|IS THE MIND AHEAD
THE BRAIN? -- REJOINDER TO BENJAMIN LIBET
by Ted Honderich
To what I thought was my unanswerable 'Is the Mind Ahead of the Brain -- Benjamin Libet's Evidence', Professor Libet did indeed make an answer, published in the next issue of The Journal of Theoretical Biology. This is my rejoinder to it. Was it a little too emollient? Its original title, by the way, was 'Mind, Brain and Time: Rejoinder to Libet'.
Libet (1978, 1981) and Libet et al (1979) report an impressive body of experimental findings and also offer a certain interpretation of those findings, which interpretation pertains to the problem of the mind-brain relation. One central experimental sequence, as Libet understands it, can be summarized as follows.
A stimulus is delivered by an inserted electrode directly to the cortex of a subject, beginning at T. It takes about 0.5 sec after the beginning of the stimulation-train before there occurs "neuronal adequacy", a sufficient physical condition, for the associated conscious sensation - there is this 0.5 sec delay before neuronal adequacy is achieved. A peripheral stimulus, say to the skin of the hand, is given later than the beginning of the direct cortical stimulation, say at T+ 0.2 or T+ 0.3 sec. This stimulus, like the direct cortical stimulation, takes about 0.5 sec to produce neuronal adequacy for its associated conscious sensation. The peripheral stimulus, by way of a specific projection pathway, very quickly produces - in about only 15 msec - a certain primary cortical response, regarded by Libet as a "time-marker". There is no such primary response with the direct cortical stimulation.
The subject is asked in what order he or she had the two conscious experiences, one of the earlier direct cortical stimulation, or one of the later peripheral stimulus. The subject reports, paradoxically, that the conscious experience of the later skin stimulus came before the conscious experience for the earlier direct cortical stimulation.
As for interpretation of these findings, it consists at bottom in a certain hypothesis. The hypothesis, whatever it is, in turn gives rise to various conclusions as to the relation of mind and brain, the mental and the neural. Some of the conclusions are to the effect that the hypothesis at least casts doubt on any Identity Theory of mind and brain (e.g. Davidson), and also the theory of lawlike or nomic connection between neural and mental events (Honderich, 1981 a,b). The hypothesis is used by others, without the dissent of Libet, also as evidence for a particular dualistic theory of mind and brain - that of the "Self-Conscious Mind", which can "play tricks with time" (Popper & Eceles, 1977, p. 364).
In my critical paper (Honderich, 1984) I made five claims, the first having two parts. Here I will consider what Libet has to say against them or of them in his reply (Libet, 1985). All page references in what follows, except when otherwise indicated, are to this reply.
1. What is the Hypothesis?
It was my contention that Libet and his colleagues, followed by Eccles and Popper, in fact stated but did not distinguish two inconsistent hypotheses. The first, named by me the "delay-and-antedating" hypothesis, is to the effect that the subject’s conscious experience of the peripheral stimulus - in the experimental sequence outlined above - actually occurs at the time when neuronal adequacy for it is achieved, which is to say at about T+ 0.7 or T+ 0.8 sec, but that it is somehow at that time taken by the subject to have occurred much earlier. That is, the experience is delayed the requisite 0.5 sec but is somehow "antedated". It is "antedated" to the time of the "time-marker" provided by the primary cortical response. That there is no such time-marker for the direct cortical stimulation fits in with this. The second hypothesis, named by me the "no-delay" hypothesis, is to the effect that the subject’s conscious experience of the peripheral stimulus actually occurs at the earlier time. It is not delayed.
It is important to one of my further claims (section 3 below) that Libet did in fact state and therefore have in mind both hypotheses. This he tends to overlook in his reply, where he tends to claim that only the delay-and-antedating hypothesis was stated by him before. He writes in his reply, for example, that it was merely the case that the delay-and-antedating hypothesis was "not always explicit" in his presentation, and that there was no more than a "potential ambiguity of meaning in our operational usage of experimentally determined terms", which potential ambiguity "led to some of Honderich’s perceived difficulties in our work" (p. 564; cf. p. 569).
I do not accept this as at all a correct account. As reported in my original paper (Honderich, 1984, p. 120), Libet wrote explicitly of the conscious experience of the peripheral stimulus that it was "actually before" the experience of the cortical stimulus. This is not an inexplicit statement of the delay-and-antedating hypothesis, but an explicit statement of the no-delay hypothesis. Secondly, he explicitly stated that the claim that neuronal adequacy for the experience of the peripheral stimulus is achieved only at about T+ 0.7 or T+ 0.8 sec is not to be taken as a claim or implication that the experience itself occurs so late as that (Honderich, 1984, p. 121). These are but two of seven passages or diagrams offered by me in support of the claim that Libet in fact stated the no-delay as well as the delay-and-antedating hypothesis, and did not distinguish them. I will not rehearse the other five items here, but invite readers to turn back to my paper.
I note, to my surprise, that this fundamental uncertainty persists in Libet’s reply to me. He sometimes states or speaks of the delay-and-antedating hypothesis perfectly clearly. "...the delay-and-antedating hypothesis does not separate the actual time of the experience from its time of neuronal production..." (p. 569). However, he writes in another place: "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." (p. 564). If he accepts the reports as valid, as truly giving the times, then presumably he must also accept the no-delay hypothesis.
He also writes, secondly: "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." (p. 567). If they accept the subject’s report of his impression (i.e. not his memory but his impression in his experience of its time), they accept the no-delay hypothesis.
Thirdly, Libet persists in contrasting his preferred hypothesis with one described as "retaining delays for the immediate subjective sensory experience" (p. 568). If the preferred hypothesis does not retain delays, it is the no-delay hypothesis.
I note, further, that while it is granted that I have "usefully" distinguished the two hypotheses (p. 563), it may be suggested elsewhere in Libet’s reply that I failed to get straight the preferred hypothesis. This is perhaps implied by the suggestion that I do not see the "significance" of the experimental findings (p. 564) and by this line: "First let us be clear about what the experiments actually tell us" (p. 564). He also says my interpretation of the significance of his findings is incomplete and partially misguided (p. 565).
There are in fact very considerable obscurities about the preferred hypothesis, and I will return to this matter below. What is to be noted now is that Libet does not show (or indeed attempt to show) that I misunderstood anything that was made clear with respect to that hypothesis. Nor does he establish, although I did not state all of it, that I failed to grasp all the evidence for it (pp. 565-6). However, I shall not contest this matter.
As it seems to me, then, Libet has not fully succeeded in part of his stated intention in his reply, which was to clarify his statements on the given issues (p. 564). The situation is not that what we have is two expressions of a single hypothesis, and that only a too-meticulous philosopher would distinguish between them. For one thing, as was pointed out in my original paper, and as will be noted again below (section 3), the two hypotheses have radically different consequences for something with which Libet is concerned, theories of the mind-brain relation.
The delay-and-antedating and no-delay hypotheses, very evidently, cannot both be true. To this, in my paper, I added: "That conclusion carries corollaries, of course. One is that if the delay-and-antedating hypothesis is true, then if there are putative findings which entail the no-delay hypothesis, those findings are false. The like corollary has to do with taking the no-delay hypothesis to be true. Any putative findings which entail the delay-and-antedating hypothesis are then false. It is not within my competence, however, to examine these findings in detail ..."(Honderich, 1984, p. 123). Later, after questioning the sense or coherence of the delay-and-antedating hypothesis, I say: "It is my... tentative conclusion here that the delay-and-antedating hypothesis is open to objection along these lines. However, it is not within my competence to judge the findings which are put in question, or whose interpretation is put in question, if the delay-and-antedating hypothesis is rejected." (Honderich, 1984, p. 125).
I do of course stick absolutely to this small point of logic. If an hypothesis is false, then, if other propositions entail it, they must be false. If the subject’s experience (of the peripheral stimulus) does occur at about T+ 0.7 or T+ 0.8 sec, then, if some proposition entails that the experience occurs much earlier, that proposition is false. A related although different point, although here we get into methodological complexities, is to be made of any proposition which falls short of logically entailing the hypothesis but does constitute good evidence for it.
Libet, although he gets me right in one place (p. 564), seems elsewhere to mistake this small and impregnable point for something else which, whatever else is to be said of it, is not impregnable at all. He writes: "Honderich claims that our ‘delay-and-antedating’ hypothesis ... involves self-contradictions which may cast doubt on some of our experimental findings" (p. 563). As will be set out again in the next section. I do indeed take the delay-and-antedating hypothesis to involve a lack of sense or coherence, to involve a self-contradiction. However, I did not anywhere claim that it follows from this that any entailers of it, or pieces of good evidence for it, are false.
Libet subsequently gives a curious description of my passage (Honderich, 1984, p. 123) quoted at the beginning of this section (p. 566). He describes the passage as containing a faulty premise. It is the faulty premise that the fact that the subject reports as he does (that his conscious experience of the later peripheral stimulus came before the conscious experience for the earlier direct cortical stimulation) is crucial evidence for the no-delay hypothesis. I nowhere in the passage or elsewhere make use of or mention such a premise. As the passage makes plain, I have no need of it.
I am now inclined to think, as Libet does not, that his experimental findings are in no close logical or evidential relation with either hypothesis. The experimental findings by themselves leave open those two hypotheses and also a third possibility. It is to the effect that the subject has a conscious experience of the peripheral stimulus at about T+ 0.7 or T+ 0.8 sec, which experience is perfectly ordinary and involves no "antedating", but that the subject in later reporting the time of his experience judges it to have been earlier than it was. He mistakenly takes it to have occurred prior to the experience of the cortical stimulation - to have occurred about 15 msec after the peripheral stimulus (Honderich, 1984, p. 119; Libet, 1985, p. 568).
To repeat, it now does not seem to me that the experimental findings entail, or, to speak quickly, are something like conclusive evidence for, any hypothesis in the field as distinct from the others. Hence the logical point is of no utility. This, of course, is not Libet’s reaction to the logical point. He is persuaded that the findings single out the delay-and-antedating hypothesis, but he does not establish this. Nor, I think, can he do so.
2. Is the Delay-and-Antedating Hypothesis Acceptable?
The delay-and-antedating hypothesis, I said in my paper, seems to come in a first part to this: that a subject has a conscious sensory experience at about T+ 0.7 or T+ 0.8 sec, which experience itself includes or involves or is accompanied by the belief, idea, impression, content or whatever that the experience is not then happening. That is, the experience somehow carries the belief or whatever that the experience is not present, but rather is past. Secondly, and differently, the experience carries some idea to the effect that it is earlier than that which in fact was earlier than it: the experience for the direct cortical stimulation. I went on to say that it is surely the case, of any actual experience had by a subject, that it is then present. To have an experience is to take it as present. Further, I suggested, more tentatively, of the very slightly later of two experiences, that the second carries an idea that it is indeed later than the other. If all this is so, I said, the delay-and-antedating hypothesis involves imputing to the subject something very like two pairs of explicit self-contradictory beliefs, the two beliefs being held simultaneously. Such an imputation, I said, was an unpersuasive ad hoc assumption.
There are obscurities and complexities in connection with conscious experience generally, as is plain enough, and in connection with time (Honderich, 1977), but Libel says nothing in his reply which refutes the above claims.
His first response is that these claims miss "the distinction between the phenomenological subjective mental content of an experience and the physical-neuronal configuration that elicits the experience" (p. 563; cf. p. 566). That very surprising and indeed stunning comment, that I do not distinguish the mental from the physical, or that my claims somehow elide the two, is presumably somewhat better expressed later. It is said:
"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. .. . 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 impression of the time he felt the experience" (p. 567).
The second sentence of this passage contains the statement noted on p. 369 above, which seems to commit Libet to what he also rejects, the no-delay hypothesis. If we put this aside, and look at the passage in terms of the delay-and-antedating hypothesis, it seems the charge it lays is that I mistakenly suppose that a subject, in having an experience, knows its actual time, or at what time it occurs relative to the neural configuration for it.
I suppose nothing of the sort. Libet is right in asserting, as I take him to, that the subject does not have to have the belief that his experience is simultaneous with the neural configuration for it. But I do not suppose and nor did I say that the subject has such a belief. I do suppose, however unaware of or correct or mistaken a subject may be about the clock-time of an experience or the time of an experience relative to the neutral configuration for it, that he takes it to be present. The analysis of this thought of his involves great difficulty, but one thing is clear: it is inconsistent with the thought that the experience is past. But the thought that it is past is exactly the thought that seems to be assigned to the experience by Libet’s account. So far as I can see, no amount of declamation can avoid the conclusion that the delay-and-antedating hypothesis assigns to a subject a pair of fully-explicit, simultaneously-held, contradictory beliefs, ideas, impressions, contents, or whatever.
Libet writes later: "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!" (p. 569).
On the contrary, to repeat, I did not and do not insist that the subject gets the clock-time right. I do insist that in having an experience one takes it to be present.
The conclusion that the given account of antedating assigns to a subject a pair of fully-explicit, simultaneously-held, contradictory beliefs is also resisted by Libet in other ways that seem distinct. The conclusion is, he says, based on "a debatable philosophical assumption...about the relation between subjective phenomena and the correlated physical, cerebral events" (p. 564). There are, so far as I can see, two possibilities as to what my supposed mistaken assumption is.
First, Libet speaks of my dismissing something "solely on the basis of [my own] assumption that the discrepancy between subjective and neuronal timings, in temporal referral, involves simultaneous, self-contradictory impressions" (p. 567). My reply is that this is not properly called an assumption, but rather an analysis of what is said of the phenomenon or experience of antedating. No alternative analysis or account of it is offered by Libet, who merely speaks of it as a "conceptually strange" and a "strange and not intuitively obvious" phenomenon (p. 563, p. 570). Rather, he mistakenly takes my demand for an alternative account as a demand for something quite different (p. 568).
The second possibility as to my mistaken assumption is that I assume that we can predict a priori, from a knowledge of the neural basis of an experience, what the content or nature of a subjective experience will be (p. 567). He writes: "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" (p. 567).
My analysis of what is said of the antedating experience, on the contrary, has nothing whatever to do with deducing or in any way concluding, from a neural premise, that an experience must be such-and-such. It is true that I accept and defend a hypothesis of nomic or lawlike connection between the neural and the mental (Honderich, 1981a,b). That hypothesis plays no part whatever in my analysis of what is said of the supposed antedating experience.
Libet seeks to make his extraordinary supposition seem ordinary by drawing an analogy between the supposed antedating and two other phenomena, but there is no relevant analogy whatever. First, we need assign nothing remotely approximate to self-contradictory beliefs to a subject in connection with the phenomenon that "neuronal spatial patterns" are of a different spatial configuration from "the subjective image they produce" (p. 566-7, p. 568). The same is true with the phenomenon that "in a visual experience the image is... projected into the external environment, not located at its neuronal site of origin in the brain" (p. 566). That each of "antedating" and the two plain phenomena involve a distinction between the mental and the neuronal, as Libet rightly says they do, goes no way at all to producing a relevant analogy.
Libet has one other objection to the claim that he is involved in assigning self-contradictory beliefs to subjects. The claim, he says, "cannot explain the experimentally observed discrepancy, between subjective timing and the empirically delayed time for cerebral adequacy for eliciting the experience..." (p. 563). Of course the claim does not explain the experimental findings. It makes no attempt to do so. It is an objection to an attempted explanation, not a substitute. It is none the worse at what it attempts for not attempting something else.
Finally here, Libet writes that "the discrepancy between subjective and neural timings...derives from experimental observations which have not been challenged on technical grounds; it cannot be refuted as a finding simply by some differing theoretical argument or concept" (p. 566; cf. p. 570).
What derives from the findings, which is described as "the discrepancy between subjective and neural timings", must of course be taken as the delay-and-antedating hypothesis. It can indeed be rejected on the basis of a "theoretical argument or concept", if one of those is the consideration that any hypothesis is unacceptable because it makes an intolerable assumption about subjects having explicit, simultaneous, self-contradictory beliefs. Libet himself elsewhere accepts the possibility of this kind of rejection, for which reason he attempts to maintain, as we have seen, that his hypothesis does not make this assumption.
3. Explanation of Mind-Brain Conclusions
Libet has taken several different views in the past of the consequences of his hypothesis for the problem of the relation between mind and brain, mental and neural events. An earlier one that he omits in his reply was the view that the hypothesis refutes or an Identity Theory or the theory that mental and neural events stand in nomic connection. He wrote: .... . on the face of it, an apparent lack of synchrony between the ‘mental’ and the ‘physical’ would appear to provide an experimentally-based argument against ‘identity theory’ ...... a temporal dissociation between the mental and the physical events would further strain the concept of psychophysiological parallelism or, if one prefers, of co-occurrence." (Libet, 1978, p. 80). Eccles & Popper take a similar view (e.g. 1977, p. 364).
Libet does not explicitly dispute but presumably does not accept my idea that such consequences have been drawn from the delay-and-antedating hypothesis by confusing it with the no-delay hypothesis, which he does agree would have the given consequences. I persist in the idea. It would be less persuasive if it were true, as I have shown Libet tends to suggest in his reply, that the no-delay hypothesis was not explicitly stated in the relevant articles. As I have maintained above, it is. Libet in his reply has more to say, which I consider in the next section here, of the connection between the delay-and-antedating hypothesis and mind-brain theories.
4. Failure of Mind—Brain Conclusions
The delay-and-antedating hypothesis, unlike the no-delay hypothesis, does not put a mental event at a different time from the neural event or whatever with which it is associated. Hence the delay-and-antedating hypothesis does not threaten either an Identity Theory or the theory of the mind which I favour. Libet accepts this, I take it. He writes, at any rate, that the hypothesis "does not provide a formally definitive contradiction of monist-identity theory". (p. 563; cf. pp. 569).
However, he pregnantly adds that the delay-and-antedating hypothesis "must be encompassed by a mind-brain theory" (p. 563). He also writes that "subjective referral of a sensory experience backwards in time is indeed a strange and not intuitively obvious concept ... ... subjective phenomena ‘can play tricks with time’… …To the extent that adherents of a monist-identity theory may be uncomfortable with such subjective ‘trickery’, they could have informal subjective (rather than formal objective) difficulties in accepting the antedating phenomenon..." (p. 570).
My own attitude, firmer now than before, is that the delay-and-antedating hypothesis is sufficiently suspect to be no great challenge to any theory of mind and brain. However, what is certain is that it is no greater challenge to an Identity Theory or the different theory I favour than to any other. Its mystery may consort better with essentially mysterious views of the mind, such as that of Eccles & Popper, but that can hardly be a significant argument for those theories.
5. Failure of No-Delay Hypothesis
The no-delay hypothesis puts a conscious experience at a different time from the associated neural configuration or events. If it were true, it would thus refute both an Identity Theory of mental and neural events, and the theory which makes those events simultaneous nomic correlates. But the no-delay hypothesis, as Libet agrees, is false. It is self-contradictory in that it asserts that a conscious experience occurs before there occurs a sufficient physical condition for it. As remarked in my original paper, Identity Theories seem to me open to philosophical refutation, and the theory of "the Self-Conscious Mind" open to both philosophical and neuroscientific refutation. The theory of nomic or lawlike psychophysical connection seems to escape philosophical objection and to have increasing confirmation from neuroscience. These claims have been defended elsewhere (Honderich, 1981a,b; Honderich, 1986).
Davidson, D. (1980). Essays on Actions and Events. Oxford: Clarendon.
Honderich, T. (1977). In: Time and the Philosophies (Ricoeur, P. ed.). pp. 141-154. Paris: UNESCO.
Honderich, T. (1981a). Inquiry 24, 3, 277.
Honderich, T. (1981b). Inquiry 24, 4, 419.
Honderich, T. (1984). J. theor. Biol. 110, 115.
Honderich, T. (1986). In: Mind Matters (Blakemore, C. & Greenfield, S. eds). Oxford: Blackwell.
Libet, B. (1978). In: Cerebra! Correlates of Conscious Experience. (Buser, P. & Rougeul-Buser, A. eds). pp. 69-82. Amsterdam: Elsevier.
Libet, B. (1981). Phil. Sci. 48, 182.
Libet, B. (1985). J. theor. Biol. 114, 563.
Libet, B., Wright, E. W., Feinstein, B. & Pearl, D. K. (1979). Brain 102, 193.
Popper, K. R. & Eccles, J. C.(1977). The Self and Its Brain. Berlin: Springer.
I am most grateful to Professor J. Z. Young for improving an earlier draft of this paper.
Journal of Theoretical Biology (1986) 118, 367-375
For a brisker account of Libet's
research than the one above, and also of later research by him, go to Is the Mind Ahead of the Brain? Behind It?