8. It is in RP that arguments are given for these assertions (410–11). ⮭
- IPML, 158, n. a, describing it as The Theory of Punishment. It refers to what is now RP, 396–7. The starred note of 1823 at IPML, 158 refers to Dumont’s recension; it states that an English translation of it, also utilising ‘the Author’s manuscripts’, is contemplated. There is a less explicit reference to material apparently intended to be included in RP at IPML, 71–2, n. r. ⮭
- The 1830 edition has nine occurrences of the phrase in 441 pages. The first occurs on p. 37. Bentham, Rationale. The lengthy and careful review of Théorie des Peines in The Edinburgh Review in 1813, which only discusses the part treating punishment, never mentions the principle of utility; the reviewer (Brougham) seems to have missed its significance. [Brougham], Review. ⮭
- Nonetheless RP Book I, Ch. 6 (an earlier version of IPML, Ch. 14) discusses how to deter potential criminals effectively and frugally. IPML in effect assumes that a good deal of psychology, philosophy and moral theory is needed to address this issue. ⮭
- IPML, 158. ⮭
- Sverdlik, Guide, 197–241. ⮭
- A condition is a relatively permanent legal status that confers certain rights and powers on an individual. Examples are husband and wife, and parent and child (470–4). Cf. IPML, 234–70. ⮭
- The references there to the ‘former’ work of IPML must have been inserted by Bentham’s editors, as RP was written first. ⮭
- A movement to improve the hygiene of institutions such as prisons and hospitals began about 1750. See Ignatieff, Just Measure, 44–6; 59–62. Cleanliness in prisons was endorsed by Hanway, Defects, 216 and Howard, State of the Prisons, 58–60. Under Howard’s influence an act of Parliament was passed in 1774; among other measures, it required prisons to be cleaned periodically. However, this requirement was not fully complied with until about 1789. See Webb and Webb, English Prisons, 38, 53. ⮭
- IPML, 185. ⮭
- This is implied by Bentham’s statement that laborious punishment is closer to perfection than any other single type (439–40), and his claim that it must always be combined with some sort of imprisonment, since offenders must be housed and fed. Bentham had sketched his ideas about the best species of prisons earlier in RP (429–31). Bentham first argued in print for the desirability of prisons in which labour would be performed in A View. There, however, he states that he is applying the ideas in ‘a work of some bulk’ that he was completing. We can now see that he is referring to the manuscripts on which RP is based. A View, 3. ⮭
- IPML, 159, 179–80; Sverdlik, Guide, 206–13, 215, 232–6. ⮭
- Sverdlik, Guide, 215–36. ⮭
- IPML, 158, n. a. ⮭
- IPML, 158, n. a, 288. ⮭
- Cf. IPML, 159, n. a, 288. ⮭
- IPML, 166–8, 181–2, 180–1. ⮭
- IPML, 175–86. It does not discuss ‘simplicity of description’ (405–6). ⮭
- The type of value that they have is instrumental. Sverdlik, Guide, 24–7, 80–2. ⮭
- These are all discussed in Sverdlik, Guide, 242–54. ⮭
- IPML, 177–8. ⮭
- IPML, 182–3. ⮭
- Beccaria warned legislators never to make punishments so severe that compassion for the criminal becomes the dominant emotion in observers. Beccaria, On Crimes, trans. Young, 49. I generally cite this excellent scholarly edition. ⮭
- Exemplarity is the other property; see below. ⮭
- Beccaria, On Crimes, trans. Young, 36–7. Bentham mentions Montesquieu in this regard, but adds that his idea of characteristicalness is ‘very indistinct’. IPML, 178. Montesquieu, Spirit, 189–91. ⮭
- Beccaria, On Crimes, trans. Young, 37. ⮭
- IPML, 178. ⮭
- Blackstone, Commentaries, 8–9. Blackstone’s thinking was clearly influenced here and elsewhere in volume 4 by Beccaria, whom he often cites. ⮭
- Public offences constitute the fourth class of offences in IPML: see 189–90, 196–203, 260, n. r4, 278–9. This argument fills in a gap at IPML, 278. See n. 25 above. ⮭
- IPML, 83. ⮭
- IPML, 141. ⮭
- IPML, 95. ⮭
- Blackstone had presented a brief version of this criticism: Commentaries, 8. ⮭
- UC lxxi. 61v1. ⮭
- UC lxxi. 197r1, UC lxxi. 46v1–r2. ⮭
- Cf. IPML, 168, n. i. The penal code states that burning an offender over a fire for arson was to occur ‘once or oftener, and for a longer or a shorter time, according to the amount of damage’. However, the value of the damage caused had to be greater than £100. If the aggravating circumstance of the death of two or more persons occurred, the offender was to be burned to death. UC lxxi. 141r1–2r1. Being burned for arson that damages property alone would be characteristical but not retaliatory. UC lxxi. 46v1–r2. The burning to death of an offender for burning people is both characteristical and retaliatory. ⮭
- Letters to Lord Pelham, 74. ⮭
- Hart, ‘Bentham and Beccaria’, 47. Hart was inaccurate in suggesting that it is only modern taste that is repelled by characteristical punishments. Objection was made to them soon after Bentham’s works were published in France. See [Brougham], Review, 10–11. ⮭
- Hart, ‘Bentham and Beccaria’, 47. ⮭
- After 1729 most types of forgery were to be punished by hanging. See McGowen, ‘From the Pillory’. A markedly high percentage of convictions resulted in execution; that is, pardons were rare. Emsley, Crime and Society, 264–5, 269–70. ⮭
- For high treason the punishment for male convicts originally consisted of dragging him to the gallows (though this was later relaxed by putting him on a sled); hanging him; cutting him down while alive; disemboweling him; burning his intestines in his presence; cutting his head off; and dividing his body into four parts. The king had the right to dispose of the head and body parts as he saw fit. Blackstone, Commentaries, 61. ⮭
- Cf. IPML, 147, n. g, 158, n. a. ⮭
- Cf. IPML, 179. But Bentham sometimes uses ‘apparent’ to refer to mental states such as thought, belief and memory, rather than perception (398, 399). Sverdlik, Guide, 2, 8–9, 167. ⮭
- Ignatieff, Just Measure, 31, 34–5, 37. Prisoners were often prevented from escaping by chaining them to the wall or floor: 34–5. ⮭
- Howard, State of the Prisons, 38–77. ⮭
- Semple, Bentham’s Prison, 50. ⮭
- Semple, Bentham’s Prison, 73. ⮭
- Panopticon, 46. Bentham’s emphasis. But he conceded in correspondence that there would be a need to search and question every visitor, and turn away those who are not ‘decently clad’. Semple, Bentham’s Prison, 143. He also eventually proposed to protect the prison with a system ‘of outer walls, palisades, and ditches constantly patrolled by sentinels [who] would ensure that the prison could be approached only along a walled avenue’. And in a departure from English practice, Bentham proposed that regular soldiers guard the prison. Semple, Bentham’s Prison, 120–1. ⮭
- Cf. IPML, 159, n. a, 288. ⮭
- IPML, 159, n. a. ⮭
- Cf. Letters to Lord Pelham, 75, which is less derogatory. ⮭
- Banishment – the term Bentham uses in this earlier part of RP – has no exemplarity, he says, because the sufferings of the convict are unobserved by his countrymen (434–5). Here, however, Bentham takes ‘exemplarity’ to mean ‘serving as an example’, that is, acting as a general preventive. But this passage also emphasises the importance of the observation of the punishment (cf. 415–16). ⮭
- Beccaria, On Crimes, trans. Young, 99, n. 4. Beccaria wrote that doing this ‘gives an example to societies [in] which he never offended; an example absolutely useless, because distant from the place where the crime was committed’. Quoted in Schofield, ‘“First Steps”’, 73. The passage is translated slightly differently in On Crimes, trans. Young, 55. More generally, Beccaria argued for the usefulness of public punishments. On Crimes, trans. Young, 49, 81. ⮭
- On Crimes, trans. Young, 7. ⮭
- Beccaria eventually came to reject public punishments. This was chiefly because he thought they are inherently degrading. On Crimes, trans. Young, 94. ⮭
- IPML, 178–9. See also Bentham, Principles of Judicial Procedure, 21, on the contrast of real and apparent justice: ‘From apparent justice flow all the good effects of real justice – from real justice, if different from apparent, none.’ ⮭
- Bentham’s emphases. The contrast between real and apparent painfulness comes up a few other times in RP at 408, 431 and 442. The contrast made on 430, which characterises the Black Prison (431), is startling. IPML also uses the contrast between the pain of an offender undergoing a punishment and the experience of other potential offenders who might be deterred by it. IPML 184–5. Here, however, Bentham is contrasting undergoing a punishment and thinking about it, as opposed to undergoing it and seeing it. ⮭
- UC lxxi. 61v1. ⮭
- UC lxxi. 61v1. ⮭
- Cf. IPML, 179–80; Sverdlik, Guide, 232–6, 250–1. ⮭
- See Promulgation of the Reasons of the Laws, 159–63 and Sverdlik, ‘Bentham’s Project’, 15–17. Cf. the drawbacks that John Rawls mentions of a putatively utilitarian public institution of ‘telishment’: ‘Two Concepts’, 11–12. In RP Bentham notes that it would be absurd to retaliate to an offence which involved making false statements about a victim by publishing false statements about the offender. If it were announced that a certain statement about her is false, Bentham says, it would not injure her (410). ⮭
- UC lxxi. 101v1, UC lxxi. 189v2–90r1. ⮭
- Shoemaker, ‘Streets of Shame?’ On public hanging, see Gatrell, Hanging Tree, esp. 225–444, 589–611. ⮭
- Shoemaker, ‘Streets of Shame?’, 245–7. ⮭
- Shoemaker, ‘Streets of Shame?’, 235, 241. ⮭
- However, this would not be exemplarity in the trompe l’oeil sense that there is in Bentham’s example of the tongue spike. Another such example is mentioned in Section 3. ⮭
- Previous writers had noted that punishments such as branding and whipping, which left observable stigma, tended to make criminals infamous; they thus became unemployable and desperate. Eden, Principles, 52. See also Hanway, Defects, 3; cf. 213. Eden’s book was popular, and probably the most influential work on penal theory in England in the 1770s. Bentham cites the book three times in RP, without mentioning Eden’s name (459, 465, 466). Bentham himself had a low opinion of the book, but he worked with Eden on the Hard Labour Bill in 1778. Semple, Bentham’s Prison, 58–9. For Eden’s role in reforming criminal law see Radzinowicz, History, 301–13; Devereaux, ‘Making of the Penitentiary Act’. ⮭
- Commensurability is never mentioned in these chapters. This makes sense, because it is a property of sets of punishments, not of any one of them. Sverdlik, Guide, 244–7. ⮭
- In Book V the punishment of transportation is examined carefully, as is the panopticon plan. However, as I observed, these chapters must have been written after 1778 by Dumont and Smith, although they relied on material written by Bentham. ⮭
- Bedau, ‘Bentham’s Utilitarian Critique’, 986–1002. ⮭
- Hanway, Defects, Letter 22, 210–35, focuses on prisons. Bentham had a high opinion of Hanway and cites the book twice in RP, without giving its title (426–7). See Semple, Bentham’s Prison, 83–9. ⮭
- This was a form of typhoid apparently unknown in other countries. Webb and Webb, English Prisons, 19–20. For Howard’s career, see Ignatieff, Just Measure, 47–59. For his relations with Bentham, see Semple, Bentham’s Prison, 71–8, 89–90, 92–4. ⮭
- Howard is often cited in A View. ⮭
- Howard, State of the Prisons, 38–77. ⮭
- Correspondence, ii. 105–7, at 106. ⮭
- IPML, 181. ⮭
- It began with an act passed in 1706. See Beattie, Policing and Punishment, 331–5. ⮭
- Howard stated that felons should not be required to work, but he was sympathetic to the idea that they be allowed to do work that earned them a more comfortable time in prison. Howard, State of the Prisons, 47. Hanway proposed encouraging prisoners to work, allowing them to profit from it and even training them in trades that could lead to employment on release. Defects, 216, 228–9, 233. ⮭
- Cf. Hanway, Defects, 210–35, esp. 225. ⮭
- Panopticon, 37–58. ⮭
- However, subserviency to compensation is simply defined as convertibility to profit (405). ⮭
- The basic idea was not new. For example, in Thomas More’s Utopia the standard punishment for major crimes is slavery, not death, on the basis that ‘live workers are more valuable than dead ones’. More, Utopia, 104–5. Cf. Voltaire, Commentary on Beccaria, 146. ⮭
- IPML, 179. ⮭
- A View, 12–13. The Penitentiary Act of 1779 permitted such payments to be made in its planned national prisons. Semple, Bentham’s Prison, 47. For the history of such schemes in local jails, see Webb and Webb, English Prisons, 82–9. ⮭
- IPML, 180–1. ⮭
- IPML, 180. ⮭
- Sverdlik, Guide, 251–2. ⮭
- Hanway, Defects, 221, seems clearly to describe reform, as distinct from particular prevention. Cf. Howard, State of the Prisons, 43, on how solitude can lead a person to reflection and repentance. ⮭
- This, too, was a claim that had been advanced before. Voltaire, Commentary on Beccaria, 146. ⮭
- IPML, 105, 111–4, 181. ⮭
- IPML, 156, n. q, 284. ⮭
- IPML, 96–142; Sverdlik, Guide, 140–77. ⮭
- IPML, 134–5; Sverdlik, Guide, 171–2. ⮭
- IPML, 116; Sverdlik, Guide, 156–60. ⮭
- For a thoughtful contemporary defence of a version of the idea that punishment should aim to reform offenders, see Duff, Punishment, esp. 91–125. Duff prefers to describe his theory as ‘communicative’ rather than ‘reformative’. On this putative distinction, see Sverdlik, ‘Punishment’. ⮭
- UC lxxi. 38r2, UC lxxi. 45r1. For both offences, other components of the punishment were to be characteristical. ⮭
- Letters to Lord Pelham, 74. The chapter in RP that discusses the panopticon, which was based on Letters, according to Dumont, likewise does not mention characteristicalness (498–503. Cf. 389). ⮭
- Devereaux, ‘Penitentiary Act’, 415–6, mentions some MPs who took this position in a debate about one of the ‘Hard Labour’ bills. ⮭
- See RP, 411 for two others. ⮭
- IPML, 21–31, 182–4. Cf. 101–2; also, Sverdlik, Guide, 34–9, 153–4, 253–4. ⮭
- Crompton, ‘Paederasty. Part 2’, 97, 106. Sverdlik, ‘Bentham’s Project’, 29. ⮭
- See Bedau, ‘Bentham’s Utilitarian Critique’, 1008. I generalise the point he makes about Bentham’s argument against the death penalty. ⮭
- IPML, 179, 185, 288. ⮭
- Emsley, Crime and Society, 21–2. ⮭
- On the ‘Bloody Code’, see Emsley, Crime and Society, 258–75. ⮭
- Given Bentham’s description of this class of people, he devotes some space to explaining why they do not simply commit suicide (446–7). See also his impression of the ‘the courage or brutal insensibility’ of criminals on the scaffold at Newgate Prison (447). A different picture of scaffold behaviour can be found in Gatrell, Hanging Tree, 29–40. ⮭
- By ‘contemplation’ Bentham may mean ‘thought’ or ‘consideration’ and not ‘observation’ of punished convicts. ⮭
- Beccaria stated that at least when such dangerous characters observed convicts doing hard labour every day in public – and being thus denied the ‘freedom and friendship’ with other citizens that they previously enjoyed – these men will decline to take their chances on, for example, robbing the wealthy. Beccaria, On Crimes, trans. Young, 50–1. ⮭
- For the first type of case see IPML, 149–51; Sverdlik, Guide, 186–90. For the second, Crompton, ‘Paederasty. Part 1’; Crompton, ‘Paederasty. Part 2’; Sverdlik, ‘Bentham’s Project’, 21–31. ⮭
- For Bentham’s understanding of the age of consent, see Schofield, ‘Bentham on adult-child sex’, 514–17. ⮭
- Smith, Wealth of Nations, 97–8 (I, viii), 180–2 (I, xi. 2); Hume, ‘Populousness’, 305, 309. ⮭
- Laqueur, Solitary Sex, 37–41, 212–3. ⮭
- Semple, Bentham’s Prison, 55–6. Bentham also visited the Cold Bath Fields Prison in July 1799, accompanied by the magistrate Patrick Colquhoun. Bentham, Picture of the Treasury, § 8, F. 3 (UC cxxi. 64). ⮭
- Marx, ‘Bentham on Social Control’, 201. ⮭
- A View, 29. Cf. Hanway, Defects, 211. ⮭
- Bentham, ‘Extracts’, 80, 91–102, 121–41. See also Quinn, ‘Bentham on Preventive Police’, 51–67. ⮭
- Letters to Lord Pelham, 238–9; Crimmins, ‘Strictures on Paley’s Net’, 28. See Radzinowicz, History, 291–3, 300, for the Tuscan experience. ⮭
- Radzinowicz, History, 395, n. 51. ⮭
- Radzinowicz, History, 497–501. This bill passed. For Romilly’s career, see 313–36; 501–25. For the earlier, and largely unsuccessful efforts at reforming English criminal law in the 1770s and 1780s, in which Romilly played a part, see 399–493. ⮭
- For Romilly and RP, see n. 10 above; for Romilly’s praise of Bentham, see Speeches of Romilly, 342. ⮭
- Speeches of Romilly, 106–94. The two bills were not passed. ⮭
- The worry about the effects of such scruples dates back at least to Blackstone’s description of the ‘pious perjury’ of English jurors who underestimated the value of stolen goods to avoid imposing death sentences: Blackstone, Commentaries, 158. ⮭
- Speeches of Romilly, 113, 115, 126–7. Romilly’s postscript to the published speech contains interesting arguments about the significance of the available statistics, 187–93. See also Radzinowicz, History, 501–3. ⮭
- Emsley, Crime and Society, 196–8. ⮭
- Tullock, ‘Does Punishment Deter Crime?’, 103. There is a discussion of many of the early studies in Zimring and Hawkins, Deterrence, 249–338. ⮭
- Bentham’s ‘Value’ is found in Baumgardt, Bentham and the Ethics of Today, 558–62. See also, Quinn, ‘Bentham on Mensuration’, 77–89. ⮭
- IPML, 38–41, 143–57, 187–91. Sverdlik, Guide, 70–85, 178–96, 260–2. ⮭
- Bentham, ‘Extracts’, 121–5. ⮭
- Cohen, Costs of Crime, pp. xi–xiv, 7–8, 90–105. This is an excellent introduction to the subject. ⮭
- As noted by, for instance, Daniel Nagin, ‘Utilitarianism and Policing’. Some economists do not accept all of Bentham’s assumptions – for example, that the pain of a punished criminal is in itself a social cost, and the pleasure a criminal gets from committing a crime is in itself a social benefit. Cf. Cohen, Costs of Crime, 30; RP, 398; Sverdlik, Guide, 31–4. ⮭
- IPML, 152. ⮭
- The reader can start examining those manuscripts at this webpage: http://transcribe-bentham.ucl.ac.uk/td/Penal_Code. ⮭
- UC lxxi. 18r1–26v1; UC lxxii. 116r1–22v2. ⮭
- IPML, 223–4. ⮭
- It had been edited and translated into French by Dumont in 1802. Bentham, Traités. ⮭
- ‘Personal Injuries Simple’, UC lxxii. 123r1–30v1; ‘Personal Injuries’, UC lxxii. 130r2–41v2. ⮭
- See, for instance, ‘Offences against the External Security of the State’, UC lxxii. 47r1–53v2; ‘Offences against the National Interest in General’, UC lxxii. 54r1–v2. See Sverdlik, ‘Bentham’s Project’, 12–13. ⮭
- See, for instance, ‘Offences against Genealogy’, UC lxxi. 114r1–15v1; ‘Of Semi-public Offences in General’, UC lxxi. 128r1–36v2. ⮭
- Bentham to Tribolet, 30 March 1779, Correspondence, ii. 248–51. ⮭
- Sverdlik, ‘Bentham’s Project’, 10–11 draws on IPML and the letter to Tribolet. ⮭
- Specimen, 164. Cf. ‘positive’ and ‘negative’ falsehoods. UC lxxi. 176v1. ⮭
- IPML, 159–60, which refers to this section of the code. UC lxxi. 1r1–7v1; Sverdlik, Guide, 198–201. ⮭
- IPML, 83, 95, 191, n. g; UC lxxi. 8r1–17r1. ⮭
- IPML, 160–2; Sverdlik, Guide, 202–6. ⮭
- Other entries mention other extenuations, meaning mitigations. See, for instance, UC lxxi. 45r2–v2; UC lxxi. 141v2. ⮭
- Hart, ‘Prolegomenon’, 17–20. ⮭
- UC lxxi. 15r1–v1; UC lxxi. 15v4; UC lxxi. 16v1. ⮭
- IPML, 83, 95, 191, n. g; UC lxxi. 18r1–26v1; Specimen, 164–5, 167–8. Other entries mention other aggravations. See, for instance, UC lxxi. 46r1–v2, UC lxxi. 118v1. ⮭
- A few notable examples: ‘Paederasty’ in Crompton, ‘Paederasty. Part 1’; Crompton, ‘Pederasty. Part 2’; ‘Adultery’, UC lxxi. 90r1–104r2, at 98r1–100v2; ‘Frauds Relative to the Coin’, UC lxxi. 160r1–86v2, at 166r1–183r1. The last is a remarkably penetrating entry. ⮭
- Sverdlik, ‘Bentham’s Project’, 9–10. ⮭
- See, for instance, UC lxxi. 66r2, 67r1–v1; UC lxxii. 173r1–4v1. ⮭
- For two possible offences where punishment would be unprofitable, see Crompton, ‘Pederasty. Part 2’, 97–100; ‘Prodigality’, UC lxxii. 62r1–v1. ⮭
- See, for instance, UC lxxi. 112v1; UC lxxi. 154v1–5r1; UC lxxi. 204r1–r2. ⮭
- According to Richard Smith, Specimen was intended to ‘show the use of a commentary of reasons’. Specimen, 164, n. ‘Theft’ also presents the reasons for a number of its features in the form of a catechism. UC lxxi. 47r2–51r2. ⮭
- UC lxxii. 145r1–50r2. ⮭
- ‘Theft’, which covers a range of acts varying in their mischievousness, is a good example. UC lxxi. 45r1–v1. Its catechism defends the various punishment options. UC lxxi. 47r2–51r2. Cf. IPML 185 on ‘complex’ punishments. ⮭
- Besides the characteristical punishments discussed in Section 2, I will mention these two severe examples: ‘the extortioner’s press’ (a device to squeeze the body of an extortionist): UC lxxi. 35v1; ‘riding the iron horse’ for robbery on horseback: UC lxxi 38r1. This was probably a variant of the torture device called the ‘wooden horse’. See ‘Wooden Horse’. Wikipedia. However, many characteristical punishments are milder. For instance, for a simple personal injury, where the aggravating circumstance is that the victim is female, part of the punishment could be a ‘characteristic penance’: Specimen, 164. In the manuscripts this is specified as ‘a woman’s cap on his head’, but this is crossed out: UC lxxi. 19r1. In other words, the characteristical component here was to be some sort of ‘simple ignominious’ punishment. Sometimes the manuscripts merely say that an unspecified ‘characteristic’ punishment is to be a part of the sanction: for instance ‘Theft’, UC lxxi. 45r1. ⮭
- However, ‘Theft’ mentions ‘penitential imprisonment in the black Dungeon’: UC lxxi. 45r1. This seems to mean the place in the Penitentiary House where temporary solitary confinement would occur (425–7). ⮭
- Two punishments in the code – namely the tongue spike and the iron pen – are described in RP, so we know they were to be deceptive: see Section 2. But were the ‘deceiver’s pallet’ (UC lxxi. 61v1) or ‘the house-breaker’s cage’ (UC lxxi. 39r1) also intended to deceive? ⮭
- See, for instance, UC lxxi. 101v1; UC lxxi. 189v2–90r1. ⮭
- UC lxxi. 197r1. ⮭
- The punishment for incendiarism causing two or more deaths was being burned to death: UC lxxi. 141v2–2r1. Cf. IPML, 168, n. i. For criminal inundation (that is, flooding) causing ten or more deaths the punishment was ‘water torture’, that is, forcing water into the offender’s mouth, until he bursts: UC lxxi. 144r1–v1. For ‘offensive rebellion’, the judge could give a death sentence, but the offender could choose the form of execution: UC lxxi. 197r1. When a state official took undisclosed payments from a foreign government, capital punishment could be imposed: UC lxxii. 54v1–r2. For poisoning a well the punishment was drenching and then death by poison, with dissection to follow: UC lxxii. 136v2. The entry on homicide (UC lxxii. 145r1–50r2) does not specify any punishments, but Bentham may well have thought that some aggravated forms of it should be punished by death. ⮭
- I am indebted for their help and suggestions to Justin Fisher, Philippe Chuard, Chase Kurth, Peter Chau, Xiaobo Zhai, Jean Kazez, Eric Barnes, Doug Ehring, Dean Thomas di Piero of SMU, Chris Riley, Becca Marin, Catherine Bradley and an anonymous reviewer. ⮭
Bibliography
Beattie, J.M. 2001. Policing and Punishment in London, 1660–1750. Oxford: Oxford University Press.
Beccaria, Cesare. 1986 [1764]. On Crimes and Punishments, trans. David Young. Indianapolis, IN: Hackett.
Bedau, Hugo. 1983. ‘Bentham’s Utilitarian Critique of the Death Penalty’, Journal of Criminal Law and Criminology, vol. 74, no. 3, 1033–65.
Bedau, Hugo. 2004. ‘Bentham’s Theory of Punishment: Origin and Content’, Journal of Bentham Studies, vol. 7, no. 1, 1–15.
Bentham, Jeremy. 1778ff. Penal Code, volunteer transcripts: http://transcribe-bentham.ucl.ac.uk/td/, UC lxxi–lxxii.
Bentham, Jeremy. c. 1778. ‘Value’, in David Baumgardt, Bentham and the Ethics of Today. New York: Octagon Books, 1966, 554–66.
Bentham, Jeremy. 1778. A View of the Hard-Labour Bill, in The Works of Jeremy Bentham, vol. 4, ed. John Bowring, 11 vols. Edinburgh: William Tait, 1843, 3–34.
Bentham, Jeremy. 1791. ‘Panopticon; or, The Inspection House’, in The Works of Jeremy Bentham, vol. 4, ed. John Bowring, 11 vols. Edinburgh: William Tait, 1843, 37–58.
Bentham, Jeremy. 1798. ‘Extracts from Jeremy Bentham’s Board of Police Bill’, in Jeremy Bentham on Police, eds. S. Jacques and P. Schofield. London: UCL Press, 77–141.
Bentham, Jeremy. 1802. A Picture of the Treasury, pre-publication text. London: Bentham Project, 2025: https://www.ucl.ac.uk/bentham-project/sites/bentham_project/files/pre-publication_text_a_picture_of_the_treasury_april_2025.pdf.
Bentham, Jeremy. 1802. Traités de Législation Civile et Pénale, vol. 3, trans. Etienne Dumont. Paris: Bossange, Masson et Besson, 302–21.
Bentham, Jeremy. 1811. Théorie des Peines et des Récompenses, vol. 1, trans. Etienne Dumont. London: Vogel and Schulze.
Bentham, Jeremy. 1830. The Rationale of Punishment. London: Robert Heward.
Bentham, Jeremy. 1843. Of Indirect Means of Preventing Crime, in The Works of Jeremy Bentham, vol. 1, ed. John Bowring, 11 vols. Edinburgh: William Tait, 533–80.
Bentham, Jeremy. 1843. Of the Promulgation of the Laws, in The Works of Jeremy Bentham, vol. 1, ed. John Bowring, 11 vols. Edinburgh: William Tait, 157–9.
Bentham, Jeremy. 1843. Of the Promulgation of the Reasons of the Laws, in The Works of Jeremy Bentham, vol. 1, ed. John Bowring, 11 vols. Edinburgh: William Tait, 159–63.
Bentham, Jeremy. 1843. Principles of Judicial Procedure, in The Works of Jeremy Bentham, vol. 2, ed. John Bowring, 11 vols. Edinburgh: William Tait, 1–188.
Bentham, Jeremy. 1843. The Rationale of Punishment, in The Works of Jeremy Bentham, vol. 1, ed. John Bowring, 11 vols. Edinburgh: William Tait, 388–532. (Pages in RP cited from this edition.)
Bentham, Jeremy. 1843. Specimen of a Penal Code, in The Works of Jeremy Bentham, vol. 1, ed. John Bowring, 11 vols. Edinburgh: William Tait, 164–8.
Bentham, Jeremy. 1843. The Works of Jeremy Bentham, ed. John Bowring, 11 vols. Edinburgh: William Tait.
Bentham, Jeremy. 1996. An Introduction to the Principles of Morals and Legislation, eds. J.H. Burns and H.L.A. Hart. Oxford: Oxford University Press (The Collected Works of Jeremy Bentham).
Bentham, Jeremy. 2009. The Rationale of Punishment, ed. James McHugh. Amherst, NY: Prometheus Books.
Bentham, Jeremy. 2010. Of the Limits of the Penal Branch of Jurisprudence, ed. Philip Schofield. Oxford: Oxford University Press (The Collected Works of Jeremy Bentham).
Bentham, Jeremy. 2011. Place and Time, eds. S. Engelmann and P. Schofield, in Jeremy Bentham: Selected Writings, ed. S. Engelmann. New Haven, CT: Yale University Press, 152–219.
Bentham, Jeremy. 2016. Defence of Usury, in Jeremy Bentham, Writings on Political Economy, vol. 1, ed. Michael Quinn. Oxford: Oxford University Press (The Collected Works of Jeremy Bentham), 44–113.
Bentham, Jeremy. 2017. The Correspondence of Jeremy Bentham, vol. 2, ed. T.L.S. Sprigge. London: UCL Press (The Collected Works of Jeremy Bentham).
Bentham, Jeremy. 2023. Letters to Lord Pelham, in Panopticon versus New South Wales and Other Writings on Australia, eds. T. Causer and P. Schofield. London: UCL Press (The Collected Works of Jeremy Bentham), 269–311.
Bentham, Jeremy. 2024. The Correspondence of Jeremy Bentham, vol. 13, eds. P. Schofield, T. Causer and C. Riley. London: UCL Press (The Collected Works of Jeremy Bentham).
Blackstone, William. 2016 [1769]. Commentaries on the Laws of England, vol. 4, ed. Ruth Paley. Oxford: Oxford University Press.
[Brougham, Henry]. 1813. Review of Bentham, Théorie des Peines et des Recompenses, Edinburgh Review, vol. 22, no. 43, 1–31.
Brougham, Henry. 1856. Contributions to the Edinburgh Review, vol. 3. London: Richard Griffin, 112–49.
Cohen, Mark. 2005. The Costs of Crime and Justice. London: Routledge.
Crimmins, J.E. 1987. ‘Strictures on Paley’s Net: Capital Punishment and the Power to Pardon’, The Bentham Newsletter, no. 11, 23–34.
Crompton, Lewis. 1978. ‘Offences against One’s Self: Paederasty. Part 1’, Journal of Homosexuality, vol. 3, no. 4, 389–405.
Crompton, Lewis. 1978. ‘Offences against One’s Self: Paederasty. Part 2’, Journal of Homosexuality, vol. 4, no. 1, 91–107.
Devereaux, Simon. 1999. ‘The Making of the Penitentiary Act, 1775–1779’, The Historical Journal, vol. 42, no. 2, 405–33.
Duff, R.A. 2001. Punishment, Communication, and Community. Oxford: Oxford University Press.
Eden, William. 1771. Principles of Penal Law. London: B. White and T. Cadell.
Emsley, Clive. 2018. Crime and Society in England 1750–1900, 5th edn. London: Routledge.
Gatrell, V.A.C. 1996. The Hanging Tree. Oxford: Oxford University Press.
Hanway, Jonas. 1775. The Defects of Police. London: J. Dodsley and Brotherton and Sewell.
Hart, H.L.A. 1968. ‘Prolegomenon to the Principles of Punishment’, in Punishment and Responsibility. Oxford: Oxford University Press, 1–27.
Hart, H.L.A. 1982. ‘Bentham and Beccaria’, in Essays on Bentham. Oxford: Clarendon Press, 40–52.
Hart, H.L.A. 1996. ‘Bentham’s Principle of Utility and Theory of Penal Law’, in IPML, pp. lxxix–cxii.
Howard, John. 1777. The State of the Prisons in England and Wales. London: William Eyres.
Hume, David. 2021 [1752]. ‘Of the Populousness of Ancient Nations’, in Essays, Moral Political, and Literary, eds. T. Beauchamp and M. Box. Oxford: Clarendon Press, 279–331.
Ignatieff, Michael. 1978. A Just Measure of Pain. New York: Pantheon Books.
Laqueur, Thomas. 2003. Solitary Sex. New York: Zone Books.
Marx, Gary. 2021. ‘Bentham on Modern Social Control: Prescient, Clairvoyant and More’, in Jeremy Bentham on Police, eds. S. Jacques and P. Schofield. London: UCL Press, 193–206.
McGowen, Randall. 1999. ‘From the Pillory to the Gallows: The Punishment of Forgery in the Age of the Financial Revolution’, Past and Present, vol. 165, no. 1, 107–40.
Montesquieu, Charles Louis de Secondat. 1989 [1748]. The Spirit of the Laws, trans. A. Cohler, B. Miller and H. Stone. Cambridge, England: Cambridge University Press.
More, Thomas. 1965 [1516]. Utopia, trans. P. Turner. London: Penguin Books.
Nagin, Daniel. 2021. ‘Utilitarianism and Policing in the US’, in Jeremy Bentham on Police, eds. S. Jacques and P. Schofield. London: UCL Press, 207–9.
Quinn, Michael. 2014. ‘Bentham on Mensuration: Calculation and Moral Reasoning’, Utilitas, vol. 26, no. 1, 61–104.
Quinn, Michael. 2021. ‘Bentham on Preventive Police: The Calendar of Delinquency in Evaluation of Policy, and the Police Gazette in Manipulation of Opinion’, in Jeremy Bentham on Police, eds. S. Jacques and P. Schofield. London: UCL Press, 35–74.
Radzinowicz, Leon. 1948. A History of English Criminal Law, vol. 1. London: Stevens.
Rawls, John. 1955. ‘Two Concepts of Rules’, The Philosophical Review, vol. 64, no. 1, 3–32.
Romilly, Samuel. 1820. The Speeches of Sir Samuel Romilly in the House of Commons, vol. 1. London: James Ridgway.
Schofield, Philip. 2019. “‘The first steps rightly directed in the track of legislation”: Jeremy Bentham on Cesare Beccaria’s Essay on Crimes and Punishment’, Diciottesimo Secolo, vol. 4, 65–74.
Schofield, Philip. 2025. ‘Jeremy Bentham on adult-child sex and infanticide’, History of European Ideas, vol. 51, no. 3, 513–28.
Semple, Janet. 1993. Bentham’s Prison. Oxford: Clarendon Press.
Shoemaker, Robert. 2004. ‘Streets of Shame? The Crowd and Public Punishments in London, 1700–1820’, in Penal Practice and Culture 1500–1900, eds. S. Devereaux and P. Griffiths. Palgrave Macmillan: New York, 232–57.
Smith, Adam. 1976 [1776]. An Inquiry into the Nature and Causes of the Wealth of Nations, eds. R.H. Campbell and A.S. Skinner. Oxford: Clarendon Press.
Sverdlik, Steven, 2014. ‘Punishment and Reform’, Criminal Law and Philosophy, vol. 8, no. 3, 619–34.
Sverdlik, Steven. 2023. Bentham’s An Introduction to the Principles of Morals and Legislation: A Guide. New York: Oxford University Press.
Sverdlik, Steven. 2024. ‘Bentham’s Project of Applied Ethics, c. 1780: A Penal Code. Part 1: Offences’, Journal of Bentham Studies, vol. 22, no. 1, 1–40.
Tullock, Gordon. 1974. ‘Does Punishment Deter Crime?’ The Public Interest, no. 36, 103–11.
Voltaire. 1963 [1766]. Commentary on Beccaria’s On Crimes and Punishments, in James Heath, Eighteenth Century Penal Theory. Oxford: Oxford University Press, 142–7.
Webb, Beatrice and Sydney. 1963. English Prisons under Local Government. Hamden, CT: Archon Books.
‘Wooden Horse’. Wikipedia: https://en.wikipedia.org/wiki/Wooden_horse_(device).
Zimring, Franklin and Hawkins, Gordon. 1973. Deterrence: The Legal Threat in Crime Control. Chicago, IL: University of Chicago Press.