Earlier in the month I responded to author Michael Prescott’s critical comments on my critique of Chris Carter’s defense of empirical arguments for life after death. After publishing my response on his blog, his readers have offered various counterpoints. I told Prescott that I would be happy to respond to “highlights” of his readers’ posts. So here’s my response to the selections he emailed me earlier in the week. Since I don’t provide much background to the various comments, I’d recommend that interested subscribers to my blog first read the comments section in Prescott’s blog where Prescott’s readers responded to my arguments. – M.S.
I’ve read the selected comments you forwarded to me from your readers. Here are some responses.
First, some of your readers brought up near-death experiences (NDEs), which I have not directly addressed at this stage. So here’s a preview. On my view, NDEs provide prima facie justification for belief in survival for those who have such experiences, and the testimonial data from such experiences may provide an interesting argument against some materialist philosophies of mind. But I don’t think there’s a good argument for survival from the testimonial data. In fact, I’d say that arguments for survival based solely on the data from NDEs are the second weakest kind of survival argument, the weakest being arguments from apparitional experiences. In addition to the widely advertised problems with NDE survival arguments, all such arguments will run into the problem of auxiliary hypotheses. I’ll discuss this in more detail in my book.
Second, as far as the alleged consistency of NDEs goes, it’s not clear to me what precisely your readers think this actually proves, shows, or otherwise establishes vis-à-vis the survival hypothesis or my critique. For reasons I’ll note here, I think the appeal to the consistency of NDEs has limited value within the larger landscape of the survival debate.
(i) As is well known from the analysis of religious experience in the philosophy of religion, it’s relatively easy to find consistent/inconsistent features in different experiences when the experience-type has vague parameters. Given the elastic parameters of NDEs, the acceptance of the survivalist interpretation of NDEs based on their consistency is just as unwarranted as the rejection of the survivalist interpretation of NDEs on the grounds of their alleged inconsistency. Proponents and opponents are equally held captive to naïve ways of conceptualizing the situation.
(ii) Even if we grant that the descriptions of NDEs are consistent and exhibit various non-trivial coherence relations, what follows? It’s unclear how this shows a “hole” in my argument. My central argument, even applied to NDEs, is entirely compatible with NDEs exhibiting coherence. It’s also compatible with the consistency of NDEs contributing to the evidential value of NDEs. However, until we formulate an argument for survival that is informed by the issues in evidence assessment I’ve raised, we don’t really know the net value of consistency. This is just another example of survivalists thinking that the demands of serious argument are met by claiming that survival is true because they’ve provided a statement of their subjective degree of confidence in unclear or contentious principles.
Third, with respect to Rouge’s comments about auxiliary hypotheses, I think he’s confused either about what independent testability involves or about how it’s applied in the sciences.
(i) Contrary to Rouge’s suggestion, the auxiliary hypotheses required to test different evolutionary hypotheses actually are independently testable in the relevant sense. See Elliott Sober’s Evidence and Evolution (chapters 3-4), where this is demonstrated, for example with reference to common ancestry and phylogenetic relationships.
But let’s be clear about what independent testability involves. Your readers seem to be operating with some highly inflated and/or idiosyncratic conception. Roughly stated, for a hypothesis H (proposed to explain observation O) to be independently testable means there’s a procedure that produces a justification for h that does not depend on our being antecedently justified in accepting H, not H, or O. I’ve provided many examples in my publications showing how this condition is widely satisfied in the sciences and in a variety of everyday applications. (I direct your readers once again to my “Getting Sober about Survival” blog series where I discussed these issues). So survival arguments fail to secure an epistemic virtue that is widely exemplified across different disciplines and modes of inquiry. In the light of this, the survivalist appeal to “consistency” looks at best like “last prize.” This tends to reinforce suspicions about survival arguments rather than rescue them from skeptical objections.
(ii) There’s no doubt that the fossil record by itself, though incompatible with certain theistic-creation hypotheses, is nonetheless compatible with a range of alternative hypotheses of the sort Rouge outlined. But this strikes me as utterly insignificant. A single piece of evidence at the scene of a crime may eliminate one suspect but still leave us with three possible suspects. This is why it’s important to locate “discriminatory evidence,” that is, observations that are to be expected given one hypothesis but not another. And, as I’ve argued, an essential aspect of such a program is locating independently testable auxiliary hypotheses in arriving at predictive consequences for both one’s preferred hypothesis and whatever hypothesis is the competitor. Again, I refer readers to Sober’s discussion (in Evidence and Evolution) of how a hypothesis is to be tested against a competitor.
Fourth, Rouge appeals to the transmission theory of consciousness, apparently to show, contrary to what I’ve argued, that some conceivable survival scenarios are more to be expected than others if consciousness survives death. This is at least an interesting suggestion.
There’s some initial confusion in Rouge’s argument, for he begins by saying:
I would argue that any version of the transmission theory is compatible with the persistence of memory, intentions, skills, and personality, and that the transmission theory in some form is by far the most likely model of the mind-brain relationship.
But, of course, the issue is not whether the transmission theory is compatible with the persistence of memories, etc. (of course it is), but rather—the stronger notion—whether these are to be expected. Rouge then switches to the stronger conception:
On the basis of the transmission theory, certain afterlife-related outcomes would be predicted to occur – not invariably, given the individual variations that are natural in any study of human consciousness, but at least in some cases. We would expect some dying patients to show heightened lucidity as consciousness begins to slough off the damaged brain – and there are cases of “terminal lucidity,” vivid and veridical deathbed visions, and NDEs in which thought and perception are heightened far beyond ordinary experience. We would expect mental confusion attributable to a damaged brain to clear up in a postmortem state, and mediumistic communications provide support for this. We would expect the deceased to retain their memories and even to experience them more vividly, and again this is consistent with mediumship, past-life studies, and NDEs (the life review). So I would suggest that, while testable predictions in this area are inevitably less certain than those in (say) chemistry or physics, the transmission theory does provide us with some predictions, and these predictions have tended to pan out.
It looks like Rouge wants to treat the transmission theory of consciousness as an auxiliary hypothesis for the purposes of developing a space of plausible survival worlds from among a larger array of merely conceivable survival worlds. So if consciousness survives death (the survival hypothesis) and the transmission theory (auxiliary hypothesis) were true, then certain afterlife-related outcomes would be predicted to occur, well, at least in some cases. Since the transmission theory has been independently tested (with success, according to Rouge), we have a survival-friendly auxiliary hypothesis for which there is independent evidence but which leads us to expect the relevant data.
This is the most interesting suggestion from among the various comments, but ultimately it’s not plausible.
(i) Where T = the transmission theory and O = any of the observational data (noted by Rouge), let’s assume that the value of Pr(O | T)—the probability of O given T—is well defined. The relevant range of data for survival arguments is considerably broader than O. Survival arguments require that the Pr(D | S) has a well-defined value, where D = the broader range of data and S = the survival hypothesis. A well-defined value for Pr(D | S) requires the kinds of auxiliary hypotheses I’ve outlined in detail, but it’s not possible to derive these auxiliaries from T. So even if Pr(O | T) is well-defined, this would be insufficient to extricate survival arguments from the problem of auxiliary hypotheses.
(ii) However, as it turns out, the value of Pr(O | T) is actually not well-defined. Rouge doesn’t actually show why T should lead us to expect O, but this is precisely what needs to be argued. And this is particularly important because Rouge has hedged the prediction with an extremely important qualifier, namely in some cases. So why should T lead us to expect O, yet only in some cases? Which cases exactly? What are the even approximate parameters here? And is this a consequence of the content of T, or T + something extra? Rouge nonchalantly refers to “individual variations that are natural in any study of human consciousness,” but this is not to be lightly passed over. Until Rouge can answer these questions, there’s no workable model here at all, and certainly no challenge to my claim that the auxiliary hypotheses required by survival arguments are not independently testable.
On the face of it, Rouge’s assertion of alleged predictive derivations strikes me as more retrofitting. He’s simply transferred this from the survival hypothesis to the transmission theory. His qualifier is quite convenient, too convenient. It allows easy confirmation but makes difficult, if not impossible, falsification. For example, it allows us to treat verified memory claims as evidence for the theory, but not their absence as evidence against the theory. If either survival or the transmission theory makes genuine predictions, I should like to know what observations we should expect if the theory is true but not if the theory is false.
(iii) As it happens, transmission theorists have taken different views concerning the degree of psychological continuity there would be between ante-mortem and postmortem consciousness, to what extent unique personality features would carry over, what causal powers would be attributed to surviving “selves,” and so forth. Understandably so. Simply proposing that the brain “transmits” consciousness rather than produces it is compatible with a broad range of survival scenarios. In fact, the language can be interpreted in terms of multiple models of consciousness. And this shows again that the value of Pr(O | T) is simply not well-defined even among those who advocate transmission theories. Indeed, on some views, if we survive death, we should not expect our ordinary personality to survive. See Tart, Charles. 1990. “Who Survives? Implications of Modern Consciousness Research.” In What Survives? Contemporary Explorations of Life after Death, ed. Gary Doore. Los Angeles, CA: Jeremy P. Tarcher, Inc., 138–151.
Fifth, several of your readers are under the impression that disposing of materialist views of mind/reality somehow renders survival more plausible. But that’s too quick in point of logic. At best, disposing of materialism removes an objection to some hypotheses of survival, but removing an objection to a hypothesis is not the same thing as providing evidence for a hypothesis. And at all events, whether materialism is true or not (or whatever metaphysical theory survivalists wish to advocate) is irrelevant to the problem of auxiliary hypotheses.
Sixth, your readers didn’t properly understand my references to Broad, Price, and Ducasse. My point there was simply that they, unlike other writers, were cognizant of there being many different conceivable survival hypotheses with varying predictive consequences. That’s not an endorsement of any of their particular flirtations at this juncture. Anyhow, any attempt to refute Broad’s “persistence hypothesis” or Price’s “place memory hypothesis” (as alternatives to personal survival) will run right into the problem of auxiliary hypotheses and get caught in the net I’ve cast into the survival debate.
Finally, I’ll leave you with a quote from Elliott Sober that sums up the plight of empirical survival arguments on my view.
The lazy way to test a hypothesis H is to focus on one of its possible competitors H0, claim that the data refute H0, and then declare that H is the only hypothesis left standing. This is an attractive strategy if you are fond of the hypothesis H and are unable to say what testable predictions H makes. (Sober, Evidence and Evolution, p. 353).
This is basically the strategy of argument in most books that try to present empirical evidence for survival, Carter’s included. They are just so many variations on lazy testing.