Free Will in the Light of Neuropsychiatry

Sean A. Spence

Figures

Self awareness is . . . one of the fundamental, possibly the most fundamental, characteristic[s] of the human species . . . an evolutionary novelty; the biological species from which mankind has descended had only rudiments of self-awareness, or perhaps lacked it altogether. Self-awareness has, however, brought in its train somber companions--fear, anxiety and death. Man is burdened by death. A being who knows that he will die arose from ancestors who did not know."

Dobzhansky (1967)

"What complicates life is having to make choices. The person who has his choices made for him lives simply."

Peter Hoeg (1993)

Abstract: If the notion of free will is to be retained by philosophers, psychiatrists and psychologists, then it will be a free will which is essentially non-conscious. The purpose of this paper is to demonstrate that a conscious free will (in the sense of consciousness initiating action) is incompatible with the evidence of neuroscience, and the phenomenology described in the literature of normal creativity, psychotic passivity, and the neurological syndrome of the alien limb or hand. In particular the work of Libet and others who have directly stimulated the brain and measured activity from its surface leads to the conclusion that subjective states (be they sensory or intentional) are preceded by predictive neural activity. Subjective phenomenology is temporally distinct from the initiation of "voluntary" action. In addition it will be argued that free will is more than a belief and that it comprises a perceptual sensorimotor component. Free will is conceptualized as a three component phenomenon consisting of a perceptual, non-linguistic experience of freedom, an experiential belief which is linguistic, and related to a specified action in time, and an abstract belief which is linguistic but remote from any particular action and akin to philosophical discourse (the abstract belief or otherwise, in free will). A theoretical explanation of some of the pathological phenomenology of possession of action is offered in terms of temporal considerations. The direction of action within mental/virtual space is hypothesized to be a function of the temporal sequence of its initiation and awareness of the latter "reaching" consciousness. It is suggested that perturbation of this temporal sequence may lead to the misattribution of host-as-alien-initiated action. Work of this kind, which is beyond subjective introspection, is necessary if we are to develop adequate models to use in conjunction with modern investigation techniques.

The purpose of this paper is to examine the status of free will in the light of evidence accrued from neuroscience. In particular it will question whether traditional notions of free will (relying as they seem to upon conscious awareness) are compatible with the temporal properties of consciousness.

Given the breadth of this undertaking the literature may appear partisan, deriving mainly from the fields of neurology, psychiatry and phenomenology. This is true. The emphasis will be upon examining free will in real case scenarios. It is [End Page 75] hoped that this will lead to some original insights into this area, and that this may provoke debate of what must be one of the central features of human phenomenology--the experience of free will.

Free Will

The following is an example of an attempt to address the problem of free will taken from a philosophical encyclopedia:

The cluster of problems about the freedom of the will arises from an incompatibility. . . .

On the one hand we believe that we can sometimes choose whether to act in a certain way or not; . . . that we are responsible for so acting or refraining from action; . . . that for those parts of our history which do not lie within our choice we cannot be held responsible.

On the other hand we believe that nature is uniform, that whatever happens results from and can be explained by a set of causes and conditions, and in particular that our actions result from our inherited character as modified by environment.

[I]f our actions arise from an inherited character as modified by our environment, then it would seem that we are no more responsible for our actions than we are for our inherited character and environment. (Urmson 113. Paragraphs and italics added by author.)

In this summary the problem of free will is couched in terms of a set of beliefs which may be accepted or rejected according to the evidence offered. In its emphasis upon belief such a definition ignores phenomenological awareness related to specific acts in the immediate present. By focusing exclusively upon belief it places "freedom of the will" at one remove from those actions that it would appear to address. Is free will merely a belief?

I would like to offer another conceptualization of free will which comprises three parts:

1. The experience that we are free to choose our actions (on at least some occasions). This experience of free will occurs at the time of an act and for its duration. It may be conceived of as a perceptual phenomenon.

2. Experiential belief, a belief "attached" to, or addressing, an ongoing experience (and temporally related to the latter).

3. The abstract belief that we are free or not free (which may occur at times remote from specific acts).

These aspects of free will are potentially separate. I may, for instance, experience myself as a free agent, free in my volition, choosing to move or act in a certain way, but I might also believe that I am not free, that my thoughts and acts result from genetic and environmental influences, and that at some ultimate level of analysis I am not as free as I feel (or experience myself).

Such a distinction would separate experience (1) and experiential belief (2) from abstract belief (3). The separation of experience (1) from experiential belief (2) is more problematic but will become apparent when we address pathological states (below).

For the present we will claim the existence of an (immediate) experience of free will. When behaving normally, e.g., lifting my arm, picking up a pen, writing, I have the subjective experience (sensation) of freedom. This is clearly distinct from the sensation of passive movement, e.g., when another lifts my arm. What is less clear is the point at which motor and sensory data, my perceptual experience, acquire a related subjective belief (my immediate experiential belief that I am free).

Part of the problem is that the "experience" of freedom is by its very nature non-linguistic. In the example of moving my arm the sense of volition involved is primarily related to motor activity and proprioception of the latter. This may provide the distinction that we seek. The experience of freedom is perceptual and (essentially) sensory and as such will access varying levels of conscious awareness, or processing. At what point does an experience become a belief? If any propositional statement (e.g., "I am lifting my arm"), is regarded as a belief, then clearly the acquisition of a linguistic label is the point at which experiential belief (2) accrues to experience (1).

Thus, for the purpose of the current paper the "problem of the freedom of the will" breaks down into 3 aspects which are phenomenologically distinct:

1. The experience (sensorimotor) of freedom coexistent with an act (pre-linguistic). [End Page 76]

2. The experiential belief which accompanies such an act (and which is linguistic).

3. The abstract belief occurring remotely from specific acts (which is linguistic and akin to philosophical discourse).

So far we have assumed that each aspect of "freedom" may access consciousness, but there is clearly a necessity to examine what is meant by this term.

Consciousness

For free will to be meaningful must we be aware of our actions? Clearly we do not invoke free will for all actions, especially those that we do not influence or of which we are only minimally aware, e.g., reflexes. A central issue in this paper is the requirement of conscious awareness for the experience of free will. To make a choice it seems appropriate (although this may be questioned) that one should be aware of that choice. (There is a difference between awareness of a choice already made, and awareness which precedes decision and indeed initiates the latter.) Are consciousness and awareness the same? For the purpose of this paper consciousness will be used in its original sense, meaning "inwardly sensible or aware", i.e., awareness of one's own thinking (Wilkes 1988). Thus, it is not being used merely to denote being awake, or non-comatose, but to connote a certain form of mental activity (metarepresentation: thinking about thinking).

Certain assumptions regarding consciousness should be acknowledged from the outset. When we discuss consciousness, as inner awareness, we assume that this will be understood by others, that they will experience a similar inner awareness. Such an assumption is essentially unverifiable: we cannot know another's consciousness. (This question is closely allied to that of theory of mind, the belief that others have minds and will behave as if they do, a faculty which is hypothesized to be deficient in autism [Frith, Morton and Leslie, 1991].)

Similarly we assume that such inner awareness is robust between different cultural groups and across historical epochs. Again, this assumption is unverifiable, and indeed may be undermined by certain linguistic evidence. Wilkes (1988) has emphasized the absence of words for "mind" and "consciousness" from some contemporary languages, and the absence of "consciousness" from English prior to the seventeenth century: "in other words . . . what strikes some of us so forcefully seems to have left little impression on others" (Wilkes 1988). We will return to this question below.

Next, the existence of consciousness as a single entity has been questioned by a number of writers, particularly Dennett (1991). He offers a multiple draft theory of consciousness, with no single current experience or phenomenological "quale" [plural "qualia"], but multiple systems functioning (and experiencing) in parallel. According to this theory an inner state reported verbally gains only apparent precedence above others, but does not truly represent the singular experience of the individual. A unitary experience is illusory (Dennett 1991). Others have questioned Dennett's approach suggesting an error in his equation of the contents of consciousness with consciousness itself; the latter being a single awareness which is independent of many competing items of content (Clark 1992).

Finally, there is the issue of the relationship of "consciousness" to "mind." Are they equivalent or is the former part of the latter. Wilkes (1988) has suggested that we place undue emphasis upon consciousness (to the detriment of "non-conscious") and that this is a product of relatively recent ways of thinking about ourselves. In particular she blames a "Cartesian catastrophe," accusing Descartes of forcing a schism between "conscious" and "non-conscious," creating a dichotomy which is unnecessary.

For Descartes conscious awareness, "I think therefore I am," became evidence for the existence of a soul (Wilkes 1988). Wilkes suggests that this was a new conceptualization and that "consciousness" enters European language (and also phenomenology) at this time.

However, for the same reasons that we cannot verify conscious experience in others or in times preceding our own, neither can it be excluded. For the present author the written report of Descartes suggests that he experienced phenomenology [End Page 77] consistent with what we would regard as "inner awareness." Similarly, records which precede Descartes may not use words such as "mind" or "conscious" but appear to describe experiences which we may recognize or translate in these terms:

The properties of a rational soul are these. She can contemplate herself, analyse herself, make of herself what she will, herself enjoy the fruit she bears . . . and always have her work perfectly complete at whatever moment our life reaches its appointed limit. (Marcus Aurelius AD 121-180)

Not withstanding the assumptions acknowledged above consciousness will be regarded as an "inner awareness" which is experienced by normal individuals and which provides the medium for the experience of free will, and beliefs pertaining to the latter.

A corollary to the above is that there will be mental processes which occur outside conscious awareness and which will be termed non-conscious (a relatively neutral term). For the purposes of this paper the assumption is that this non-conscious is a heterogeneous entity comprising sensory information processing, storage of (semantic and episodic) memory, procedural memory, and the possibility of dynamic or learned complex behavioral patterns which, in psychiatric terms, may be regarded as "defenses." The veracity of elements of this non-conscious will not be debated here. However it is clear that there are components that are described elsewhere as preconscious, subconscious and unconscious by other authors.

The Experience of Free Will: The Normal Case

Under normal conditions many acts may be performed without undue attention or conscious processing being devoted to them. A common example is driving a car while talking or listening to music. Under these conditions behavior is clearly "automatic" to some extent and may be accompanied by very little subjective volition. Indeed it is not uncommon to reach the end of such a process without any conscious awareness of the preceding interval (e.g., the details of a familiar journey).

When activity is complex or novel it receives greater attention, and is more likely to be remembered (Frith 1992). Under these circumstances the sensory aspects of the activity are more likely to be attended to, and awareness of one's own volition increasingly apparent (e.g., the first faltering steps of learning to drive or to play a musical instrument). In this normal situation it can be difficult to distinguish the sensory experience of freedom from the associated experiential belief in the same, but as stated before, there exists a distinction between motor and sensory perceptual phenomena e.g., the sequence of finger movements on a keyboard, and the thoughts (beliefs) which address these phenomena.

When anxious an individual may become acutely aware of his own actions, interrupting the flow of normally fluent activities e.g., walking and speech. In such a condition the experience (perception) of voluntary activity is heightened, but there is not necessarily any change in the experiential belief in the freedom of a specific act.

In normal individuals engaged in creative activities there is frequently a phenomenological account which describes altered volition. Such cases have been documented elsewhere by Boden (1992) as evidence for non-conscious processing producing novel computations in response to creative problems. These solutions subsequently access consciousness. The following descriptions taken from different sources speak of an alteration in the phenomenology of volition.

A1. "Sometimes when I'm singing there's no sensation in my throat. It feels as if some outside influence has taken over. The air is coming through me--as if I'm being sung like a flute." Sheila Chandra (quoted in "The Wire", May 1994)

A2. "I just play what comes into my mind." Thelonious Monk (quoted in Jazz at Ronnie Scott's,1979)

A3. "So . . . I did one of those things you do and never find out why, even though something in you knows you could never have done anything else." William Gibson (1993)

A4. "When you blurt something out and haven't censored yourself, your first instinct is to cover up because it doesn't make sense, rather than just letting it be there. It's a mistake to censor those kind of irrational [End Page 78] utterances. They are often the deepest things you say." David Byrne (The Wire, 1994)

A 5. "How can I tell what I think until I see what I say?" E. M. Forster (1960)

A 6. "It was late before the two guests left and Russell was alone with Lady Ottoline. They sat talking over the fire until four in the morning. Russell, recording the event a few days later, wrote, "I did not know I loved you till I heard myself telling you so--for one instant I thought 'Good God, what have I said?' and then I knew it was the truth.'" Clark (1975) quoted by Dennett, 1991, 246

A 7. "I was sitting in a chair in the Patent Office at Berne when all of a sudden a thought occured to me: 'If a person falls freely he will not feel his own weight.' I was startled. This simple thought made a deep impression on me. It impelled me toward a theory of gravitation." Albert Einstein, quoted in Highfield and Carter (1993)

In these accounts from normal individuals there are circumstances described in which the subjective experience of voluntary action is diminished or lost, and experiential belief is altered. They comprise altered motor performance e.g., singing, playing the piano, speaking, and also, in the case of Einstein, insightful thought (or "inner speech"). None of the individuals attribute the phenomenon to an outside force, although, in the case of A1 there is clearly a metaphorical allusion to such passivity ("as if I'm being sung like a flute"). However, despite the obvious fact that they are all physically responsible for their respective acts they do not experience as freely willed action the impulses which produced these acts.

Thus, in "normals" it appears possible for the experience of free will to be heightened, diminished or lost, and for the experiential belief in free will (pertaining to a specific act) to be altered (but nevertheless retained). Any allusion to outside forces remains metaphorical and does not reach delusional intensity.

Pathological Cases

A variety of pathologies underlying neuropsychiatric illnesses may disturb "willed" action. From these "natural experiments" it may be possible to infer aspects of normal functional organization. In some cases the pathology may be so devastating as to leave little prospect of willed movement, e.g., the limbs affected by dense hemiparesis, or the "off" state of Parkinson's disease. Similarly in severe psychiatric illness the disturbance may be so profound as to offer little prospect of teasing apart aspects of volition e.g., the depressed individual with both psychomotor retardation and nihilistic delusions: "I am dead; I have no feelings and no will" (Sims 1988, 235).

In such a case we may clearly hypothesize that mechanisms of willed action and their representation in the subject's consciousness are disturbed. But is this disturbance one of higher processing, the abstract belief in freedom, or is it perhaps "lower," related in some part to an experience of impeded action, a failure to initiate volition at some level?

In an attempt to dissect volition some specific examples will now be addressed.

1. The Alien Hand Syndrome

The "alien hand" syndrome is a rare but potentially informative example of a pathological process affecting willed movement. Initially the term was used to describe the inability of the affected subject to identify their (affected) limb (Brion and Jedynak 1972). However, in later descriptions the emphasis shifted to the subject's inability to control the latter (Gasquoine 1993). The limb is experienced as being autonomous. The syndrome is usually found in association with lesions of the medial frontal lobe (contralateral to the affected limb) and anterior corpus callosum (Gasquoine 1993). (However, it is common for the pathology e.g., cerebral infarction or tumor to encroach on neighboring structures.)

The striking phenomenology encountered has been reported by a number of authors:

B1. "[A]t one point it was noted that the patient had picked up a pencil and had begun scribbling with the [affected] right hand. When her attention was directed to this activity, she reacted with dismay, immediately withdrew the pencil, and pulled the right hand to her side using the [normal] left hand. She then indicated that she had not herself initiated the original action of [End Page 79] the right arm. She often reacted with dismay and frustration at her inability to prevent these unintended movements of the right arm. She experienced a feeling of dissociation from the actions of the right arm, stating on several occasions that 'it will not do what I want it to do.'" (63 year old right handed female with left medial frontal infarction; Goldberg et al 1981)

B2. "He viewed these (autonomous) actions as unwanted, unintentional, and uncontrollable. He described the [affected] right hand as "the bad one, it has a mind of its own," and that it was 'always trying to get into the act.' A very specific complaint, which was chronic and persisted for a year after the infarction, was that the patient felt the right hand anticipated future actions and performed movements prior to the patient actually intending them. The [normal] left hand did not show this tendency." (68 year old right handed male with left medial frontal and callosal infarction; Feinberg et al 1992)

B3. "Astonishment and frustration were expressed regarding her inability to voluntarily release her grasp of objects in the [affected] left hand. She would attempt to restrain the unwanted movements of the left hand by keeping her hands folded together or by gripping an object in the left hand. The left hand would perform activities with objects around her in a compulsive way that was not associated with a feeling of voluntary control. . . . Her left hand was noted to wander around in the 'alien mode' when not restrained. . . . Another incident occurred when the [normal] right hand picked up a bowl of hot soup and the left hand threw it to the ground. On another occasion, her left hand began to remove a cigarette from her mouth as she was about to light it with the right hand. She stated that the left hand 'was trying to keep me from smoking.'" (53 year old right handed female with right medial frontal lobe and corpus callosum infarction; Goldberg and Bloom, 1990)

B4. "She often spoke to her own [affected] left hand, asking it to perform some movement, but the hand 'did only what it wanted to.' In bimanual tasks (e.g., lighting a match) she tried to use only her [normal] right hand but was regularly impeded by her left hand's groping. At that time she experienced the lack of ownership of her left hand and could not indicate her own among others' hands.." (65 year old right handed female with right medial frontal infarction extending to anterior cingulate gyrus, supplementary motor area and anterior fibres of corpus callosum; Trojano et al 1993)

B5. "[S]he noticed that her [affected] left arm began to move 'on its own.' The left hand would tenaciously grope for and grasp any nearby object, pick and pull at her clothes, and even grasp her throat during sleep. These autonomous left arm and hand movements tended to occur in bouts usually signaled when her hand did 'silly things' such as scratching her knee. She slept with the arm tied to prevent nocturnal misbehaviour. She never denied that her left arm and hand belonged to her, although she did refer to her limb as if it were an autonomous entity." (39 year old right handed female with trauma to medial frontal white matter bilaterally, corpus callosum, right basal ganglia, internal capsule and thalamus; Banks et al, 1989)

B6. "Sometimes the [affected] left hand would pick up a cup and move it into the path of the [unaffected] right hand so as to interfere with eating. While playing checkers on one occasion, the left hand made a move that he did not wish to make, and he corrected the move with the right hand; however, the left hand, to the patient's frustration, repeated the false move. On other occasions, he turned the pages of a book with one hand while the other one tried to close it; he shaved with the right hand while the left one unzipped his jacket; he tried to soap a washcloth while the left hand kept putting the soap back in the dish; and he tried to open a closet with the right hand while the left one closed it." (40 year old right handed male with ruptured anterior communicating aneurysm, and subsequent resection of right frontal gyrus rectus; Banks et al 1989)

As may be seen from these accounts the alien hand is experienced as behaving in an autonomous and purposeful way. Its interference with normal activity is such as to provoke patients to developing strategies to restrain or "distract" the hand. The patients acknowledge that the limb belongs to them but that its actions are not under their control. Yet they do not attribute its activity to outside forces (cf. passivity, below). Their appreciation of their predicament suggests that the subjects retain a sense of their own will, in that they form action plans (which fail). There is a circumscribed inability to impose that will upon one domain of their body/action, i.e., the affected limb.

Thus it is clear that the experience of free will may be absent with regard to part of a subject's anatomy and repertoire of voluntary behavior (limb movement). This is particularly noticeable in those examples where these activities occur during sleep (e.g., B5). Experiential belief relating to the action of the "alien" limb is clearly disturbed. [End Page 80] Once the aberrant activity of the limb is noticed the patient tries to exert conscious control but cannot. The limb is experienced then as having a "mind of its own." This belief is not delusional since it retains an "as if" quality, and is clearly supported by the subjectively and objectively unusual behavior of the limb.

The subject loses the experience of free will (with respect to the affected limb), and develops an accommodating abnormal experiential belief that the limb is autonomous.

That these subjects have sustained lesions of their medial frontal lobe and anterior corpus callosum is of interest, given the role of these areas in the genesis of willed action (see "Temporal Aspects of Action", below) and information transfer between the hemispheres. We will return to the issue of the timing of initiation of activity, but for the present it is interesting to note example B2 in which the subject experienced his right hand as anticipating his intentions. In this case the autonomy of the affected limb is further extended (to anticipating the will of the subject).

What does the alien hand tell us of free will? It suggests that the sense of free will accruing to action may have a neurological component which is discrete from (but related to) the neurological correlates of action itself; in other words that the initiation of action and the "labelling" of that action as "intentional" ("willed" or "belonging to the subject") are potentially separate.

2. Passivity Phenomena

Following this line of argument we will consider passivity phenomena. These phenomena, comprising "made" thoughts, actions, affects, and impulses (together with somatic passivity) are included in Schneider's first rank symptoms of schizophrenia (Schneider, 1974, first published 1957). The subject experiences their thoughts, actions, feelings and drives as having been influenced or replaced by those of external agencies. These symptoms were grouped as a diagnostic aid, independent of any specific pathology. The symptoms themselves are not specifically confined to schizophrenia, but are found in a variety of organic diseases (an exclusion criterion in Schneider's original paper), affective psychoses, and dissociative states. Trimble (1990) has suggested that they are localizing signs of temporal lobe pathology.

The relevance of these symptoms in the present context is that they clearly implicate a disturbance of the individual's experience of "self," and "willed action." Whereas they are classified as delusions (and hence abnormal beliefs), I would like to question the exclusivity of this definition. Might not passivity phenomena be at least partly due to abnormal experiences?.

Patients with delusions of alien control (passivity) perform abnormally on paradigms designed to test "self monitoring." In Frith and Done's (1989) study of performance of a motor task in the absence of visual feedback, schizophrenics with delusions of control were significantly less likely to make error corrections than either "normals" or other schizophrenics (without such delusions). This result is consistent with a disturbance of monitoring of self-generated acts. Other authors have reported disturbance of error correction in schizophrenia (Malenka et al 1982), and aberrant cortical potentials preceding willed action (Timsit 1970).

Thus we have some empirical indication that schizophrenic subjects (reporting passivity phenomena) might be "monitoring" (and perhaps "experiencing") their own volitional processes in an abnormal way.

The patients' own accounts appear to reveal two aspects: firstly an alienation from their own volitional processes e.g., movement, or thought; and, secondly, a delusional elaboration of an often bizarre quality which "seeks" to explain the former. Some examples are given below, the first of which reflects the bewildering nature of the disorder as experienced by the subject.

C1. "A former patient of the Heidelberg clinic, a well-educated individual who a little later became acutely ill with schizophrenia, wrote an account of these "made" phenomena: 'On the following morning I was put into a most peculiar mood by this machine or whatever it is. . . . All the night through I was fully sensible and quite clear . . . the machine . . . was fixed in such a way that every word I spoke was put into me electronically and I could of course not avoid expressing the thoughts in [End Page 81] this peculiar mood. . . . A joyous mood was put into me which prevented me from thinking of anything. At times I was given back my natural way of thinking. . . . I feel the machine is getting me down mentally more and more and I have several times asked for the current to be turned off and my natural thinking returned to me. . . . One evening the thought was given to me electrically that I should murder Lissi. . . . [T]o a large extent these are not my own thoughts. . . . The remarkable machine can also suddenly give me sleep . . . and withhold sleep . . . give me any kind of movement . . . also when I read, whatever it is, I cannot give sufficient attention to the contents and with every word I get a side-thought--I want to make another point . . . when one reads all this it seems the greatest nonsense ever written but I cannot say anything else except that I have really felt all this, though unfortunately I have never understood it.'" [Jaspers 1963; original italics, 579-580]

C2. "A 29 year old shorthand typist described her actions as follows: 'When I reach my hand for the comb it is my hand and arm which move, and my fingers pick up the pen, but I don't control them. . . . I sit there watching them move, and they are quite independent, what they do is nothing to do with me. . . . I am just a puppet who is manipulated by cosmic strings. When the strings are pulled my body moves and I cannot prevent it.'" (Mellor, 1970; 17-18)

C3. "A patient said: 'I never shouted, it was the vocal cords that shouted out of me. . . . The hands turn this way and that, I do not guide them nor can I stop them.'" [Jaspers 1963, 23]

C4. "A 26 year old engineer emptied the contents of a urine bottle over the ward's dinner trolley. He said, 'The sudden impulse came over me that I must do it. It was not my feeling, it came into me from the X-ray department, that was why I was sent there for implants yesterday. It was nothing to do with me, they wanted it done. So I picked up the bottle and poured it in. It seemed all I could do.'" [Mellor 1970, 17]

C5. "A 29 year old housewife said, 'I look out of the window and I think the garden looks nice and the grass looks cool, but the thoughts of Eamonn Andrews come into my mind. There are no other thoughts there, only his. . . . He treats my mind like a screen and flashes his thoughts on to it like you flash a picture.'" [Mellor 1970, 17]

These subjects report a range of passivity phenomena: made actions (C1, 2 and 3), impulses (preceding acts; C4), affect (C1), and thoughts (C1 and 5). As well as experiencing a separation from the initiation of their own actions and thinking, they also explain these phenomena in delusional terms, e.g., an electronic machine (C1), cosmic strings (C2), the X-ray department (C4) and a television personality (C5). They lack the experience of freedom, and derive an abnormal experiential belief.

From our previous consideration of the alien hand syndrome, cases C2 and C3 are particularly interesting given their superficial resemblance to phenomenological reports of the former (e.g., B1). Again we see subjects who do not experience a sense of volition, despite acts occurring which appear purposeful. A difference is seen when the experience is further elaborated in a delusional sense e.g., "I am just a puppet who is manipulated by cosmic strings" (C2).

Example C4 is more subtle. The subject experiences an alien impulse to act prior to his own initiation of the act (emptying the urine bottle). He seems to acknowledge that he performed the act (experiential belief), but he did not experience it as a freely chosen act ("It seemed all I could do"). Once again we may make a case for this phenomenon resembling that alien hand case described above (B2) in which the subject experienced his intentions as being anticipated by his (affected) hand.

The purpose of this exposition is not to attempt to "localize" passivity phenomena or their neural correlates but to demonstrate the similar patterns of disruption to willed action which may be seen in neurological and psychiatric disorders, and also their particular characteristics which may cast light upon normal activity. As we have seen above (the normal case), normal individuals may experience their freedom to act in a variety of ways, which at times may appear to approach automaticity or passivity (without the delusional intensity). In particular the examples of creativity given suggest that subjects may experience their subjective consciousness as the "recipient" of formed cognitions or acts. In this regard consider Einstein's report (A7). Again there are similarities with the passivity of thought described by the subject in example C5 (above), in that thought is recognized as "arriving." But the delusional elaboration of the schizophrenic subject distinguishes [End Page 82] the experience from normal. This point is being labored at present, for we shall return to it when we consider the "timing" of neural events. Can a thought or act be "willed" consciously? For the present we can say that in both "normal" individuals and those with neurological and psychiatric disorders, it is possible for thought and action to be isolated from the sense of "will" normally associated with them. The symptomatic elaborations may vary, but it would seem that we have sufficient evidence for suggesting that seemingly "purposeful" action and "insightful" thought may be neurally distinct from the sense of "will" or "ownership" which normally attends them. This is hardly a new idea:

"A man's thoughts are sometimes not due to himself but come without his will." Philo, De Mutatione (Clark 1991, 226)

"We speak, not only to tell others what we think, but to tell ourselves what we think." Hughlings Jackson (Dennett 1991, 194)

The Temporal Aspects of Action

Does a meaningful concept of free will require that we may, at least on some occasions, consciously choose a certain course of action? Must choice and decision be conscious? If so, what temporal relationship must pertain between conscious awareness of intent and the initiation of action?

As alluded to in the introduction, there is an assumption embedded within the concept of free will that conscious choice may occur (or appear to occur) on at least some occasions. For this to be a real choice, however, it is here argued that the choice made, e.g., a decision to act, would of necessity be prior to the initiation of the specified act. Initiation would "need" to follow decision. By decision here we refer to the conscious "making" of a decision to act. Is this sequence of events a biological possibility?

For the sake of example I will assume that observable motor acts provide a model for all acts, mental or physical. Although this is an assumption I hope to justify it in due course.

Voluntary finger movement in the human is preceded by a well documented pattern of cerebral electrical activity measured from surface electrodes (Kornhuber and Deecke 1965; Deecke, Scheid and Kornhuber 1969; Libet et al 1982). There are 3 potentials identifiable using electroencephalography (EEG):

1. Readiness Potential (RP): a slowly increasing, negative potential occurring 850 msec prior to voluntary movement (as recorded by EMG). The RP is bilateral, symmetrical, and seen over pre-and post-central regions, being maximal at the vertex.

2. Pre-motion Positivity (PMP): a symmetrical positive potential seen 86 msec prior to voluntary movement.

3. Surface negative motor potential: maximal over the contralateral, pre-central hand area, 56 msec prior to voluntary movement (Deecke, Scheid and Kornhuber 1969).

Thus the early stages of the initiation of a motor act precede its objective execution by approximately 850 msec. At what point does the subject become subjectively aware of the "decision" to move? If the voluntary decision is itself initiating the movement, then one would intuitively suggest that the former precede the latter in time. (One would then need to explain how a "thought" might "produce" such an action.) If, however, the neural substrates of action precede the subjective decision to act, then clearly the latter cannot be causal to the former (at least in terms of our current understanding of causality), i.e. conscious "decisions" do not cause acts.

Libet et al. (1983) addressed the above issue by requiring subjects to note when they decided to act in a movement paradigm. Subjects observed an oscilloscope screen on which a spatial clock was represented. While their EEG and EMG data were recorded, the subjects noted the "time" according to the oscilloscope clock at which they "decided" to move. Their RPs began approximately 350 milliseconds prior to their subjective "decision." Thus Libet et al. conclude that volitional activity is initiated non-consciously (Libet et al., 1983). [See fig. 1]

If the findings of Libet and co-workers are replicated, then we must conclude that "decisions" [End Page 83] to act arise prior to our conscious awareness of them. Thus our "decision" or "freedom" is illusory (if by these terms we mean conscious phenomena). It occurs after activity has been initiated. Libet et al. (1983) wish to retain free will, however, when they suggest that the subject may still have the power of veto over arising intentions. Thus an act may be planned, and initiated out of consciousness, but might still be rejected when entering the latter. Could this be so? I will suggest that in the light of Libet's own work this is unlikely to yield "true" freedom.

When the somatosensory cortex is directly stimulated, at liminal intensity, stimulus train durations of about 500ms are necessary to elicit any reportable conscious sensation. Stimuli below that intensity or train duration produce no sensation at all (Libet, 1993 for review). Similarly an afferent sensory stimulus may be altered (by direct cerebral stimulation) for up to 500ms prior its subjective "appearance" (Libet 1993). Libet suggests that a period of "neuronal adequacy" of approximately 400-500 ms is required for neuronal activity to yield a conscious sensation. For the purposes of this discussion, it does not matter what duration in time, or magnitude of stimulus, is required for such a sensation to "arise." The important point is that a finite period of neuronal activity precedes conscious awareness of the latter. Neuronal events precede subjective phenomena and hence phenomenology.

Thus, to suggest as Libet does that free will is still sustainable, as a power of veto, is flawed. It is flawed because the "decision" (in consciousness) to act or not to act is itself the "result" of preceding neural activity.

The decision does not make itself. It is a product or correlate of preceding neuronal events.

Similar conclusions have been reached by Velmans in his analysis of the literature concerning human information processing (Velmans 1991). The weight of the evidence derived from [End Page 84] cognitive neuroscience points to sophisticated analysis of both incoming sensory information and ongoing volitional activity "taking place" out of consciousness. "[O]ne becomes aware of a stimulus only after one has analysed and selected it, and aware of one's own response only after one has executed it" (Velmans 1991, 666).

Discussion

The purpose of this paper has been to consider whether the notion of free will is compatible with neuroscience. We returned repeatedly to the question of whether conscious awareness is necessary for such freedom. If we regard it as central to freedom of the will that we have subjective phenomena "choosing" prior to neural activity, then clearly free will is refuted by the evidence provided. The work of Libet and others has the clear implication that the "experience" of free will occurs after the neural activity which initiates action. Thus, subjectivity is placed at one remove from initiation (and the two are therefore potentially separable). The necessarily materialist and reductive evidence presented has demonstrated that human beings may lose experiential control over aspects of their behavior, and that their experience of control may be altered both normally and pathologically. When we consider the neural correlates of phenomena we find that conscious experience is always temporally post neural activity and thus (within our current understanding) "caused" not causal. Our "control" is a sense which may be lost.

Such a conclusion seems to devalue consciousness, our awareness, our "thinking about thinking." Our sense of agency is apparently illusory:

"The human brain generates an internal model to fit incoming information about the external world, and experiences the model rather than the information." (Picton and Stuss 1994)

"Most of us share in common the useful delusion that we have free will. Schizophrenic patients have lost this experience. As a consequence they develop idiosyncratic, and in many cases harmful, delusions instead." (Frith 1987; my italics)

Or again in Bollas (1992):

"The self does not evolve unconsciously. . . . [T]he self is unconsciousness, a particular inner presence, reliably vectored by the forms "it" uses to find expression." (51)

In fact, we are left with more of a conundrum than might appear at first. We have excluded a conscious freedom of the will. We have relegated any freedom to the non-conscious. Are we left with the conclusion that human acts are predicted by neuronal events, that our consciousness is not necessary for what may appear purposeful? (If we retain the notion of "purpose" then we must cede that "purpose" to the non-conscious.)

The identity of the individual becomes more than the conscious "I" or the "ego." It must comprise both conscious and non-conscious.

Einstein's consciousness experiences the "thought" "If a person falls freely he will not feel his own weight."

What is the cause of such a thought? If it is not the concept, or the "reason" that he holds in consciousness at the time, then it is a process occurring out of, and prior to awareness, but with apparent purpose, and a bias towards truth, an "appreciation" of "reason."

As Dennett (1991) and Hoffman and McGlashan (1993) have highlighted, many processes occur in parallel within the nervous system, only some of which appear to be conscious. Some flavor of this multiplicity, this "reporting" on processes and what it might comprise, may be appreciated in the pathological schizophrenic symptom of voices commenting:

"A 41 year old housewife heard a voice coming from the house across the road. The voice went on incessantly in a flat monotone describing everything she was doing, with an admixture of critical comments. 'She is peeling potatoes, got hold of the peeler, she does not want the potato, she is putting it back, because she thinks it has a knobble like a penis, she has a dirty mind, she is peeling potatoes, now she is washing them.'" (Mellor 1970, 16)

If there are multiple processes occurring out of the individual consciousness, then it is possible to conceive of how these might appear persecutory and alien if they acquired/accessed consciousness; becoming the persecutory "bizarre objects" of Bion (1955), or the autonomous, isolated islands [End Page 85] of parallel distributed processes of Hoffman and McGlashan (1993). This point was also made by Vygotsky (1934): "Each one of us carries schizophrenia in a latent form, i.e. in the mechanism of thought which, when uncovered, becomes the central figure in the drama of schizophrenic thought" (1934, 1071).

Conscious awareness, thinking about thinking, however we conceptualize it, is not the initiator of action in the nervous system. Whereas "we" are normally shielded from this realization by our "sense" of volition and freedom, it is clear that such a sense is vulnerable to disturbance. The symptoms arising from the latter in the normal process of creativity, and the passivity phenomena of schizophrenia, comprise the attribution of volition to other forces which (from the materialist perspective) must reside (out of consciousness, but) within the nervous system.

The role of consciousness in volition is not that of initiator. But for any purpose to be realized consciousness seems "necessary." The "purposeful" acts which arise without consciousness seem either futile and ill-conceived (e.g., in somnambulism, fugue and delirium) or mundane and repetitive (as in driving along a familiar route, or performing a simple manual task). As we have seen with the alien hand syndrome, purpose may be apparent but disruptive and "self"-defeating. In schizophrenia, action and thought may be out of control, apparently purposeful, but blunted. Stereotypic acts and recurrent intrusive thoughts impede the patient rather than prompting change. Only in the normal process of creativity do we see such a passive consciousness "receiving" "worthwhile" insight or performing actions with subtlety and invention (and Boden [1992] has highlighted the non-random nature of this process and how its contents relate to the prior contents of consciousness). Consciousness is necessary for purposeful (intentional) action, but it is not the initiator.

Time

Materialist and reductive theories of mind (including the psychoanalytic) have placed great emphasis upon the antecedents of behavior. These have tended to be long term e.g., genetic inheritance and early childhood experience. In the present paper I have emphasized the minutely short term; the neural events which predict behavior. There [End Page 86] exists the possibility of a continuum, with genetics and environmental factors integrated into a neural activity which precipitates action (and, on the same basis, thought). However such an equation is inadequate in several ways. It fails to account for the nature of the generation of creativity and action, and places the issue of their genesis "further back" in terms of both time (preceding conscious intent) and location in "inner space" (generation is non-conscious cf. conscious). The problem has been moved but not solved.

Nor have I accounted for aesthetics or "reason." I have failed to address the future:

"I" may "choose" x or y in the present, not only because a happened to me in the past, but because I would like b rather then c to happen to me in the future.

The "purpose" which we may attribute to non-conscious processes, determining choice, is orientated temporally, towards the future, and thematically, towards food, music or physics as the context "requires." It is also externally orientated; it acts upon the world.

Space and Time in Virtual/Mental Space

My "acting" in time is essentially a spatial and temporal phenomenon. But as well as acting within physical space, "I" act within a mental (virtual) space. I would like to suggest that in this (virtual) space, time and the temporal relationship of events become dimensions.

I am suggesting that in the "normal" state although acts are initiated prior to my awareness of my own decision, the temporal relationship that pertains is such that I experience my decision prior to information regarding the initiated action. There is a delay between the initiation of the act and such initiation reaching my consciousness.

While "correct" timing is maintained, "I" have the impression that my conscious "I" generates acts. This illusion is fostered by the chronology of my awareness (both of "intent" and "resulting" act). While the chronology is in place "I" act externally upon the world (in mental/virtual space).

However, if the chronology of internal initiation of action breaks down, if for pathological reasons, the timing of events is altered, then my (conscious) "intention" to act may appear to lag behind action or to precede it by an unduly long period (figures 4-6). I may then experience my action as imposed (in the former) or impeded (in the latter). Timing and the "direction" of action are thus related. [End Page 87]

Both scenarios may be seen in pathological states above, where in B2 the (alien hand) patient feels the affected limb anticipates his wishes, in C2, C3 and C4 where schizophrenic patients' actions are determined independent of them, and, in C1 where the patient feels an external machine prevents him from thinking. When the chronology of action is altered, the experienced "direction" of action is also changed (in the internal world of the subject). ¹

External Context

I have attempted to explain some aspects of the experience of free will, and of its pathology. With regard to schizophrenic "passivity," I have hypothesized a cause of abnormal experiences, but not of their delusional elaboration (e.g., "cosmic rays" or persecuting machines). ²

I have also not sought to address the place of the individual in external space or environment. If the conscious "I" does not generate action, and if the non-conscious is the generator, then such action may be generated in response to imperatives from the environment (of which consciousness may be unaware, at least temporarily) (and see Velmans 1991). This serves to emphasize the importance of context in volition. If we learn of our own intentions by speaking to others then this emphasizes our communal ecology.

I will not expand upon this question further in this paper except to say that in this respect I have been guilty of a weakness exhibited in much of our neuropsychiatric and psychological literature; namely that I have repeatedly considered individuals divorced from their context, their environmental milieu. Whereas the purpose of this paper has been to question the emphasis upon the conscious (over the non-conscious) in considerations of free will, there is clearly a case to be made for work which questions the emphasis of individual above community in the exercise of healthy mental life.

Concluding Remarks

The reductionist, materialist perspective underlying the current analysis has been "pushed" to its logical conclusion. Consciousness is not the initiator of willed action. Yet the effect of this analysis is to raise further issues which perhaps speak to the limits of reductionism itself as an exhaustive account of human action. We require a new perspective on the relationship between the temporal and spatial distribution of activity within the nervous system and the subjective "mental/virtual space" which is its correlate. With brain imaging techniques we may approach some of these questions. When subjects imagine movements they "activate" motor areas within their brains (Stephan et al, 1995). When schizophrenics "hear" voices it is their speech areas which they "activate" (McGuire et al 1993). But whereas neuroscience is well equipped to measure and characterize such phenomena, it is clearly on less secure ground with respect to "motive," reason, and the "purpose" of our action, our acting in the world.

We repeatedly return to notions of "cause," initiation, and the antecedents of subjective and neural events. We require a new conceptualization of the non-conscious "parts" of ourselves.

My choice, such as it is, occurs out of consciousness, prior to my knowledge of it. My consciousness [End Page 88] appreciates the consequences. Without conscious experience my non-conscious is ill-equipped to "decide." My experience of freedom, unimpeded thought and action, is distinct from, but related to, beliefs both immediate and abstract. Experience or belief may be subject to pathology.

But I know not "why" my experience is as it is, any more than I know why I experience at all. And my "decision" is made before "I" make it.

". . . but a person can only walk one way and that is the way I still walk."

Angela Carter
Review Independent on Sunday,
8th January, 1995.

Acknowledgment

The author wishes to thank the following for their comments upon earlier drafts of this paper: Professor S. R. L. Clark, Professor C. D. Frith, and Professor K. W. M. Fulford. He particularly wishes to express his gratitude to Professor Steven Hirsch with whom he has debated many of the ideas contained within this manuscript over the past 3 years. Sean A. Spence was supported by the Charing Cross Hospital Special Research Trustees.

Sean A. Spence, Research Fellow, MRC Cyclotron Unit, Hammersmith Hospital, Du Cane Road, London W12 ONN; Academic Department of Psychiatry, Charing Cross Hospital, Fulham Palace Road, London W6 8RP.

Notes

1. These hypotheses are partially borne out by the experiment of GreyWalter (1963). In this experiment neurosurgical patients had electrodes implanted in their motor cortices. They were instructed to look at slides on a carousel and to advance the latter by pushing a control button. In fact the latter was a fake and the slides were advanced by the amplified signal originating in the subjects' own motor cortices (as they prepared to press the button). The patients were apparently startled by the effect (Dennett 1991). It seemed to them that the slide projector was anticipating their decisions. Now if, instead of a signal being sent to a slide projector, a neural signal by-passed the usual pathways and initiated action at a temporal variance to the normal state it is possible to imagine that such a phenomenon might give rise to a subjective experience not dissimilar to some of those described above (in particular B2). A subject might indeed "feel" that his movements were being anticipated, or that a limb "had a mind of its own." If his thoughts were affected might he not experience thought insertion?

2. [I have described elsewhere cases in which such elaboration is clearlyrelated to environmental stimuli and prior exposure to such stimuli (Spence 1993; 1994). In the case of Nintendo hallucinations, a stimulus in the environment is clearly incorporated in to the phenomenology of the psychotic process, itself becoming persecutory (Spence 1993); and similarly in female survivors of child sexual abuse suffering from mania and experiencing tactile sexual hallucinations the former abuser in childhood is recreated in hallucinatory form in the acute psychosis (Spence 1994).]

References

Banks, G., P. Short, J. Martinez et al. 1989. The alien hand syndrome: Clinical and post mortem finding. Archives of Neurology 46: 456-459.

Bion, W. R. 1957. Differentiating the psychotic from the non-psychotic personalities. International Journal of Psychoanalysis 38: 266-275.

Boden, M. 1992. The Creative Mind. London: Abacus.

Bollas, C. 1992. Being a Character. London: Routledge.

Brion, S. and C. P. Jedynak. 1972. Troubles du transfert interhemispherique. A propos de trois observations de tumeurs du corps calleux. Le signe de la main etrangere. Revue Neurologique 126: 257-266.

Clark, S. R. L. 1991. How Many Selves Make Me? In Human Beings, edited by D. Cockburn. Cambridge: Cambridge University Press. 213-233.

Clark, S. R. L. 1992. Minds, Memes, and Rhetoric. Inquiry 36: 3-16.

Deecke, L., P. Scheid and H. H. Kornhuber. 1969. Distribution of readiness potential, pre-motion positivity, and motor potential of the human cerebral cortex preceding voluntary finger movements. Experimental Brain Research 7: 158-168.

Dennett, D. C. 1991. Consciousness Explained. New York: Penguin Books.

Dobzhansky, T. 1967. The Biology of Ultimate Concern. New York: The New American Library.

Feinberg, T. E., R. J. Schindler, N. G. Flanagan et al. 1992. Two alien hand syndromes. Neurology 42: 19-24.

Foucault, M. 1981. London Review of Books 21 May-3 June: 5.

Frith, C. D. 1987. The positive and negative symptoms of schizophrenia reflect impairment in the perception and initiation of action. Psychological Medicine 17: 631-648.

Frith, C. D. 1992. Consciousness, information processing and the brain. Journal of Psychopharmacology 6: 436-440.

Frith, C. D. and D. J. Done. 1989. Experiences of alien control in schizophrenia reflect a disorder in the central monitoring of action. Psychological Medicine 19: 359-363.

Frith, U, J. Morton and A. M. Leslie. 1991. The cognitive basis of a biological disorder: Autism. Trends in Neuroscience 14: 433-438.

Gasquoine, P. G. 1993. Alien hand sign. Journal of Clinical and Experimental Neuropsychology 15: 653-667.

Gibson, W. 1993. Burning Chrome. New York: HarperCollins.

Goldberg, G, and K. K. Bloom. 1990. The alien hand sign: Localization, lateralization and recovery. American Journal of Physical Medicine and Rehabilitation 69: 228-238.

Goldberg, G., N. H. Mayer and J. U. Toglia. 1981. Medial frontal cortex and the alien hand sign. Archives of Neurology 38: 683-686.

Grey, W. W. 1963. Presentation to the Osler Society, Oxford University. In Consciousness Explained, by D. C. Dennet: New York: Penguin. 167.

Highfield, R. and P. Carter. 1993. The Private Lives of Albert Einstein. London: Faber and Faber.

Hoeg, P. 1994. Miss Smilla's Feeling For Snow. London: Flamingo.

Hoffman, R. E. and T. H. McGlashan. 1993. Parallel distributed processing and the emergence of schizophrenic symptoms. Schizophrenia Bulletin 19: 119-140.

Hughlings, J. S. 1915. Quoted in Consciousness Explained, by D. C. Dennett: 194.

Jaspers, K. 1963. General Psychopathology, translated by J. Hoenig and M. W. Hamilton. Manchester: Manchester University Press.

Kornhuber, H and L. Deecke. 1965. Hirnpotentialanderungen des Menschen: Bereitschaftspotential und reafferente Potentiale. Pfluegers Arch Gesamte Physiol Menschen Tiere 284: 1-17.

Libet, B. 1993. The neural time factor in conscious and unconscious events. In Experimental and Theoretical Studies of Consciousness. Ciba Foundation Symposium 174. Chichester: Wiley. 123-146.

Libet, B., C. A. Gleason, E. W. Wright et al. 1983. Time of conscious intention to act in relation to onset of cerebral activity: Readiness-Potential. Brain 106: 623-642.

Libet, B., E. W. Wright and C. A. Gleason. 1982. Readiness-potentials preceding unrestricted "spontaneous" vs. pre-planned voluntary acts. Electroencephalography and Clinical Neurophysiology 54: 322-335.

Malenka, R. C., R. W. Angel, B. Hampton et al. 1982 Impaired central error-correcting behavior in schizophrenia. Archives of General Psychiatry 39: 101-107.

Marcus Aurelius. 1964. Meditiations, translated by M. Staniforth. London: Penguin Books.

McGuire, P. K., G. M. S. Shah and R. M. Murray. 1993. Increased blood flow in Broca's area during auditory hallucinations in schizophrenia. Lancet 342: 703-706.

Mellor, C. S. 1970. First rank symptoms of schizophrenia. British Journal of Psychiatry 117: 15-23.

Picton, T. W., and D. T. Stuss. 1994 Neurobiology of conscious experience. Current Opinions in Neurobiology 4: 256-265.

Schneider, K. 1974. Primary and secondary symptoms in schizophrenia. In Themes and Variations in European Psychiatry, edited by S. R. Hirsch and M. Shepherd. Bristol: John Wright. 40-46.

Sims, A. 1988 Symptoms in the Mind: An Introduction to Descriptive Psychopathology. London: Bailliere Tindall.

Spence, S. A. 1993. Nintendo hallucinations: a new phenomenological entity. Irish Journal of Psychological Medicine 10: 98-99.

Spence, S 1994 Tactile sexual hallucinations occuring in mania in women. Psychiatry in Practice 13, no. 4: 20-21.

Stephan, K. M., G. R. Fink, M. C. Ridding et al. 1995. Motor imagery and motor performance: Electrophysiological characteristics and functional anatomy in precentral gyrus. Journal of Cerebral Blood Flow and Metabolism 15: S43.

Timsit, M. 1970. The Kornhuber and Deecke phenomenon in schizophrenia and borderline cases. Electroencephalography and Clinical Neurophysiology 29: 535.

Trimble, M. R. 1990. First-rank symptoms of Schneider: A new perspective? British Journal of Psychiatry 156: 195-200.

Trojano, L, C. Crisci, B. Lanzillo et al. 1993. How many alien hand syndromes? Follow up of a case. Neurology 43: 2710-2712.

Urmson, J. O. 1991. Freedom of the will. In The Concise Encyclopedia of Western Philosophy and Philosophers, edited by J. O. Urmson and J Ree. London: Routledge.

Velmans, M. 1991. Is human information processing conscious? Behavioural and Brain Sciences 14: 651-726.

Vygotsky, L. S. 1934. Thought in schizophrenia. Archives of Neurology and Psychiatry 31: 1062-1077.

Wilkes, K. V. 1988.-, yishi, duh, um, and consciousness. In Consciousness in Contemporary Science, edited by A. J. Marcel and E. Bisiach. Oxford: Oxford University Press. 16-41.