by Benjamin Libet

--The Determinism and Freedom Philosophy Website --

This is Benjamin Libet's response to a critical paper by Ted Honderich about neuroscientific research by Libet and colleagues and the philosophical use made of the research. This was by Libet himself and also by Karl Popper and J. C. Eccles in their book The Self and Its Brain. Honderich's critical paper, to which you can turn, is
Is the Mind Ahead of the Brain? -- Benjamin Libet's Evidence Examined. For Honderich's rejoinder to Libet's response below, see in turn Is the Mind Ahead of the Brain? -- Rejoinder to Benjamin Libet. The research and interpretation under discussion, having to do with sensations, was followed by more Libet research of a different kind, about actions. For a brisker paper that discusses both periods of research, there is Honderich's Is the Mind Ahead of the Brain? Behind It?, from the book On Determinism and Freedom (Edinburgh University Press, 2005).

: Honderich claims that our "delay-and-antedating" hypothesis, of a delay in cerebral production combined with a subjective antedating of a conscious sensory experience, involves self-contradiction, which may cast doubt on some of our experimental findings and on the hypothesis. This claim misses the distinction between the phenomenological subjective mental content of an experience and the physical-neuronal configuration that elicits the experience; also, it cannot explain the experimentally observed discrepancy, between subjective timing and the empirically delayed time for cerebral adequacy for eliciting the experience, found when stimulating a subcortical sensory pathway. Honderich usefully distinguishes between our stated (delay-and-antedating) hypothesis and a different though unac-ceptable one which would have serious implications for mind-brain theories. The delay-and-antedating hypothesis does not provide a formally definitive contradiction of monist-identity theory (of the mind-brain relationship). However, our experimentally based hypothesis does dissoci-ate subjective/mental timing from the actual physical/neuronal time of an experience. This phenomenon, though conceptually strange, must be encompassed by any mind-brain theory.

     Honderich reviewed our experimental findings and analyzed our hypothesis on the relation of subjective timing of a conscious sensory experience to its neuronal production (Libet et al., 1979). The hypothesis states: (a) Delays of up to 0-5 sec or more occur before achieving cerebral "neuronal adequacy" for eliciting the sensory experience. (b) However, there is a subjective referral or antedating of the experience back to a time close to the early initial signal, arriving in 10-20 msec via the fast specific projection pathway and represented by the primary evoked potential at the sensory cortex; subjectively, then, there would appear to be no delay of the experience (Libet et al., 1979; Libet, 1978, 1981, 1982). This is referred to by Honderich as the delay-and-antedating hypothesis.
      Honderich claims that, although we clearly preferred the delay-and-antedating hypothesis, a second hypothesis was implied by our language and work, a so-called "no delay" hypothesis. In the latter, there would be the same delay of up to 0-5 sec for neuronal production of the experience, but the actual experience appears before then, with no delay. He suggests that our failure to distinguish the two ditlerent hypotheses may cast doubt on some of the experimental findings and led us to an unwarranted view of certain consequences of the work for mind-brain theories. I shall attempt to clarify my statements and views on these issues.
      Additionally, Honderich argues that our delay-and-antedating hypothesis is itself doubtful and perhaps unacceptable. I shall aim to show that his argument is based on a misapprehension of the significance of some of the experimental findings as well as on a debatable philosophical assumption he makes about the relation between subjective phenomena and the corre-lated physical, cerebral events.

1. The Experimental Discrepancy between Timings

     First let us be clear about what the experiments actually tell us. Subjects reported the relative timing order of two sensory experiences, i.e. reporting which one was felt first. The times of delivering the two stimuli and the required minimum durations of each stimulus for eliciting its sensory experience were known to the observer, but not to the subject. From these (and accessory) data, inferences were made about subjective timing of a sensory experience relative to the minimum times required to produce the experience (i.e. to achieve a state of neuronal adequacy for eliciting it) (Libet, 1973, 1981, 1982; Libet et al, 1979). In discussing and diagramming the data we tended to place the subjective experience at the times reported by the subject, these reports being the only available directly valid indicators of such times. Strictly speaking, the reported time represented the subject's recalled impression of when the experience occurred, relative to another sensory experience, and not necessarily the time of actual appearance of the experience.* Although this qualification is clearly implicit in our treat-ment, indeed it forms the basis of the delay-and-antedating hypothesis, it was not always explicit in our presentation. I believe this potential ambiguity of meaning in our operational usage of experimentally determined terms led to some of Honderich's perceived difficulties in our work.
      Honderich (1984) provides a reasonable listing of our findings, but his interpretation of their significance is incomplete and partially misguided, particularly for his subset # 2 of our findings. In our finding listed as # 2.1 by him, the experience of a skin stimulus was reported by the subject to begin well before the experience for a cortical stimulus, when both stimuli were started simultaneously; the cortical stimulus empirically required about 0.5 sec of repetition to produce any sensation, while the adequate skin stimulus was a single brief pulse. This experiment was done to test our previous hypothesis of "simple delay", namely that awareness of all sensory inputs would require a delay similar to that for the cortical stimulus. Finding 2.1 was indeed a crucially initial one in that it appeared to contradict our simple delay hypothesis. But we had previously obtained strong though indirect evidence that a delay, of up to 0.5 sec to achieve neuronal adequacy for eliciting the sensory experience, did occur even for a skin stimulus (listed in Libet, 1973, 1978, 1981 ; Libet et al. 1967, 1972, 1979, but not detailed here). This paradox led us to propose the delay-and-antedating hypothesis; retaining the requirement of a neuronal delay, the early subjec-tive timing of the skin sensation was now explained as a subjective referral or antedating of the actually delayed experience backwards to the time of the initial fast ascending signal.
      The crucial experimental test of the delay-and-antedating hypothesis lay in matching the time order of experiences for a skin pulse and a subcortical stimulus (in the medial lemniscus pathway, LM); these are listed as findings 2.2-2.4 by Honderich, who missed their full significance. The LM stimulus, like a cortical one, requires the same substantial period of repetitive pulses to become effective; with intensity adjusted so that at least 0.2 sec of repetitive pulses (at 20-60 per sec) was required to elicit any sensation, we knew that the cerebral production of the sensory experience had to be delayed for at least that 0.2 sec. However, unlike the cortical stimulus, each LM stimulus pulse in the train produces the same fast projection response (the primary evoked potential) at sensory cortex that is elicited by a skin stimulus, thus providing the putative early signal needed for subjective antedating of the experience. The experimental test, which could potentially have falsified the delay-and-antedating hypothesis, in fact confirmed it. When the onset of the LM stimulus train was simultaneous with a skin stimulus pulse, subjects reported that both sensory experiences began simul-taneously (Libet et al., 1979). That is, the LM-elicited sensation was experienced with no delay relative to that for a single skin stimulus pulse, even though the LM stimulus could not have become adequate until 0.2 sec had elapsed; an LM stimulus train shorter than 0-2 sec elicited no sensation whatsoever! This extraordinary finding, taken together with the different result when matching onset of a cortical stimulus with a skin pulse (finding 2.1), provides conclusive experimental proof of a discrepancy between the subject's timing of his experience and the delayed time at which the LM stimulus became adequate. It seemed warranted to assume that a similar discrepancy exists for skin stimulus, in view of the supporting indirect evidence; the alternative possibility, that the delay-and-antedating hypothesis applies specifically and only to the subcortical LM stimulus, is unattractive and unnecessary.
      In view of the foregoing, Honderich's argument (near the end of his section 1), that finding 2.1 is itself either false or may make other finding; entailing the delay-and-antedating hypothesis false, is based on a faultv premise. Finding 2.1 does not provide crucial evidence for a no-delav hypothesis; it can be satisfactorily and even more attractively encompassed (with the added indirect evidence) by the delay-and-antedating hypothesis. It should be emphasized that the discrepancy between subjective and neuronal timings, directly observed when testing with LM stimuli, derives from experimental observations which have not been challenged on tech-nical grounds; it cannot be refuted as a finding simply by some differing theoretical argument or concept.

2. Subjective Content and Neuronal Configuration

     The distinction, between the phenomenological subjective content of an experience and the observable physical feature.s of the neuronal activities that give rise to or are associated with the experience, is a general one for all mind-brain relationships. Discrepancy between the spatial features of a sensory experience and the spatial pattern of the neuronal activity eliciting the experience has long been recognized and spoken of as "subjective referral in space". For example, in a visual experience the image is not only projected into the external environment, not located at its neuronal site of origin in the brain, but the subjective configuration of the image differs drastically from the underlying neuronal patterns. There is in addition the neurologically well-known phenomenon of subjective "filling in" of the missing portion of a blind area in the visual field.
     Our findings in support of the delay-and-antedating hypothesis led to an analogous concept in the temporal dimension. In temporal referral the sensory experience is antedated or subjectively referred back to a time preceding the delayed time at which neuronal adequacy is achieved; in spatial referral, analogously, the spatial form of a somatosensory experience, for example, is subjectively referred to a part of the body and in a pattern quite different from the spatial configuration of neuronal activity in the cerebral cortex. Our further evidence supported the view that the same ascending pathway to sensory cortex, the specific projection system (lemnis-cal for the somatosensory modality), provides the signal utilized for both types of referral. Honderich dismisses this clear analogy solely on the basis of his own assumption that the discrepancy between subjective and neuronal timings, in temporal referral, involves simultaneous, self-contradictory impressions (see below).
      Indeed, we have no way of predicting a priori, from a knowledge of the neural basis of an experience, what the content or nature of a subjective experience will be, even if one assumes the relationship between the "phys-ical" (externally observable) and the "mental" (internally observable) is one of psychophysical lawful correlation (see Libet, 1982; Nagel, 1974). The experience can only be described in an introspective report by the subject. Honderich appears to violate this principle when he insists that he knows, a priori, what must be the subject's impression of the timing of an experience.
      Honderich insists it is "surely true that the actual having of any conscious sensory experience -- is accompanied by the belief or impression that the experience-is happening now." He claims that our postulated "antedating or referring back (of the experience in time) involves imputing something like simultaneously-held, fully explicit self-contradictory beliefs to subjects...". But Honderich appears to confuse the content of an experience with the actual time of its appearance. We do not regard the subject as being unaware of his experience at the actual time it appears, but we accept the subject's report of his impression of when he felt the experience appeared, and we attempt to account for the discrepancy. At whatever time an experience occurs, the subject's impression of when it appeared could depend on other modulating factors. In our hypothesis, the early, primary cortical response, evoked by the fastest arriving neural component initiated by the sensory input, serves as the signal that leads to a temporal modulation of the subject's impression of when the experience occurred; the localized pattern of these same responses incidentally also serves to modulate the subject's spatial impression. With the stimulus located on somatosensory cortex where it does not elicit this special primary evoked response, the subject reports a delayed time of experience that does in fact coincide with the delay required for the stimulus to become effective.
      Contrary to Honderich's belief, the subject is not forced by the hypothesis to have two contradictory impressions or beliefs about when he felt the experience; he would only have one impression of the time he felt the experience. Operationally, the subject in fact reports an awareness of only one sensory experience with a single time onset. The subject has no aware-ness of any discrepancy between his subjective impression, of when he had the experience, and the actual time of the experience (when neuronal adequacy for producing it occurs). To the extent that some kind of neuronal retention of the early ascending signal may be involved, such a memory process is completely unconscious and not accessible to conscious aware-ness, unlike ordinary memories of events or experiences. The same uncon-scious automaticity applies to subjective referral in space; we are all, as "subjects", completely unaware of the discrepancy between our neuronal spatial patterns and the subjective image they produce. Interestingly, however, the relationship between image and neuronal pattern can be altered
      Honderich's apparent demand for an "account of the supposed phenomenon" of subjective referral backwards in time, or a description of "what processes" mediate the unconscious and automatic referral, is some-what strange. One can only describe relationships between subjective phenomena and neural events, not how one gets from one to the other. The critical components of the neural correlates in the delay-and-antedating hypothesis are, in fact, as or more complete than is the case for most mind-brain correlations.
      Finally, subjective referrals in space or time are fundamentally different from sensory illusions. In illusions there are distortions of the real stimulus configurations, whereas subjective referrals serve to "correct" the experience so as to be closer to the actual time or form of the sensory stimulus than is the neuronal timing or spatial form (Libet, 1981). Honderich appears to prefer the suggestion by Donald MacKay that subjective antedating may be due to an illusory judgment made by the subject when he reports the timing (discussed by Libet et al., 1979). But, as already noted (Libet et al., 1979) such a proposal entails a serious difficulty, aside from arguments about simplicity of hypotheses. "By retaining delays for the immediate subjective sensory experiences, when they initially and actually occur, this alternative suggestion of MacKay's becomes unable to explain the absence of subjective "jitter" or asynchrony in our experience, when a variety of peripheral sensory stimuli are applied synchronously. At least one factor that should produce differences in the delays for achieving neuronal adequacy with different stimuli, is the strength of the stimulus. (See dis-cussion in Libet et al., (1979) and Libet, (1981).) One attractive feature of our (delay-and-antedating) hypothesis is in fact its ability to deal with this difficulty. Subjective referrals, that are retroactive to the early primary evoked response to each sensory input, would make irrelevant any differen-ces among the timings for neuronal adequacy in a group of synchronously initiated inputs; delays for the primary evoked potential are short (10 to 20 msec), and the differences produced by differing intensities of peripheral somatic stimuli are known to be so small as to be negligible for the purpose of subjective timing" (Libet et al., 1979, p. 220).

3. Import for Mind-Brain Theories

     Honderich (1984) rightly points out that a no-delay hypothesis would flatly threaten identity theory and psychophysical lawlike correlation, but that such a hypothesis "must be false, in that a mental item occurs before the physical item on whose later existence it depends". We did not espouse such a hypothesis, as Honderich concedes, but some of my language in the earlier papers led him to a conjecture that we were unwittingly adopting this hypothesis as a basis for our conclusions about the mind-brain relation. I had previously stated that our findings of a discrepancy between subjective timing of an experience and the time of its neural production (i.e. the delay-and-antedating hypothesis) "is not any denial of correspondence between mental and physical events, but (is) rather (a description of) the way in which the correspondence is actually manifested" (Libet, 1981, p. 195); and that "a discrepancy between the `mental' (i.e. subjective timing of an experience in this case) and the `physical' in the temporal dimension can be regarded, in a manner analogous to that for the discrepancy in the spatial dimension, as not contradicting the theory of psychophysical parallelism or correspondence" (Libet et al. 1979, pp. 221-222).
     I did add that such a "temporal discrepancy creates relative difficulties for identity theory, but that these are not insurmountable" (Libet, 1981, p. l96). Honderich disagrees with this, stating that our delay-and-antedating hypothesis presents "no problem whatever for an identity theory" (his section 4). The kind of difficulties I referred to is perhaps exhibited in part by Honderich himself, in his expression that the subjective antedating of a sensory experience "may indeed be of a strange self-contradictory kind". Although the delay-and-antedating hypothesis does not separate the actual time of the experience from its time of neuronal production, it does eliminate the necessity for simultaneity between the subjective timing of the experience and the actual clock-time of the experience. But we find Honderich insisting that the opposite is true, namely that there must surely be a simultaneity between the subject's impression of an experience and the actual time of the experience!
     Many will concede that an absence of isomorphism, say for a visual image relative to its corresponding neuronal pattern, is acceptable to and compatible with identity theory. On a strictly formal basis the same may be said for the temporal discrepancy, as expressed in the delay-and-ante-dating hypothesis in which simultaneity of the actual experience and its neuronal counterpart is preserved; this is also Honderich's position. To the extent that my previous statements may be taken by Honderich and others to imply a different position, I would agree that such an implication should be removed.
      However, the phenomenon of subjective referral of a sensory experience backwards in time is indeed a strange and not intuitively obvious concept (Libet et al. 1979; Libet, 1981) (although, as I argue above, it is not self-contradictory). In it, subjective mental timings of experiences become dissociated from actual clock-times of the experiences. It shows that subjec-tive phenomena "can play tricks with time", as Eccles aptly put it (Popper & Eccles, 1977, p. 364), although admittedly these tricks are accomplished unconsciously and in relation to specifiable neural information. To the extent that adherents of a monist-identity theory may be unconfortable with such subjective "trickery", they could have informal subjective (rather than formal objective) difficulties in accepting the antedating phenomenon, even though there is a solid experimental basis for the temporal discrepancy in the timings.
The above paper appeared in the Journal of Theoretical Biology (1985) 114, 563-570. I am very grateful to Professor John Searle of the Department of Philosophy, University of California, Berkeley, for helpful comments on an earlier draft of the manuscript, though responsibility for all stated positions is mine.


HONDERICH, T. (1984). J. theor. Biol. 110, 115.
LIBET, B. (1973). In: Handbook of Sensorv Physiology. (Iggo, A. ed.). Vol. 2, pp. 743-790. Berlin: Springer-Verlag.
LIBET, B. (1978). In: Cerebral Correlates of Conscious Experience (Buser, P. A. & Rougeul--Buser, A. eds). pp. 69-82. Amsterdam: Elsevier.
LIBET, B. (1981). Phil. Sci. 48, 182.
LIBET, B. (1982). Human Neurobiol. 1, 235.
LIBET, B. (1985). Behav. Brain Sci.
LIBET, B., ALBERTS, W. W., WRIGHT, F. W. & FEINSTEIN, B. (1967). Science 158, 1597.
LIBET, B., ALBERTS, W. W., WRIGHT, E. W. & FEINSTEIN, B. (1972). In: Neurophysiology Studied in Man (Somjen, G. G. ed.). pp. 156-168. Amsterdam: Excerpta Medica.
LIBET, B., WRIGHT, E. W. Jr., FEINSTEIN, B. & PEARL, D. K. (1979). Brain 102, 191.
NAGEL, T. (1974) Phil. Rev. 83, 435.
POPPER, K. R. & ECCLES, J. C. (1977). The Self and Its Brain. Berlin: Springer-Verlag.
* The question of the relationship between the reported introspective timing and the actual time of a subjective experience, especially for operational experimental purposes, has been considered at greater length elsewhere (Libet, 1981, 1985).
For Ted Honderich's rejoinder to this article, go to Is the Mind Ahead of the Brain? -- Rejoinder to Benjamin Libet. For a more readable criticism of both the Libet research considered here and also some different later research, go to Is the Mind Ahead of the Brain? Behind It?

HOME to Det & Free website front page
HOME to T.H. website front page