Guilt and Choice in Criminal Law Theory – A Critical Assessment

In Germany and other European countries, courts and legal scholars assume that the Principle of Guilt (Schuldprinzip) is of central importance for the legitimisation of criminal punishment. The German Federal Constitutional Court (Bundesverfassungsgericht) assigns the Principle of Guilt constitutional status. Although it is not mentioned in the German Constitution (the Grundgesetz from 1949), the Federal Constitutional Court refers to it in several judgments such as the following (the Lissabon-decision) from 2009:


Guilt and Choice in Criminal Law Theory -A Critical Assessment
TATJANA HÖRNLE* 1

The Principle of Guilt and the Principle of Alternate Decisions
In Germany and other European countries, courts and legal scholars assume that the Principle of Guilt (Schuldprinzip) is of central importance for the legitimisation of criminal punishment.The German Federal Constitutional Court (Bundesverfassungsgericht) assigns the Principle of Guilt constitutional status.Although it is not mentioned in the German Constitution (the Grundgesetz from 1949), the Federal Constitutional Court refers to it in several judgments such as the following (the Lissabon-decision) from 2009: Criminal law is based on the principle of guilt.This principle presupposes a human being's own responsibility, it presupposes human beings who themselves determine their actions and can decide in favour of right or wrong by virtue of their freedom of will.The protection of human dignity is based on the idea of Man as a spiritual and moral being which has the capabilities of defining himself, and of developing, in freedom (see BVer-fGE 45, 187 <227>).In the area of the administration of criminal law, Article 1.1 of the Basic Law determines the idea of the nature of punishment and the relationship between guilt and atonement (BVerfGE 95, 96 <140>).The principle that any sanction presup-poses guilt thus has its foundation in the guarantee of human dignity under Article 1.1 of the Basic Law (see BVerfGE 57, 250 <275>; 80, 367 <378>; 90, 145 <173>). 1 The Constitutional Court did not invent the Guilt Principle -it traces back to older developments in post-war criminal law doctrine.In a ruling from 1952, the German Federal Court of Justice (Bundesgerichtshof, the highest appellate Court for criminal and civil cases) emphasised the importance of the Guilt Principle (in a rather brief way, obviously meaning to express a self-evident truth): and the intensity of reactions to Frankfurt's attempt to debunk the Principle of Alternate Possibilities underscores this point. 6e German courts (in the decisions cited above) use a somewhat narrower concept of alternate possibilities than Frankfurt.While his focus is on alternate actions ('could have done otherwise'), the crucial point is the possibility of alternate decisions ('could have decided otherwise').To distinguish between a broader Principle of Alternate Possibilities and a somewhat narrower Principle of Alternate Decisions is helpful for an analytical purpose.It allows separating internal freedom from external freedom -that is freedom from coercion, by persons or other sources of imminent danger.
For those who are familiar with the English literature on the subject of criminal responsibility, the general reasoning behind the German insistence on a Principle of Alternate Decisions must appear familiar, too.It is a common assumption in contemporary legal theory that the foundation for criminal responsibility is choice. 7Modern conceptions of individual fault are deeply rooted in subjectivism and the notion of having made the wrong choice. 8r the following discussion it will be helpful to have an example, therefore let me introduce one: suppose that Mrs. Miller on March 1, 2016, had an argument with her husband; at 10.05 pm she grabs a kitchen knife and stabs him deep in the chest.She intends to kill him, but he survives.I will leave aside offence descriptions and it should also be assumed that Mrs. Miller did not act under circumstances that would constitute a justification, such as self-defence.The questions I am going to address here are: Is it a necessary requirement for convicting Mrs. Miller that a judgment about wrongdoing can be supplemented by a judgment about guilt?Would blame for intentionally stabbing her husband only be fair if it can be shown that at 10.05 pm she could have made another decision?The 1952-ruling by the Bundesgerichtshof answers these questions in an affirmative way.In German criminal law doctrine, the expression Schuldvorwurf is used to denominate this particular content of blame, and the distinction between the two 6 The number of publications in reaction to Frankfurt's article is huge.To  categories of 'wrongdoing' and 'guilt' is regarded as an essential feature of the Allgemeine Verbrechenslehre (general structures of crimes). 9e of the aims of this article is to challenge the view that criminal punishment is only legitimate if the actor could have decided otherwise.First, I will point out that references to 'free will' are permissible in some legal contexts, but should be avoided in the criminal law as a consequence of its blaming features (see below 2).Blame must be based on accurate descriptions compatible with empirical data.The Principle of Alternate Decisions, however, is not compatible with the most plausible model of human decision-making (3).Further, I will argue that this does not mean to give up on criminal law as a blaming institution.Rather, the blame implicit in criminal convictions should be conceptualised as blame for wrongdoing -such judgments are legitimate even if alternate decisions are not possible at any given moment in time (4).However, the simple compatibilist conclusion proposed by some authors (who accept determinism, but argue that there is no need to change criminal law doctrine) is not convincing.To question the notion of choice as expressed in the Principle of Decisions and to eliminate blame for guilt makes it necessary to re-consider some parts of criminal law doctrine (5).

Why the Term 'Free Will' Can Be Misleading
The questions to be discussed in this article are often framed with the expression 'free will' .However, it is not the best approach to ask in a general, unspecified way: ' Are people free?' Freedom is not a natural attribute that can be observed and measured -for this reason, to talk about the possibility or the impossibility of 'proving' free will is misleading.'Free' is a normative term -the (possible) result of a normative evaluation.Criteria for this evaluation must depend on the context.In practical ethics and in law, applying the label 'free' serves a purpose, namely to justify certain consequences.The decisive question ought to be: '… free enough for this consequence'?It is one thing to consider the unique features of a human being in a philosophical discourse or within constitutional theory, but it is quite another to do so in a criminal verdict.
In other contexts than the criminal law, there might be good reasons to emphasise specific human properties in order to distinguish us from other species, for instance, our ability to evaluate first order volitions with second order volitions. 10From the perspective of psychology, attribution is a universal practice.With the help of the attributions of others already from the early childhood, human beings constitute themselves as selves able to make autonomous decisions and actions. 11A somewhat idealistic vision of human beings also serves the purpose to demand respect from each other and from the state.For instance, the German Federal Constitutional Court's reference to the 'idea of Man as a spiritual and moral being which has the capabilities of determining himself, and of developing, in freedom' has to be seen in this context. 12Within political and constitutional theory, such attributions have the function of promoting a decent state -a state that treats people with respect and leaves them room to organise their lives without too much interference from the state.If the task is to sketch the relationship between state and citizens, the demand 'be realistic about human beings' can be relaxed to some degree. 13wever, a heavier burden of justification must be shouldered if a reference to 'free will' becomes an essential part of the message delivered in a criminal conviction.This message is much more specific than general praise for the abilities of human beings -its context is a negative one and it is a particularly serious judgment.Criminal convictions blame the individual, and this blame expressed by the state must be justified carefully.Therefore, the underlying description that this person `could have decided otherwise´ at the time of her offence must be a correct description.Its blaming features and the (at least potentially) very serious consequences of criminal convictions distinguish the criminal law from other branches of the law.
The notion of 'free individuals' is a Janus-faced concept.Liberal conceptions of the relationship between the state and its citizens and the interactions between citizens stress freedom of action.Political philosophy and constitutional theory do well in emphasising the individuals' need for freedom and self-determination.Moral philosophy and criminal law, however, have to deal with the second 'face' -the more problematic one which appears if strong visions of 'freedom' are used to justify blame and hard treatment.If one insists that the Principle of Alternate Decisions is essential for criminal convictions, it must be shown that it is consistent with what we know about human decision-making.Other than in constitutional theory, it is not sufficient to point to some lofty ideal of humanity if the consequence is to blame an individual with criminal sanctions for not having taken the rightful decision.The question; 'to what degree' must be left open (obviously, some foothold in reality is necessary, otherwise constitutional theory would become a pointless exercise in day-dreaming).My point here is only: requirements must be stricter if we turn from constitutional theory to criminal law.

Cognitive Sciences and the Consequences for the Criminal Law
At this point, the cognitive sciences come into play as they shed serious doubts on the Principle of Alternate Decisions.Criminal law theorists have tried to avoid this conclusion with evasive strategies, arguing for instance that different academic fields employ their own methods, or introducing the notion of different languages (Sprachspiele) to isolate criminal law from science. 14This is, however, not a convincing approach. 15If the criminal court in my example expresses a judgment about Mrs. Miller's guilt, blaming her for making the wrong decision although she could have decided otherwise, this is a statement about a factual circumstance, 16 which must be compatible with scientifically sound models of human decision-making. 17A lot of attention has been paid to a group of experiments called the Libet-experiments.Their point, to put it in a nutshell, is to demonstrate that our experience of consciously having decided between two alternatives is preceded by a not yet consciously experienced willingness to act.The latter can be registered as neuronal activities before the subject expresses to have made the decision. 18 Philosophical writings about the 'could have done otherwise'-problem often focus on counterfactual approaches ('could have done otherwise if she had made another decision'; see for instance Moore 2016 pp.66-89).However, if one takes the perspective of judges, counterfactual scenarios are not very helpful: they have to make statements about the events as they were.Counterfactual reasoning is not going to answer a defendant's question: 'Why do I deserve blame?' in a convincing way.17 This argument applies to all kinds of criminal law judgment.For instance, the role of the 'reasonable person' that is needed for the verdict 'negligent behaviour' should be based on research in the cognitive sciences which warns against modelling role expectations in a perfectionist way.gued extensively about which conclusions could be drawn from these experiments. 19cholars in philosophy and criminal law often point to the limits of the experimental design and therefore conclude that the criminal law may continue to rely on the Principle of Alternate Decisions. 20But an exclusive focus on an analysis of the Libet experiments misrepresents the state of knowledge achieved by neuroscience and the cognitive sciences.There are other important sources for insights into the neurological foundations of decision-making, for instance changes of behaviour after injuries or diseases in certain areas of the brain. 21spite the increasing body of research in the area of cognitive sciences, those who defend the Principle of Alternate Decisions can still point to the incompleteness of knowledge.Because neuroscientists cannot claim to have discovered all the minutiae of how the human brain works, some criminal law theorists assume that we can ignore their work until a final model of decision-making is presented. 22However, this strategy is not convincing, either.Anyone who has done some basic reading in epistemology knows that all scientific models are preliminary.For scientists, it is a matter of methodology to admit that assumptions about human decision-making are and will be incomplete.But in the meantime, legal judgments do make statements about facts such as in my example: Mrs. Miller could at 10.05 pm have decided otherwise.Criminal courts do not and cannot wait until scientific knowledge is 'complete' .The solution for coping with such situations cannot be to ignore science altogether and work with fictions.Rather, criminal law theorists should attempt to grasp what the most plausible model of human decision-making available is.
How would the most plausible model of human 'decision-making' look like?For our purpose, the salient point is: At the given time, could an individual have made another decision?In answering this question, the starting point must be that decision-making depends on physiological structures, that is: the brain and other neurological structures.The possibility of another entity (called, for instance, `spirit') that could circumvent these structures has to be discarded.On the one hand, such an entity would need to be able to supervene the neurological structures that stand behind what is called decision-making.On the other hand, the entity would need to causally influence those neurological processes which lead to bodily movements or circumvent them in initiating movements,this is highly implausible. 23statement on which all neuroscientists are most likely to agree is that the interactions and processes involved in human decision-making are highly complicated. 24For our purpose, however, it is only necessary to focus on some features.Human beings are capable of reasoning and deliberation -we can access stored information about moral or prudential reasons (judging the likelihood of rewards and costs and weighing these consequences) and apply them to a present situation.Neuroscience can show which parts of the brain are involved in these activities.25Some elements of the complex interactions and loops that are called 'decision-making' are, however, beyond conscious reflection.Brain researchers even claim that non-deliberative activities in brain sections such as the limbic system are having the 'final word' . 26 the assumption is correct that crucial parts of what is called decision-making occurs in processes that are not open to conscious deliberation, the argument 'you could have decided otherwise' looses force.And even if the networks for deliberative activities would exercise final control, the main problem remains: decision-making depends on existing physiological structures.In the example used here: what Mrs. Miller decides at 10.05 pm is determined by the conditions of her neurological system at 10.05 pm.This infrastructure has been shaped by her life history up to that moment.It will include stored information about morality, if she has been exposed to moral arguments about not injuring other human beings -which we can safely assume, and most likely also about prudential reasons for obeying the law.But, although neural structures are flexible and open to changes at any given point in time, decisions and actions depend on the resources, as they are available at this very moment. 27As already mentioned, we should not assume that some immaterial entity mysteriously intervenes.The only somewhat more serious hypothesis is that there might be randomness in the way the neural cells interact, perhaps caused by indeterminism on the molecular level.If this were the case, it would only exacerbate our difficulty -if the outcome is determined by random events, it becomes even less plausible to blame Mrs. Miller for having made that decision. 28iminal law theory with an emphasis on the Principle of Alternate Decisions and the paradigm of choice contradicts itself.The inherent problem is the timeframe applied.The Principle of Alternate Decisions refers to a 'snapshot' timeframe 29 -but if one dismantles events into such narrow timeframes, it becomes incoherent to blame persons for not having decided otherwise.Consequently, criminal law theory should avoid the concepts of alternate decisions and choice.
The criticism against the Principle of Alternate Decisions has different force if one distinguishes two arguments that refer to it.It figures both in justifications of blame as a general practice and in arguments describing the content of blame in individual cases.The two decisions cited at the beginning of this article illustrate these two modes of referring to the Principle of Alternate Decisions.The German Federal Constitutional Court talks about the 'idea of man' as a prerequisite and legitimatory basis for the institution of punishment, using a preamble-like reference.The more serious problems arise with regard to the content-orientated notion of blaming offenders for their guilt (Schuldvorwurf), which the Federal Court of Justice (Bundesgerichtshof) presented in its 1952-ruling.If one turns from punishment theory to actually blaming and punishing individuals because they had not decided otherwise, the question of individual fairness adds weight to the problem.If the content of blame is formulated in such a way (applying a 'snapshot'-timeframe), it 27 See for this point Schiemann, Kann es einen freien Willen geben?-Risiken und Nebenwirkungen der Hirnforschung für das deutsche Strafrecht, 57 Neue Juristische Wochenschrift (2004)

The Role of Guilt in German Constitutional Law
If these critical remarks about the Principle of Alternate Decisions are presented to a German audience, the following objection will be raised immediately: the notion of Schuldvorwurf is indispensable because the Principle of Guilt has constitutional status. 30owever, the Federal Constitutional Court's decisions reveal that its conception of 'guilt' is very broad, much broader than the conception used in German criminal law doctrine.My analysis does not amount to a full-scale criticism of what the Federal Constitutional Court puts under the heading 'guilt' .The Court stretches references to 'guilt' far beyond preamble-type references to the 'idea of man' .It points to different principles that are, and undoubtedly should remain, central to the criminal law -principles that can be explained and justified independent of the Principle of Alternate Decisions. 31For instance, both constitutional and the criminal law demand rules to establish a close connection between an individual's actions and the outcome for which this individual is blamed.It would not be compatible with human dignity if a person who had nothing to do with the event would be picked out for punishment as a scapegoat. 32Criminal convictions must be based on rules for the attribution of responsibility such as the act requirement, causal contribution and, in the case of omissions and negligence, specific relationships and duties to act.Of course, it is open to debate exactly where human dignity draws the 'red line' for attributing personal responsibility, but it is not necessary to discuss details here.The crucial point is: one does not need the Principle of Alternate Decisions to argue against, for instance, punishing relatives of the offender who were not involved in the commission of the crime.The concept of personal responsibility continues to play a central role in criminal law theory even if the Principle of Alternate Decisions is abandoned.Another set of rules that the Federal Constitutional Court mentions due to its broad conception of 'guilt' concerns proportionality between seriousness of the offence and sanction severity.There are good reasons to adopt a principle of proportionality 33 -but again, these reasons do not depend on a Principle of Alternate Decisions.

The Options for Criminal Law Theory: Choice as Fiction -Foregoing Blame -Blame for Character -Blame for Wrongdoing
Having argued that blame for not having decided otherwise is incompatible with an empirically sound model of human decision-making, the question arises: What should be the next move in criminal law theory?
A way out of the dilemma could be to shift the content of blame from choice to character.Character-based theories of criminal punishment are not popular in contemporary Germany because they tend to be equated with National Socialism.Nevertheless, there are noteworthy exceptions: Rolf Herzberg recently has made a case for a character-based theory of criminal law, on the basis of determinism. 34Within criminal law theory in Britain and the US, the topic of character seems to be less politically contaminated: character-based theories are discussed as a genuine alternative to choice-based theory. 35Contemporary character theories tend to come closer to choice theory: they do not claim to make a holistic judgment about the 'whole person' , but focus on attitudes that manifest themselves in the criminal act. 36But with a focus on character, blame could still have two contents: blame for having wronged another person (by violating his or her rights) and blame for those features of the offender's character which show up in the commission of the offence.This way of defining the content of blame does, however, not offer a promising solution for those in search for deep responsibility. 37Forces that are evidently important, but beyond one's own influence (such as genetical factors and foetal development) have shaped any human biography.Not being an empathic and considerate person can probably largely, or at least partially, be explained with deficient neurological structures acquired at early stages in life.Even if one can identify later junctions which turned out to have disastrous consequences for this person's future development: if we were to blame her for having taking this wrong turn, another 'snapshot'-timeframe would be applied that invites the same kind of criticism as blame for the decision to commit the criminal act.
Another somewhat cynical approach admits, within academic circles, that the Principle of Alternate Decisions is indeed not justifiable, but claims that criminal law practice should nevertheless uphold the fiction of legitimate blame for not having decided otherwise.Two reasons could be offered for this strategy: either a functional reason (crime control) 38 or reasoning that points to the offenders' own subjective experiences.Many persons continue to passionately defend 'free will' in a strong sense, including the Principle of Alternate Decisions, based on subjective daily experiences to consciously choose between options.This first-person-perspective is misleading 39 -but for many, it cannot be transcended.Picking up on this phenomenon, it has been argued that it is not unfair to treat offenders according to their first-person-perspective 40 -if they feel like having had a choice, why not blame them based on their own perceptions?Another conclusion could be to abandon the notion of blame in criminal law altogether.A few authors have spelt out this radical consequence, opting for a preventive system of reactions without blaming features. 41However, this is neither a necessary nor a recommendable consequence.Criminal law involving blame should be preferred to a system of merely preventive reactions because a statement about wrongdoing is owed to the victim of the crime (punishment theory cannot be discussed in full here but I will come back to the victims-argument below).
The solution proposed here is that we should neither abolish the criminal law nor continue the 'either choice or character'-debate, nor rely on fictions of alternate decisions being possible.The simpler solution is to limit both the justification and the content of blame to wrongdoing in the legal sense, that is, to the judgment 'you have violated the rights of another person or important collective interests' . 42The legitimacy of such a blaming statement does not require additional assumptions about choice, character, maintenance of public order or the offenders' first-person-perspectives. If the offender has committed wrong, no more additional content of blame needs to be added.The legitimacy of blame for wrongdoing can be justified without nurturing fictions such as the Principle of Alternate Decisions and without the concept of guilt.
Admittedly, many criminal law theorists and moral philosophers will not be comfortable with this solution.A concern is raised in discussions: If one abandons the notions of choice and guilt, does this mean to abandon a much larger set of normative concepts such as 'wrongdoing' and 'blame'?Is it possible to defend blaming practices while questioning the focus on guilt that is so firmly anchored in our moral and legal practices?The answer is that our social practices of attributing moral and legal responsibility to persons are not just historical developments that could or should be reversed.Rather, they can be justified in ways that are neither dependent on the Principle of Alternate Decisions nor on sociological arguments such as: we are used to these concepts and they are functionally important. 43Using the key concept of 'wrongdoing' to make a person responsible for her actions is an essential, non-contigent feature of developed societies.Questioning the Principle of Alternate Decisions does not render all normative concepts meaningless that presuppose human responsibility.The scope of the argument made in this article is much narrower.The judgment: 'X has committed a wrong and deserves blame for it' retains its sense even if the additional description 'and he could have decided otherwise' is ommitted.
Many philosophers and legal theorists will nevertheless continue asserting that the concept of personal responsibility must be richer -that 'ultimate responsibility' beyond an attribution of wrongdoing is required.These philosophers and legal theorists share the strong moral intuition that the description 'he or she could have made another decision' is an indispensable requirement of doing justice to this individual.If intuitions are widely shared and deeply felt, their roots often remain obscure.One possible explanation for the contemporary insistence on 'ultimate personal responsibility' might be found in religious traditions.Academics beyond the field of theology would not cite religious sources explicitly or may not even be clearly aware of them.However, they have shaped our collective history of ideas and the intuitions flowing from this pool.Judgments from a religious perspective, especially those modelled after the notion of one 'Last Judgment' , tended to focus on the person rather than on isolated incidents (not a feasible option for human judges 44 who can at best grasp incomplete fragments of the defendant's personality).The consideration that there are religious roots of common moral judgments might also add something to the explanations for the historical shift in criminal law that Nicola Lacey has described: the gradual replacement of character-based with choice-based ascriptions 43 Wallace, Responsibility and the moral sentiments (Harvard University Press 1994).44   See Duff, What kind of responsibility must criminal law presuppose?In Free Will and Modern Science, ed.Swinburne (The British Academy 2011) pp.178-199, at 194. of responsibility during the last two centuries. 45Several factors probably promoted the replacement of a holistic, character-based judgment with the modern 'snapshot'-timeframe.One factor could be the increasing awareness of epistemological difficulties related to character-based judgments.Another factor might be the desire to emancipate human beings from the grip of religious commands by emphasizing their capacity for autonomous decision-making.From today's perspective, looking back on literature from the period called the Enlightenment, one can observe that 'enlightened thinkers' in some ways still were closely attached to their religious roots. 46The desire to find a substitute for God's 'the whole person'-perspective in the 'wrong choice' standard could be interpreted as a lasting subliminal attachment to Christian thinking.As long as one feels the need to find an alternative for a certain concept, one has not quite liberated oneself from it.The most straightforward solution, however, is the one proposed here: it is unnecessary to find a substitute for character judgments (be they ambitiously holistic or more moderately restricted to the attitudes manifested in the criminal act).It is legitimate to express blame simply because the offender has disregarded the other person's rights and disregarded his social duties.
The wish to look deep into a person's soul or alternatively into her mind before passing judgment about wrongdoing looses its appeal once the differences between state punishment on the one hand, religious punishment and/or moral judgments on the other hand have become clear.State punishment serves different purposes.The purpose of blame through criminal convictions is to clarify and demarcate the sphere of rights and duties that citizens (and persons who are only temporarily located in the state's territory) 47 have.The judgment that wrong has been done clarifies boundaries for conduct and clarifies social duties.These duties depend on the particular social role the offender had -they could be comprehensive professional duties, such as those of doctors or lawyers, or much less ambitious general social duties, such as for instance to maintain buildings and other objects in a state which does not endanger the public.
Criminal law judgments should primarily focus on what went wrong in the specific social interaction that constitutes the criminal offence.From a victim's perspective, blame is justified if the offender has disregarded his or her rights.For instance, from the perspective of Mr. Miller in my example, who regains consciousness in the intensive care unit, it is appropriate and sufficient to blame his wife in the following way: 'you should not have stabbed me in the chest' .This does not amount to what Nicola Lacey has called 'outcome responsibility' , assuming that this was the only alternative to either 'choice' or 'character' . 48It is not sufficient to point to the harm caused through the act (the injured body) as such.Rather, the core feature that calls for blaming Mrs. Miller is the violation of the duty not to attack others.Because, for good reasons, victims are prohibited from reacting to the wrongdoing themselves, the state has to pronounce blame on behalf of the victim and to administer the punishment that serves to emphasise the seriousness of the crime.However, both the legitimacy and the content of blame through criminal punishment should be based on a 'victim-as-disregarded-co-citizen'-perspective.Inquiries into the offender's character or her ability to have decided otherwise are superfluous and should be avoided.Some readers might worry that this might lead to a harsher system of punishments.However, the approach proposed here does not amount to treating offenders in a remorseless way.On the contrary: a realistic, non-metaphysical view of human beings should promote an anti-perfectionist stance.An anti-perfectionist stance eschews ambitious moral demands.It avoids hindsight biases and exaggerated expectations when considering what is 'reasonable' within a particular social role. 49And, with regard to punishment levels, the insight that we all are fallible beings and that 'conscious choice' might not work quite as it appears to us should promote scepticism towards a system with high overall severity.The solution is not to focus on 'diminished personal guilt' in some cases (and not in others), but to scale down the severity of punishment in general.

Consequences for Topics in Criminal Law
Adopting the aforementioned premises has implications for several topics in criminal law: some doctrines and legal rules should be modified.So far, scholars have not discussed the necessity of partial changes, as criminal law theorists fall into two camps when confronted with the cognitive sciences and particularly neuroscience.As mentioned above, a small minority declares total surrender and concludes that criminal law must be abolished in favour of a preventive system.The vast majority, however, either ignores the empirical models of human decision-making or brushes them aside, 50 or develops a compatibilist position with the conclusion that criminal law doctrine is not in need of change. 51A third camp's position would be that foundational features (criminal law as a blaming practice) can be upheld, but that revisions need to be made regarding some provisions in criminal codes or non-codified teachings and doctrines.So far, this camp lacks a sufficient amount of members.In this article, I can only sketch which issues are in need of closer review if criminal law no longer is based on the notions of choice and guilt. 52The main point is: criminal law doctrine will not be left as a 'pitiful ruin' 53 once the Principle of Alternate Decisions is abandoned, but it needs to be re-examined with regard to circumstances that have been traditionally categorised as issues of guilt, such as insanity, diminished responsibility and excuses.Furthermore, the central role of offenders' volition needs to be re-considered and downgraded if the focus is not on the 'guilty mind' , but on the violation of duties towards co-citizens.

From Guilt to Communicative Requirements: Why Some Pathological Mental States (but Not Others) and Immaturity Are Reasons to Refrain from Punishment
If wrongdoing is seen as sufficient reason for blame and criminal punishment, should we discard the circumstances traditionally labelled 'exemptions from guilt' as irrelevant? 54eaders accustomed to the English scheme organised around 'actus reus; mens rea; defences' 55 will come up with a similar question, namely: does my proposal include the abolition of the insanity defence?Should we punish persons who suffer from a severe psychiatric disease -in my example, punish Mrs. Miller if the background was not an 'ordinary' argument between spouses, but an acute state of psychosis leading her to believe that an evil demon had seized the body of her husband?Children can also commit wrong, but should we punish a ten years old child who stabs another with a knife?With regard to the examples just mentioned, a truly radical departure from established legal doctrines is not required.The explanation for exempting these offenders from criminal punishment needs to be re-thought -but the outcome will be the same.Starting with the premises developed here, it would not be a satisfactory explanation to argue that these persons could not have decided otherwise and therefore acted without guilt.This is not a criterion that allows distinguishing between cases involving child offenders or offenders who suffer from a severe mental pathology and ordinary cases.A mentally sane adult could also not have decided otherwise.The rationale for exempting the insane and the very young must be found elsewhere.
Crucial to the evaluation is the persons´ inability to be involved in a meaningful communication about what they have done.References to the communicative function of criminal trials and criminal convictions are the key to explaining why children and mentally ill persons should not be punished.The basic claim, developed in detail by Antony Duff, Sandra Marshall and others, states that criminal law in practice should not be understood as a one-way-communication but rather as a dialogue between the accused and those who call him to answer. 56To take part in this enterprise, communicative skills in a wider sense are necessary.Defendants must not only possess the cognitive skills necessary for any sensible communication: to hear and to talk, to understand vocabulary as well as complex sentences and to respond to others in an intelligible way.In addition, they must also have a basic understanding of both reality and abstract normative concepts such as 'rights of others' and 'duties towards co-citizens' . 57Animals, very young or intellectually severely handicapped persons and persons with a grossly distorted view on reality, for instance believing in demons disguised as humans that ought to be killed, do not have the basic skills and understanding.
If one thinks in schemes how to check criminal liability, the intellectual and cognitive pre-requisites necessary to 'answer for crime' 58 should be the first item on the list, not an item to be checked at the bottom under the label 'guilt' .For this reason, contemporary German criminal law doctrine (Allgemeine Verbrechenslehre) 59 should be revised.According to it, the category 'guilt' appears late in the evaluative sequence: after it has been established that the conduct falls under the offence description (objektiver Tatbestand), that it was committed with the necessary intention (subjektiver Tatbestand) and without justification (Rechtswidrigkeit).Only as the next step, guilt (Schuld) is to be examined by asking if exemptions (Schuldausschließungsgründe) or excuses (Entschuldigungsgründe) play a role.However, if the focus shifts from guilt to communication, it becomes a matter of logical priority to establish if the offender can enter a dialogue before examining the possible content of blame. 60e claim that the offender must have basic intellectual and cognitive abilities is also raised as a matter of procedure, under the heading 'fitness for trial' .The communicative aspects also matter for substantive criminal law -if the necessary abilities were not present at the time of the offence, criminal proceedings should not be initiated.This statement requires explanation.In many cases, the time span between offence and trial is significant.How should we deal with constellations in which the actor at the time of the offence lacked basic cognitive abilities and understanding, but later is able to understand that what he has done was wrong?The claim that wrongdoing suffices for criminal liability might be interpreted as meaning that we only have to postpone criminal proceedingsand try child offenders or the insane once these offenders have reached maturity or have recovered from an illness.Considering this possibility, it should be discarded for two reasons.First, there is the identity problem. 61In criminal trials, diachronic identity usually is assumed even if many years have passed since the offence.It is, however, not self-evident that such presumptions are defendable for all cases, particularly when convicting very old persons for their actions as young adults.For our topic, the identity problem is even more pressing.It would be inconsistent to claim that due to her newly developed abilities we can now communicate with a person, while ignoring this fundamental change when blaming her for past conduct.Secondly, 'answering for crime' presupposes not only that the accused can understand the description of their prior conduct as wrongdoing.In addition, it requires that they are capable of presenting their view, drawing attention to factors that might be relevant for the evaluation.A defendant could not take this active communicative role if he at the time of the offence suffered from mental pathologies or was a child.He might have some dim and distorted recollections, but the person he was is not really accessible retrospectively to himself.For these reasons, basic competences must have existed at the time of the offence.So far, the difference between traditional guilt-based criminal law theory and a focus on wrongdoing has pointed towards relatively minor changes -changes regarding schemes on how to structure the evaluation, and different reasons why mental deficits and immaturity exempt from punishment.However, more substantial differences are also likely to surface.Some criminal codes contain an extensive range of 'exemptions from guilt' , such as the German Criminal Code and other civil law codes 62 , as well as the American Model Penal Code. 63In those provisions, offenders will not be punished if, due 60 See Duff 2007  to some pathological mental state, they either are unable to understand wrongfulness or unable to act according to this understanding.One can argue about how ambitious 'ability to understand wrongfulness' ought to be conceptualised.While consensus can be expected about the necessity of basic intellectual abilities, the threshold for offenders' minimal normative understanding might be more controversial.For the 'victims-as-disregarded-co-citizens'-approach proposed here, offenders must be capable to understand basic normative concepts such as citizens' rights and duties, but it is not necessary that they share all aspects of the relevant moral reasoning.A more demanding notion of reason-responsiveness (or, in German: normative Ansprechbarkeit) 64 claims that the offender must have a thorough understanding of the particular normative demand and of himself as being bound by it.Differences appear, for example, with regard to the condition called psycho-or sociopathy. 65Should persons' different abilities to truly understand the meaning of moral demands and to evoke affective responses necessary for empathy and considerateness 66 be relevant for criminal law judgments?From the victims' perspective, the answer is: no.The cognitive sciences can clarify the role of emotions for reason-responsiveness.For the purpose of the criminal law, however, it is legitimate to treat offenders' moral and emotional inner lives according to a 'black box'-model.
Legal systems with wide-ranging exemptions from guilt 67 also apply them to offenders who are not capable to act in accordance with their normative understanding.The idea is that offenders must be able to regulate their impulses.However, this reasoning is based on the notion of choice.If we stop blaming offenders for having made the wrong decision and focus on communication about wrongdoing, it is no longer evident that the effectiveness of personal impulse control should be a matter of interest for the criminal law.Criminal law should take an outside view.From the perspective taken here, it is not a lamentable shortcoming that insanity defences in some states within the US or § 20 Nor-wegian Criminal Code 68 tend to be conceptualised more narrowly 69 than, for example, the wide exemption clause in § 20 German Criminal Code.

Excuses
Criminal codes contain provisions that address unusual external circumstances such as duress or self defence with excessive force.How far do such provisions depend on the Principle of Alternate Decisions?It is often assumed that extreme circumstances may constitute an excuse, because the offender could not have decided otherwise. 70Again, this rationale is not a convincing one because it does not provide a relevant distinction: no actor, with or without this excuse, could have decided otherwise at the time of the act.But this insight does not require us to ignore very exceptional circumstances, which are typically called 'excuses' .These should be taken into account.To justify some rare exemptions from criminal punishment, it is not necessary to point to 'limited choice' , nor is a character-based logic (arguing that the particular circumstances shed a more favourable light on the offender's character) 71 the only alternative.Rather, excuses can be based on an assessment of mutual rights and duties of the persons involved.Even if the conduct in question is correctly categorised as 'wrong' , the relative proportion of wrong attributable to the offender, in relation to the responsibility attributable to others, might be small enough to forego punishing him.Take excessive self-defence as an example (see § 33 German Criminal Code).The crucial point is that responsibility can to a large part be attributed to the attacker.Although the attacker then became the victim of excessive force, he lost his standing to complain about this because he has created the situation through his unlawful attack.With regard to duress by threats, also this outcome can be largely attributed to another culprit, that is, the person issuing a severe threat.Somewhat more complicated is the explanation why duress by circumstances (see § 35   See Gardner 1998.interaction between offender and victim and the mutual duties involved should in some cases be supplemented by other reasoning.If a threat to the offender's life, or the life of persons close to him, was the reason for the crime, blaming him for the crime means to demand 'you should have sacrificed yourself or the person close to you' .Because the state must not demand that citizens sacrifice their lives, this can be reason to forego criminal punishment despite the fact that the offender has done wrong to the victim of the crime. However the rationale of norms commonly called 'excuses' is explained in detail: it is not necessary to keep a category labelled 'guilt' within the evaluative scheme (the Allgemeine Verbrechenslehre) in order to accommodate such rules.In substance, they should be treated as sentencing rules.The consequence 'no punishment' is appropriate if only a small proportion of the wrong done is attributable to the offender, such as in the case of excessive self-defence or in the case of threat to life.In other cases, the better solution might be a mitigation of punishment.Legislatures should consider re-designing the relevant norms with inflexible outcomes.More flexible norms that allow substantial mitigations (this is the Norwegian solution for excessive self-defence and duress, § 80 lit.d), e) Norwegian Criminal Code) seem preferable.

Mitigation of Punishment Because of Offenders' Pathological Mental States
Under German law, in criminal trials for serious accusations, it is not uncommon that experts in forensic medicine and forensic psychiatry testify concerning whether offenders' abnormal mental states had substantially diminished their capacity to appreciate the unlawfulness of their actions or to act in accordance with any such appreciation.Other legal systems also allow for mitigation, with somewhat different requirements concerning mental impairments. 72However, such mitigation clauses deserve critical review.It is necessary to establish, as a pre-requisite for criminal proceedings, that the offender could have participated in a communication about the relevant factual and normative issues at the time of the offence (see 5.1).However, if this threshold has been reached, further attempts to assess mental states in a fine-grained way need not be undertaken.

Volition
Another tricky subject is the role of volition in criminal law.In modern legal systems, intention plays a pre-eminent role and volition features centrally.The fact that the defendant wanted a certain outcome is not always a necessary condition for criminal liability 72 Stuckenberg 2016, in this issue, at 63; see also § 80 lit.f) Norwegian Criminal Code requiring substantial deficits in the offender's grip on reality due to severe mental disease.
-recklessness or negligence can be sufficient to classify the act as a crime.But for both labelling the crime (for instance: conviction for murder) and sanction severity, the law often relies on the difference between recklessness and purpose.This corresponds with the widespread moral intuition that 'having wanted this outcome' matters when judging behaviour. 73A detached assessment, however, invites the question: why do so many assume this?The concept of volition as discussed within the cognitive sciences is much more complicated than folk psychology assumes.Focusing on conscious will as the central feature of decision-making might be an exaggeration, simplification or perhaps even a misrepresentation of its actual relevance. 74It is doubtful whether the idea of 'mental causation' should be afforded the central role it has had in traditional moral philosophy and the law.
On a descriptive level, several reasons can account for the stubbornness of intuitions about the importance of volition.Subjectively, we constantly perceive ourselves as mental causators -with all mental processes under the control of our will.It is an easily made mistake to accept subjective experiences straightforwardly as evidence.Another explanation can perhaps point to normative, deep structures in societies with monotheistic religions that tend to model human moral judgment parallel to the perspective of a judgmental God. 75From this point of view, a human being's inner act of rebellion against God's commands is the most important feature -the actual harm done to another human being matters less.However, this perspective should not be applied to state punishment.The wrong done to a fellow-citizen must be the basis for blame expressed by state officials, not the offender's guilty mind as such, or, more precisely, snippety accounts of volition -even the most honest and thoughtful persons cannot provide an accurate account of their own mental state.
Would a 'victims-as-co-citizens-based' method of evaluating crimes mean to judge crimes exclusively according to the severity of harm done or the risk of harm?No.This is not the consequence of shifting away from an out-of-the-world version of third-person-perspective.The description of circumstances must be richer than merely measuring the harmful outcome or the probability of it.Two acts that have caused the same kind of outcome, such as the wounding of another human being, still can and should be judged differently.The difference that Antony Duff describes with the categories 'attack' and 'en- dangerment' 76 is of high importance, although Duff argues that the central feature of attacks is intention. 77One could describe those categories from an outside perspective without necessarily making a reference to volitions.The crucial part is how a hypothetical observer of the actual act and its surrounding circumstances would describe this event.The description 'attack' refers to conduct that an observer would describe as directly targeting another person in a hostile way.We typically infer the actor's purpose to hurt from such descriptions.However, volition might only be a typical accompanying circumstance without being necessary for the assessment of wrongdoing.I suspect that volition stands as a proxy for something else, this something else being the high blameworthiness of targeted, aggressive, fear-creating violations of the personal sphere of others.If assessments of offences' seriousness are based on the importance of citizens' duties and the corresponding seriousness of violating this duty, the observer's description 'an attack' is more important than the volitions that might be inferred from this description.It is our primary and most important duty not to attack others physically, particularly not with dangerous weapons.The violation of this duty is a serious one and thus leads to the assessment 'serious crime' , for instance in the case of Mrs. Miller in my example.It is feasible, and in my view recommendable, to de-psychologise the criminal law with regard to the importance of volitions.

Summary
The central role of 'guilt' in German criminal law and, more generally, the well-established focus on choice in criminal law theory across legal systems should not be taken for granted.Closer attention to the conditions of human decision-making reveals that the Principle of Alternate Decisions is not a suitable basis for blame.However, it does not follow that the criminal law should be abolished or replaced with a system of preventive measures.Neither the criminal law's legitimacy nor the content of blame expressed with criminal convictions depend on the Principle of Alternate Decisions.Offenders are legitimately punished for the violation of duties towards their fellow citizens.
The approach defended here belongs to the wider group of compatibilist theories; however, it deviates from the mainstream of compatibilist thinking, which argues 'nothing needs to be changed' .The rules and doctrines in criminal law need to be scrutinised.Some changes will be necessary once 'snapshot'-timeframe notions of choice are given up and the wrong done to victims, that is, the violated duties towards co-citizens, becomes the focus and benchmark for de-psychologised offence evaluations.Criminal law doctrine can be constructed without a category labelled 'guilt' .What is traditionally called 'exemptions from guilt' can largely be re-formulated as conditions necessary for communication with offenders.More intense inquiries into the mental states of offenders are not the task of the criminal law.A separate category labelled 'excuses' is unnecessary.Mitigation or waivers of punishment, if, for instance, large parts of the responsibility for the offence can be assigned to other persons, are a matter for sentencing rules.Another important implication is that the central role of volition in criminal law needs to be re-examined.
The proposals made here would lead to a simpler scheme of offence evaluation.I suggest to de-psychologise the criminal law and not to rely on forensic experts for finegrained assessments of 'guilt' .This would simplify criminal trials.Rather than attempting to tailor sanctions as precisely as possible to 'individual guilt' , the overall level of punishment should be kept moderate -to acknowledge the abundance of unfavourable circumstances in all offenders' lives.

19
See for instance Conscious Will and Responsibility: A Tribute to Benjamin Libet, eds.Sinnott-Armstrong and Nadel (Oxford University Press 2010); Roskies, The Neuroscience of Volition, in Decomposing the Will, eds.Clark, Kiverstein and Vierkant (Oxford University Press 2013) pp.33-59, at 39.

30
See for this objection Hillenkamp 2015, 46. 31 See for the different contents of the constitutional Schuldprinzip Hörnle, Die verfassungsrechtliche Begründung des Schuldprinzips, in Festschrift für Klaus Tiedemann, eds.Sieber et al. (Carl Heymanns Verlag 2008) pp.325-343, at 340-343.32 In his impressive analysis of medieval sources and the development of our modern concept of guilt, Harald Maihold focuses on this aspect: punishment should not be imposed for others' guilt but only for 'one's own guilt': Maihold, Strafe für fremde Schuld?Die Systematisierung des Strafbegriffs in der Spanischen Spätscholastik und Naturrechtslehre (Böhlau Verlag 2005).

73
See for example Moore, Intention as a Marker of Moral Culpability and Legal Punishability, in Philosophical Foundations of Criminal Law, eds.Duff and Green (Oxford University Press 2011) pp.179-205.74 See Wegener, The Illusion of Conscious Will (MIT Press 2002); Bargh, Bypassing the Will: Toward demystifying the nonconscious control of social behaviour, in The New Unconscious, eds.Hassin, Ulemann and Bargh (Oxford University Press 2005) pp.37-58; Roskies 2013 p.51. 75 See for the historical development of our modern notion of mens rea among medieval Christian authors Maihold 2005 pp.174-176, 211-213.
mention a few examples: Mele and Robb, Rescuing Frankfurt-Style Cases, 107 The Philosophical Review (1998) pp.97-112; Widerker and McKenna (eds.),Moral Responsibility and Alternative Possibilities: Essays on the Importance of Alternative Possibilities (Ashgate Publ.2006); Fischer, The Frankfurt Cases: The Moral of the Stories, 119 The Philosophical Review (2010) pp.315-336.See, for the emergence of this idea over time Lacey, Responsibility and Modernity in Criminal Law, 9 Journal of Political Philosophy (2001) pp.249-276.
7 8 See Loughnan, Asking (Different) Responsibility Questions: Responsibility and Non-Responsibility in Bergen Journal of Criminal Law and Criminal Justice, this issue, pp.25-47, at 29-30.
Scientists and philosophers have ar- unfair -the convicted person (like any other person) could not have decided otherwise.

51
Different versions of compatibilist positions have been proposed (among others) by Duff 2011; Morse, Common Criminal Law Compatibilism, in Neuroscience and Legal Responsibility, ed.Vincent (Oxford University Press 2013) pp.27-52; Mackor, What Can Neuroscience Say About Responsibility? in Neuroscience and Legal Responsibility, ed.Vincent (Oxford University Press 2013) pp.53-84; Bröckers 2015; Moore 2016.Hörnle 2013 pp.69-78 (however, in the present article I propose more changes to the criminal law than I thought initially to be necessary).53 This is Thomas Hillenkamp's diagnose, Hillenkamp 2015 p.55.Among other things, he points to 'voluntariness' , a notion that figures in different sections of criminal law doctrine, for example concerning consent, ibid at 44-45.The term 'voluntary' should remain part of the criminal law, but it can be interpreted as lack of coercion.This makes sense independent of the Principle of Alternate Decisions.
German Criminal Code, excusing necessity) should matter.If a life-threatening situation has natural causes, dividing persons' responsibilities is not possible.But one can argue that assessing the 68 See for the new Norwegian Criminal Code (in force since October 1, 2015): Jacobsen and Sandvik, An outline of the new Norwegian Criminal Code, 3 Bergen Journal of Criminal Law and Criminal Justice (2015) pp.162-183 and for the insanity rules Gröning and Rieber-Mohn, NOU 2014/10: Proposal for new rules regarding criminal insanity and related issues, Norway post-22 July, 3 Bergen Journal of Criminal Law and Criminal Justice (2015), pp.109-131.I have used the German translation by Cornils and Husabø, Das norwegische Strafgesetz -Lov om straff (straffeloven) vom 20.Mai 2005 nach dem Stand vom 1. Juni 2014 (Max Planck Institut für ausländisches und internationales Strafrecht 2014).