42  Tomorrow Never Comes

Apathy, Abulia, and the Failure to Initiate

42.1 A represented future does not initiate itself

Chapter 41 ended at a threshold. Exploration can make an option available. Valuation can make it desirable. Prospective bodily change can make its consequences felt, and maintained control can keep the option effective after its cue has disappeared. None of those operations guarantees that action will begin.

A person may know that a bill must be paid, explain what will happen if it is ignored, and possess the money and motor capacity needed to pay it. The envelope can remain unopened. A patient may agree that walking would preserve strength, describe the route down the corridor, and stand when firmly asked. Left alone, the patient does not rise. The possible action is not absent. It has failed to organize behavior.

This gap is often hidden by the word motivation. We say that someone wanted an outcome enough to act or did not want it enough to begin. The word compresses several problems. The outcome must be represented. Its expected benefit must matter in the current bodily and social state. Effort, delay, uncertainty, and risk must be tolerable. One policy must gain priority over alternatives. The organism must prepare to expend energy, release a starting action, sustain the course, and resume after interruption.

The syndromes in this chapter separate some of those operations. Apathy reduces goal-directed activity. Abulia produces a more severe loss of initiative and spontaneous behavior. Auto-activation deficit makes thought and action strikingly dependent on external stimulation. Akinetic mutism can leave a person awake but producing almost no spontaneous speech or movement. Their boundaries are not perfectly stable, and different injuries can produce partly different mixtures. Together they establish one central fact:

A represented future does not initiate itself. An available goal must acquire enough control to recruit effort, bodily preparation, action selection, motor release, and persistence.

The chapter’s title comes from an observation made during the era of prefrontal lobotomy. Some operated patients could discuss plans for the next day, yet the plans did not organize what they did. “Tomorrow never comes,” Freeman and Watts wrote [@freemanwatts1939frontal]. The phrase remains clinically revealing if it is treated as a metaphor. Tomorrow may still be available in speech. What fails is its power to reorganize today.

42.2 Apathy, abulia, auto-activation deficit, and akinetic mutism

The broadest term is apathy. An international consensus group defined it as a quantitative reduction in goal-directed activity relative to the person’s previous level of functioning. The reduction can appear in behavior and cognition, emotional responsiveness, or social interaction [@robertetal2018apathy]. This definition is deliberately descriptive. It identifies a change that can be observed without assuming that every case has the same cause.

Apathy is not simply laziness, weak character, or a failure to understand what should be done. It is also not identical to anhedonia, the reduced capacity to experience pleasure, although the two can coexist. A person may enjoy an activity after someone else starts it yet rarely initiate it. Another may begin tasks but derive little pleasure from their outcomes. Apathy can accompany depression, but it can also occur without pervasive sadness, guilt, hopelessness, or suicidal thinking. Conversely, the absence of reported sadness does not prove that every other affective or cognitive process is intact [@husainroiser2018apathy].

Abulia usually refers to a more marked reduction in initiative. Speech becomes sparse, answers are delayed, and behavior may stop unless another person repeatedly prompts it. Basic movement can remain possible on command. The distinction from severe apathy is one of degree and pattern rather than a universally accepted boundary.

The older term psychic akinesia is most useful when attached to a more specific syndrome now called auto-activation deficit. In the classic cases, spontaneous thought and behavior were profoundly reduced, while sustained external stimulation could produce unexpectedly elaborate speech or action. Some patients described a mental emptiness when nothing in the environment called for a response. The syndrome was first associated with bilateral basal-ganglia lesions and was later reported after frontal and bithalamic injury [@laplaneetal1984psychic; @laplanedubois2001autoactivation].

Akinetic mutism describes an extreme paucity of spontaneous movement and speech in a person who appears awake. The eyes may open and track. Elementary movements may sometimes be elicited, but the person rarely initiates them. The syndrome can follow bilateral injury involving anterior cingulate and supplementary-motor territories, but it also occurs after thalamic, basal-ganglia, midbrain, brainstem, or connecting white-matter damage [@nemethetal1988akinetic; @darbyetal2018volition].

The terms overlap, but they are not interchangeable. One influential framework separates emotional-affective, cognitive, and auto-activation routes to reduced goal-directed behavior, while emphasizing that several can coexist in one person [@levydubois2006apathy].

Clinical term	Characteristic reduction	What may remain accessible	Main caution
Apathy	Goal-directed activity across behavioral, cognitive, emotional, or social domains	Many capacities may remain, with variable benefit from structure or prompting	A broad syndrome with several possible component failures
Abulia	Initiative, spontaneous speech, and self-started action	Wakefulness, comprehension, and movement on command may be relatively preserved	Boundaries with severe apathy and akinetic mutism vary across traditions
Auto-activation deficit	Internally generated thought and behavior	Sustained external stimulation can produce a striking temporary restoration	Most closely related to classic psychic akinesia, not a synonym for every abulic state
Akinetic mutism	Nearly all spontaneous speech and movement	Eye opening, visual tracking, and occasional commanded action can remain	Can reflect failures of arousal, motivation, selection, initiation, or several together

The table is not a severity ladder. A patient can be apathetic yet verbally active. Another can be nearly mute because language production is impaired rather than because initiative is lost. Paresis can prevent action after a goal has been selected. Fatigue can make sustained effort physiologically expensive. Catatonia, delirium, severe depression, aphasia, disorders of consciousness, and medication effects can all produce superficial similarities. Clinical interpretation therefore asks what the person can do under carefully varied conditions, not merely how much spontaneous behavior appears at the bedside.

42.3 When external structure restores action

The classic cases of auto-activation deficit make the role of the environment unusually visible. Laplane and colleagues described patients with bilateral basal-ganglia lesions who showed very little spontaneous behavior. They might sit silently, fail to begin ordinary activities, and report little spontaneous mental content. Persistent questioning or direct instruction could nevertheless evoke appropriate answers and organized actions. The underlying capacities had not vanished. They had become difficult to recruit without external structure [@laplaneetal1984psychic; @laplanedubois2001autoactivation].

Two cases with left caudate lesions supplied another striking clinical contrast. Both patients became markedly abulic after hemorrhagic basal-ganglia injury, with sparse spontaneous speech, prolonged response latencies, emotional indifference, and dependence on repeated prompting. Yet they also showed severe memory, attention, and executive difficulties, and the extent of caudate damage differed between them. Their cases therefore illustrate the chapter’s central problem without proving that initiative alone had been subtracted. Elementary movement could recover while spontaneous, organized behavior remained profoundly impaired, but cognition was not otherwise intact [@benkeetal2003caudate].

That contrast is often described as preserved reactivity with lost self-activation. It should not be translated into a division between the present and the future. An external prompt can concern next week: “Begin preparing for your appointment.” A self-generated action can concern the immediate present: reaching for a glass because one is thirsty. The relevant distinction is whether the environment supplies part of the control structure or the person must generate and sustain that structure internally.

A cue can help in several ways. It can specify the goal, identify the first movement, reduce competition among possible starting actions, increase arousal, divide an extended task into manageable steps, or re-engage the person after behavior has stopped. “Get dressed” may be too underspecified. “Pick up the shirt” supplies a first state transition. Once the shirt is in hand, its affordances and the examiner’s continued prompts can support the next transitions. The cue has not necessarily restored a missing desire. It may have supplied a sequence that the damaged system could no longer organize reliably on its own.

Experiments with self-initiated and externally triggered movement show why the distinction cannot be assigned to one anatomical switch. In a PET and movement-potential study, healthy participants and patients with Parkinson’s disease extended a finger either at a self-selected time or in response to a tone. The two conditions recruited substantially overlapping motor systems. Among healthy participants, right dorsolateral prefrontal cortex was the clearest region distinguishing self-initiated from externally triggered movement. In Parkinson’s disease, self-initiated movement was accompanied by reduced early motor preparation and lower activation in supplementary-motor, anterior-cingulate, putaminal, and related regions [@jahanshahietal1995selfinitiated]. External triggering did not bypass the action system. It changed which information organized that system and when preparation began.

The clinical response to prompting therefore establishes less—and more—than it first appears to. It does not identify a single self-activation center. It does show that the difference between capacity and access can be enormous. A patient who rarely speaks may still produce a coherent account when sufficiently engaged. A patient who does not begin a routine may still execute its components when the environment supplies a starting point and maintains the sequence.

That distinction matters clinically and conceptually. Sparse spontaneous behavior should not automatically be read as absent comprehension, absent preference, or absent personhood. At the same time, a response obtained under intensive prompting does not prove that cognition is otherwise normal. The prompt has changed the task. It has temporarily supplied organization that everyday life usually requires the person to generate.

External structure can make a latent capacity behaviorally accessible without revealing exactly which endogenous operation the structure has replaced.

42.4 From expected value to effort, vigor, and initiation

Chapter 37 showed that an outcome’s value depends on bodily state, learned consequences, delay, uncertainty, and available alternatives. This chapter adds a boundary that value cannot cross by itself. An option can be judged beneficial and still fail to recruit action.

One rare case makes reward sensitivity visible. A man developed profound apathy after focal bilateral lesions involving the globus pallidus. In oculomotor choice tasks, his decisions were unusually insensitive to reward. Levodopa partly increased reward sensitivity; a dopamine-receptor agonist produced a larger change, accompanied by greater social engagement and reduced clinical apathy [@adametal2013dopamine]. The case supports a causal contribution of pallidal and dopaminergic circuitry to one form of apathy. It does not show that dopamine is the chemical source of will. One person, one lesion pattern, and one pharmacological response cannot define every failure of initiation.

A larger study in Parkinson’s disease separated several variables that ordinary language folds together. Participants chose whether to accept offers combining different rewards and physical effort. Apathy was associated especially with weak incentivization by the available reward: low-reward actions often failed to become worth doing. Dopamine medication had a different pattern. It increased acceptance of high-effort, high-reward offers and also increased the vigor of force production after an offer had been accepted. Apathy itself did not produce the same motor-vigor effect [@leheronetal2018effort].

The distinction is important. Reward sensitivity concerns how much an expected outcome changes the attractiveness of an action. Effort valuation concerns what energetic or cognitive cost the organism will pay. Vigor concerns the speed or force with which an action is performed after it has gained control. Initiation concerns whether a starting action is released. Persistence concerns whether the course continues when reward is delayed, progress is uncertain, or interruptions occur. These variables interact, but none is a synonym for the others.

A failure can therefore occur at several points:

represented outcome → state-dependent value → effort and risk → policy selection → bodily and motor preparation → initiation → persistence → feedback

The arrow is a conceptual sequence, not an anatomical assembly line. Expected effort changes value. Movement feedback can change commitment. Bodily preparation can begin before a final policy wins. Basal-ganglia, frontal, thalamic, hypothalamic, brainstem, and motor systems participate recurrently at several points.

The same caution applies to wanting. A person can report wanting an outcome while doing little to obtain it. Another can work vigorously through a learned routine without experiencing a strong conscious desire. A cue can release an action whose value was already established. A medication can change vigor without correcting every aspect of apathy. There is no single quantity called wanting that is poured into an action until movement begins.

This decomposition also connects the chapter to Chapter 38. Initiation and maintenance are different control problems. One patient may begin repeatedly but abandon each course. Another may fail to start yet continue effectively once externally launched. A third may complete overlearned routines but not construct a new sequence. The apparent unity of motivation dissolves into dissociable operations.

42.5 A distributed initiation system

The anatomy follows the behavioral decomposition. Severe failures of spontaneous action can result from lesions in medial frontal cortex, basal ganglia, thalamus, brainstem, or the white matter linking them. That distribution is not noise around a hidden will center. It is what a recurrent control system predicts.

42.5.1 Medial frontal systems relate goals to control and action

The medial frontal wall includes several territories with different architectures and connections. Anterior cingulate and anterior midcingulate regions receive information about outcomes, pain, bodily state, conflict, effort, and control demand. More posterior cingulate motor, pre-supplementary motor, and supplementary motor territories participate in selecting, organizing, preparing, and sustaining actions. None performs initiative alone.

The clinical evidence is strongest when lesions are bilateral or interrupt connected loops. Németh and colleagues described three patients with akinetic mutism whose common damage involved the rostral anterior cingulate gyri and extended into neighboring supplementary-motor territory [@nemethetal1988akinetic]. Such cases do not mean that anterior cingulate cortex stores motivation. They show that damaging a region positioned between outcome evaluation, autonomic regulation, and motor organization can prevent those variables from cohering into action.

Electrical stimulation provides a complementary observation. Stimulation of anterior midcingulate sites in two patients produced autonomic changes, an expectation that a challenge was approaching, and a determined disposition to meet it. The stimulation sites belonged to a wider network that included frontoinsular, frontopolar, and subcortical regions [@parvizi2013persevere]. The result is not the discovery of a perseverance button. It demonstrates that altering one node can reconfigure bodily state, anticipated demand, and action readiness together.

42.5.2 Basal ganglia, dopamine, and thalamus alter access to action

Basal-ganglia circuits influence which cortical and brainstem controllers gain access to action, how strongly learned consequences bias that competition, and how vigorously selected actions are expressed. Damage to caudate, putamen, globus pallidus, or their connections can therefore reduce spontaneous behavior even when elementary movement remains possible. The classic auto-activation cases make this especially clear [@laplaneetal1984psychic].

The globus pallidus is not an engine that sends desire forward. It is a major output structure within several partly segregated loops. Changes in pallidal output alter thalamocortical and brainstem states. Dopamine modifies learning, reward sensitivity, effort acceptance, and vigor through projection-specific actions. The focal pallidal case and the Parkinson’s disease study show that these contributions can dissociate [@adametal2013dopamine; @leheronetal2018effort].

Thalamic participation is equally important. Mediodorsal, intralaminar, and related thalamic territories help sustain recurrent cortical–subcortical states. Bilateral thalamic injury can produce profound reductions in self-generated behavior, sometimes resembling the basal-ganglia syndrome. White-matter injury can produce a similar result without destroying either cortical or subcortical gray matter directly. The functional unit is the loop.

42.5.3 A prepared organism must be available

An initiated course is more than a motor command. It requires wakefulness, postural readiness, cardiovascular and respiratory adjustment, and a willingness to expend limited energy. Hypothalamic and brainstem systems contribute arousal and autonomic preparation. Midbrain and pontine lesions can produce akinetic or abulic states partly because these systems no longer support the cortical–subcortical configuration needed for sustained action [@darbyetal2018volition].

This does not make apathy a disguised disorder of arousal. Many apathetic patients are awake. It means that arousal is one necessary input among several. A goal can be valuable yet fail to recruit an adequately prepared body. Conversely, increased arousal can amplify poorly selected actions as well as useful ones. Preparation must be coordinated with value, context, and policy.

42.5.4 Different lesions can enter one network

Darby and colleagues assembled 28 published cases in which focal lesions produced abulia or akinetic mutism. The lesions were anatomically heterogeneous. They involved anterior cingulate cortex, globus pallidus, thalamus, caudate, and brainstem. Lesion-network mapping showed that the sites shared functional connectivity with a region of anterior cingulate cortex [@darbyetal2018volition].

The result supports a network account, but its limits are as informative as its conclusion. The cases were collected retrospectively and differed in syndrome, severity, and assessment. Normative connectivity cannot show that the damaged connection was identical in each patient. Most importantly, the method cannot determine whether a particular case reflects reduced value, impaired movement selection, failure to initiate a selected movement, inadequate arousal, or a combination. Common connectivity does not imply one common psychological subtraction.

The distributed account can now be stated precisely:

An available goal becomes action only when interacting systems make it valuable enough, affordable enough, selected enough, physiologically prepared enough, and stable enough to begin and continue.

42.6 The lobotomy record: historically important, anatomically uncontrolled

The title of this chapter came from lobotomy reports, but those reports require unusual care. The operations were destructive, the patients were heterogeneous, and the clinicians who promoted the procedures often judged their own results. The record is historically important. It is not a selective lesion experiment.

42.6.1 Several operations became one name

Egas Moniz introduced prefrontal leucotomy in Portugal in 1935 and published an English-language account in 1937. The procedure disrupted frontal white matter through openings in the skull. Moniz presented it as a treatment for severe mental disorders and claimed a degree of safety and effectiveness that later evidence did not justify [@moniz1937leucotomy].

Walter Freeman and neurosurgeon James Watts introduced their own prefrontal operation in the United States in 1936. Their standard prefrontal lobotomy used a different instrument and plane of section. It was this earlier operation—not the later transorbital method—that produced the 48 cases discussed in their 1939 paper on frontal-lobe function [@freemanwatts1939frontal].

Freeman subsequently promoted transorbital lobotomy, in which an instrument entered through the thin orbital roof. His principal report appeared in 1949 [@freeman1949transorbital]. The cuts produced by this method were neither small nor consistent. Postmortem study showed variable lesions extending through frontal white matter and sometimes involving cortex, vessels, and structures beyond the intended plane [@freemanwilliams1951lesions]. The familiar phrase the lobotomy therefore conceals changing techniques and highly variable damage.

42.6.2 Treatment, discharge, and manageability were entangled

The operations spread through a psychiatric system with severely ill patients, overcrowded hospitals, limited staffing, and few treatments that reliably altered chronic psychosis, severe mood disorder, or relentless agitation. Surgeons and psychiatrists did not describe only one goal. They claimed relief of distressing symptoms, reduced agitation, discharge from hospital, restored family life, and return to employment. Institutional manageability was also an explicit and powerful outcome.

Those goals could conflict. A quieter ward was not necessarily evidence that a patient had recovered. Discharge could reflect family resources and postoperative supervision as much as symptom change. Employment was often treated as a sign of success, embedding expectations about productivity and social conformity in what appeared to be a medical outcome [@pressman1998lastresort; @raz2010psychosurgery]. Institutionalized people and patients dependent on family decisions were especially vulnerable to judgments made by physicians, administrators, and relatives rather than by the patients themselves.

Freeman’s own later follow-up of 3,000 patients illustrates the problem. Outcomes differed by diagnosis, procedure, private versus state-hospital setting, and the support available after surgery. Hospitals operating with discharge as a goal reported different results from hospitals using the procedure mainly to control disturbed behavior [@freeman1957followup]. The variation undermines any claim that one standardized lesion produced one standardized psychological effect.

42.6.3 Outcomes were heterogeneous

Some patients were described as less tormented, less agitated, or able to leave an institution. Others remained severely ill, relapsed, or underwent additional operations. Reported adverse outcomes included seizures, hemorrhage, infection, cognitive impairment, reduced initiative, emotional blunting, disinhibition, socially inappropriate behavior, personality change, and death. Contemporary controlled comparisons already noted uncertainty about surgical plane, lesion extent, therapeutic specificity, and the emotional and intellectual deficits produced [@pressman1998lastresort; @simonetal1951controlled].

The fact that some patients or families reported improvement does not rescue the practice. The procedures were generalized far beyond the quality of the evidence. Consent was often inadequate or coercive. Anatomical damage was irreversible and poorly controlled. Benefits were judged through standards that sometimes privileged institutional quiet, employability, or family convenience over the person’s autonomy and subjective life. Tens of thousands of lobotomies and related psychosurgical operations were performed in the United States during roughly two decades [@pressman1998lastresort].

The ethical conclusion does not require a cartoon in which every practitioner sought only placidity or every patient had the same outcome. The documented history is severe enough: crude destructive surgery was promoted rapidly, evaluated inadequately, and applied to people whose power to refuse was often limited.

42.6.4 A perceptive hypothesis from an uncontrolled record

Freeman and Watts nevertheless noticed something worth preserving. In their 1939 series, they described patients who could complete a task but did not reliably proceed to the next, appeared less troubled by error or failure, and discussed plans that did not govern later behavior. They summarized one recurring impression with the phrase “tomorrow never comes” [@freemanwatts1939frontal].

They proposed that frontal cortex contributes to the “projection of the whole individual into the future”: using past experience, anticipated consequences, and emotional significance to organize action before the outcome arrives [@freemanwatts1939frontal, pp. 536–537]. The formulation was remarkably prospective for 1939. It also bundled together several operations that the rest of this unit has worked to separate—representation, valuation, bodily anticipation, rule application, inhibition, initiation, and persistence.

The historical observation can therefore motivate a modern control problem. It cannot establish that lobotomy selectively removed future representation. The patients had different diagnoses before surgery, different lesions afterward, and variable changes in cognition, affect, social behavior, and movement. There was no matched control group capable of isolating one function. Even the apparent loss of initiative coexisted in some patients with impulsive or disinhibited acts.

The appropriate conclusion is narrower and stronger:

Extensive frontal white-matter injury can disrupt the ability of anticipated consequences and extended goals to organize behavior. The lobotomy record cannot tell us which component failed, where that component was located, or whether the same failure occurred in every patient.

42.7 The patient who could judge afterward

One patient in the Freeman and Watts report makes a narrower dissociation especially vivid. After surgery, he slapped nurses and pulled a fire alarm. When questioned afterward, he could recognize that the acts had been inappropriate. His difficulty, he said, was that “beforehand I couldn’t say whether or not it would be alright” [@freemanwatts1939frontal, pp. 536–537].

The statement contrasts retrospective evaluation with prospective control. Once the act and its social consequences were real, the patient could discuss them. Before acting, the same information did not constrain behavior effectively.

That observation does not reveal why. The patient may have failed to construct the likely consequence, assign it enough value, recruit an anticipatory bodily response, apply a social rule at the relevant moment, inhibit a prepotent action, or keep the consequence effective while action competition unfolded. Extensive frontal disconnection could disturb several of these operations at once.

The case therefore belongs beside Chapter 39, but it does not prove the somatic-marker hypothesis. No physiological signal was measured. The historical report cannot show that a missing bodily warning caused the act or that intact anticipatory arousal would have prevented it. The most that can be concluded is that retrospective knowledge did not become prospectively effective.

The behavior also warns against merging lobotomy effects with abulia. Abulia produces too little spontaneous action. This patient produced actions that were excessive, poorly constrained, and socially inappropriate. A common framework is possible, but the shared failure lies at the level of extended control. In one case, an available goal fails to initiate action. In another, represented consequences fail to restrain it. They are not the same syndrome.

42.8 What “tomorrow never comes” can mean

The phrase should now be heard differently from the way the original lesion story invited us to hear it. It does not mean that an entire future has disappeared from the brain. A patient may still understand calendars, describe plans, remember obligations, or imagine an outcome. The phrase names a failure of transfer: what is represented does not acquire enough control over present behavior.

Several dissociations make that failure possible.

A future outcome can be represented without receiving much current value. It can be valued without overcoming expected effort, risk, or delay. A course can win that comparison without yielding a clear starting action. A starting action can be selected without sufficient autonomic, postural, or motor preparation. Action can begin and then stop. After an interruption, the original goal can remain verbally available while no longer re-engaging the sequence. External structure can sometimes supply one or more of these missing operations.

This is why the syndromes do not reveal a reservoir of will. What ordinary language calls initiative is an achievement of interacting systems. Memory and imagination provide possible states. Orbital, medial, striatal, hypothalamic, and interoceptive systems help determine what is at stake. Lateral systems maintain rules and intermediate goals. Medial frontal systems relate expected benefit, demand, progress, and control to action. Basal ganglia and thalamus influence selection and release. Brainstem and autonomic systems prepare an organism capable of paying the cost. Motor systems act, and feedback changes what happens next.

No chief is missing when one of those relations fails. The anti-homunculus argument rests on dissociation. Representation can survive without initiation. Reward sensitivity can change without an equivalent change in vigor. External prompting can restore a sequence without repairing every deficit. Anatomically different lesions can produce overlapping clinical pictures because they interrupt the same loop at different points.

The language used for the patient also matters. A person does not become more or less “there” according to how much spontaneous behavior remains. Injury can alter initiative, emotional responsiveness, insight, language, memory, or movement while the patient remains a person whose welfare and preferences matter. The ethical failure of the lobotomy era included treating reduced disturbance or increased manageability as though it exhausted that question.

Frontal systems have a major role in the chapter’s control problem, but they do not contain tomorrow. The hippocampus and connected cortex provide scene, event, and relational structure. Semantic memory supplies knowledge of regularities and consequences. The body and hypothalamus determine current need. Amygdala, insula, striatum, and orbital systems contribute learned significance and value. Thalamus, basal ganglia, medial frontal cortex, brainstem, and motor systems help convert those variables into an initiated course.

“Tomorrow never comes” is a clinical metaphor for a future that can remain cognitively available while failing to become behaviorally effective.

That conclusion is less dramatic than literal future subtraction, but it explains more. It accommodates apathy and abulia without pretending that all cognition is intact. It accommodates akinetic mutism without reducing wakefulness to motivation. It accommodates disinhibition without calling it the same deficit as inactivity. It also preserves the central insight of the title: governing present action by an absent outcome requires work that can fail.

42.9 Coda: constructing a future from remembered structure

This unit has asked how an organism becomes governed by information that is not fully specified by the immediate scene. Value can depend on a delayed outcome. A rule can remain effective after its cue disappears. An imagined consequence can change the body. Another person’s possible response can alter present action. Exploration can enlarge the repertoire of available policies. The final chapter has shown that none of these achievements guarantees initiation.

The unresolved first step is representation. Before a future can influence value, effort, or action, some structure of that future must be constructed. The nervous system does not create it from nothing.

Episodic memory contributes details from particular experiences. Hippocampal and connected posterior systems help bind people, objects, spatial settings, and relations into coherent events. Patients with bilateral hippocampal damage can be impaired not only in remembering episodes but also in constructing coherent new experiences, especially their spatial setting [@hassabis2007imagine]. Functional imaging likewise shows substantial overlap between remembering past events and imagining possible future events [@addisetal2007pastfuture].

Overlap is not identity. Future construction places additional demands on recombination, novelty, prospective relevance, and control. Semantic knowledge supplies schemas and facts that need not come from a remembered episode. Spatial and relational maps constrain where events can occur. Current bodily state determines which possibility matters. Frontal and posterior systems help select, combine, and elaborate details. The constructive episodic simulation account proposes that flexible memory supports future simulation; it does not establish that all memory evolved for imagining or that one system performs both operations without distinction [@schacteretal2007prospective].

The transition to Unit 7 is therefore not from a future organ to a memory organ. It is from one control problem to the distributed constructive systems on which that problem depends.

Unit 6 has asked how absent outcomes acquire control over present behavior. Unit 7 turns to the predictive maps and memories from which routes, scenes, events, and possible futures are assembled.

TipEstablished findings, working synthesis, and open questions

Established findings. Apathy is a measurable reduction in goal-directed activity relative to a person’s prior functioning. Abulia and akinetic mutism describe more severe failures of spontaneous speech and action, while auto-activation deficit identifies a striking dependence on external stimulation. These syndromes can follow lesions involving medial frontal cortex, basal ganglia, thalamus, brainstem, and connecting white matter. Reward sensitivity, effort acceptance, motor vigor, action selection, initiation, and persistence can dissociate. Historical lobotomy procedures produced variable lesions and heterogeneous outcomes and cannot be treated as selective experiments on one frontal function.

Working synthesis. “Tomorrow never comes” is a metaphor for a conversion failure. A possible action may be represented, discussed, and even valued without acquiring enough control to recruit effort, bodily preparation, selection, motor release, and persistence. The failure does not reveal a single reservoir of will or the literal removal of future representation. It reveals the distributed work required to turn an available goal into an initiated course of action.

Open questions. It remains uncertain how best to subtype apathy across neurological and psychiatric disorders; which cases arise mainly from reduced reward sensitivity, altered effort valuation, deficient action-state generation, impaired arousal, or failure of motor release; why external prompts restore behavior in some patients; and how dopaminergic and non-dopaminergic systems interact. It is also unresolved how remembered episodes, semantic knowledge, spatial maps, and current goals are combined into future simulations, and how those simulations gain enough control to initiate and sustain behavior.