Hannah Maslen's ambition in this book, to show the relevance of penal theory to the more “real” world of sentencing practice, is exciting. For her, “examining the practical implications of a normative argument serves to prompt reflection on its validity” (pp. 4, 135). For me, the excitement is not so much exploring the validity of her theory, but exploring whether theory can actually help us resolve what to many practitioners appear as practical rather than theoretical questions. The key question of this book – whether remorse should affect the sentence an offender receives – is enormously important. Does Maslen succeed in her ambition? I am not sure. It is a very readable book (unlike many books of penal theory!), and one which summarises theoretical writings in ways which are very engaging. But, rather than bridging the chasm between the world of theory and the world of practice, Maslen manages to illustrate the size of the gap.

For a start, in the first half of the book, she sticks within a retributive framework for reasons which are not entirely clear. She chooses five potential desert-compatible arguments for justifying mitigating a sentence on grounds of desert:

• • The Changed Person argument (this she argues is unacceptable unless the changes are really dramatic, transformative. Its acceptance would require us to buy into “character retributivism”, itself a problematic theory. And it would only allow us to justify mitigation in the most serious cases).

• • The Reduced Harm argument (which may provide evidence of lesser culpability but cannot retroactively alter facts about the offender's state of mind at the time of the offence, so offers very limited scope for reducing the harm (or wrong) and is therefore to be ruled out).

• • The Already Punished argument (remorse as self-imposed punishment, or self-condemnation: a lively analysis of various censure-based retributive theories results in the conclusion that penance could be punishment, but with very limited justification).

• • The Responsive Censure argument (this is her preferred option: mitigation is a proper response to the offender's genuine communication of remorse).

• • The Merciful Compassion argument (which seems to offer Maslen several challenges, but is in any case “redundant” since the Responsive Censure argument is she says a more principled approach).

My bullet-point summary does not, of course, do justice to the arguments as they unfold. It is a delight to see Maslen unpick different versions of “modern retributivism”, to follow her nifty analyses. But I am not entirely convinced – surely remorse can, in fact, revise the “qualitative badness” of the harm? It is a problem for some retributivists that they seem to need to fix (or “set”, in Maslen's language) the measure of harm at the time of the offence. But it seems to me obvious that remorse (and apologies) can reduce harm. With Smith (see Justice Through Apologies: Remorse, Reform and Punishment, Cambridge, 2014), I would seek to argue that an apology/remorse can assuage a victim's fear or guilt, and indeed it can (although Maslen would deny this) assuage the wider community's fear or guilt. Maslen seems to downplay this argument by pointing out that a guilty plea in many jurisdictions results in a significant discount whether or not this is connected to remorse. But why is this relevant, and, in theory, should it? Discounts for guilty pleas may (perhaps) be justified in practice, but not in theory?

Why does the Responsive Censure argument win? First of all, we have to distinguish between a person's blameworthiness and the extent to which we should blame or censure them. Then we have to consider how punishment communicates censure. Maslen builds on Duff's conception of punishment as explicitly “dialogical” or reciprocal. An offender must be given the opportunity to be remorseful – and when there is such a response it must not be ignored. It is certainly attractive to think of censure as reflection-facilitating (p. 105). What Maslen calls a “dialogical” model of censure seems to be right, whether you take a retributivist or more utilitarian position. It seems intuitively obvious that censure (blame) which is responsive to the offender has greater moral and communicative legitimacy. But it is not just the Responsive Censure argument which leads us to this conclusion: there can be good utilitarian arguments for communicative accounts of punishment. Nor am I comfortable with all her arguments on mercy, on which she seems unduly tough: does feeling compassion for an offender really have “strong overtones of pitying the offender” (p. 121) and, if it does, is that necessarily wrong? Remorse is not the same as shame, guilt or regret: what is so wrong with giving Merciful Compassion a small role, carefully explained to avoid inconsistent and discriminatory application?

The book is full of neat summaries and juxtapositions of conflicting theoretical approaches. For example, when it comes to the Already Punished argument, Maslen argues that Duff's theory could allow for penance as punishment, but that, from von Hirsch and Ashworth's perspective, remorse could not constitute punishment, as censure must be imposed by an external authority. (Would they agree?) Students of penal theory will enjoy her ostensibly simple summary of retributivism, her tidy classifications. Modern retributivism is thus divided neatly into:

• • benefits and burdens retributivism (Morris, Finnis, Murphy);

• • penance perspectives on retributivism (Duff);

• • censure and sanction retributivism (Von Hirsch).

Even if this book is no easy read for the non-expert, one gets a wonderful flavour of Duff, von Hirsch, Tasioulas, etc. all battling it out. It is like eaves-dropping at a conference of penal philosophers, with a glass of wine in your hand.

It gets more difficult. Once you agree that remorse does have a role, how does it fit in with other aggravating and mitigating factors? The second part of the book examines the implications of Maslen's normative conclusions for sentencing practice. Chapter 7 raises important questions about whether, for example, offence type or offence seriousness should influence the weight that should be assigned to remorse. Here, to my surprise (at p. 147), Maslen admits that “bearing in mind the multiple aims of sentencing, we must not forget that imposing a limit on the mitigating role of remorse might also, or instead, be justified on consequentialist grounds” (p. 147, Maslen's emphasis). This seemed to pull the rug from under the main argument of Part One. I had been reading the book as simply a justification of remorse as mitigation from a retributivist perspective. I had by this point accepted that the Responsive Censure argument usefully reveals what is of foremost value for “remorse-based” mitigation (“the prior full acceptance and agreement with the message communicated by censure” (p. 160)), but now it appears that Maslen is happy for us to agree this on utilitarian, consequentialist grounds as much as retributive grounds?

I was looking forward to the theorist's answer to difficult questions relating to how a sentencer can justify more or less mitigation for remorse for different types of offence or offences of different seriousness, but the answer seems to be that “the Responsive Censure argument renders remorse-like responses consistently relevant and … neither limits or maximises their influence for the most serious offences… . Responsive Censure requires judicial consideration of the offender's prior acceptance of the message, whatever the content of that message” (p. 160). So, sentencers should always dialogue, and always listen out for remorse. That makes sense – but how much is the discount? Before we get to that, we have Chapter 8, which explores the challenge of the remorseful recidivist. Here, Maslen acknowledges what is clearly true: genuine remorse and attempts at desistance are not incompatible with slipping back into crime. Desistance is truly difficult. Conclusion – “previous convictions should not always be taken to discredit claims of remorse” (p. 178). This does not seem to add very much to what Maslen calls elsewhere “theoretical precision”. In Chapter 9, Maslen is rightly critical of the vagueness of Sentencing Guidelines in England and Wales. Her international comparisons are disappointingly limited to Australia, New Zealand and the US: we can learn much from European sentencing systems (my expectations may be unreasonable, and Maslen has a clear agenda for future research). Her analysis of data collected by the Sentencing Council under their now defunct Crown Court Sentencing Survey suggests that “remorse has a significant effect on the likelihood of custody for assault offences, although its effect decreased as offence seriousness increased, presumably because the most serious assaults almost always attract custodial sentences” (p. 189). Is she prepared to say explicitly that sentencers give too little “discount” for remorse in serious cases? I am confident that Maslen is right to conclude that “remorse most likely came to have a role to play [in Guidelines] because it was intuitively thought relevant” (p. 202): but, at the end of the day, does she really offer any more principled guidance? I would encourage Maslen to attempt a draft Guideline for judges on the role of remorse.

I am delighted that Maslen is prepared to tell retributivist theorists of all hues that remorse has an important part to play in mitigating sentences, that Responsive Censure permits the widest application of remorse-based mitigation. That is an important theoretical contribution. But I am left slightly disappointed: this book promised so much and offered me less than I (unreasonably?) hoped for. At the end, Maslen suggests that the implications of the Responsive Censure argument reveal deficiencies in some aspects of sentencing guidance AND renders convincing the theoretical position itself. There is much that is attractive in her theory; and much that is deficient in current sentencing guidance. But the connections are not quite there for me. What Maslen does less successfully is show the relevance of the theoretical debate to real life. She has not yet convinced this avowedly “real-world” criminal lawyer how this discussion of penal theory actually “helps”. In that sense, the book is a disappointment. In other ways, it is a lively and instructive read.