Cup of Nirvana Philosophical and Contemplative Explorations

What’s Wrong with Survival Literature?

newcoverThere continues to be a plethora of articles and books published on near-death experiences, children who claim to remember past lives, apparitional experiences, and mediums who deliver ostensible messages from the deceased. And here I refer not to the trumped up, obviously exaggerated if not fabricated stories perpetuated through the tabloids and “ghost hunting” television programs, but to phenomena studied in academic settings such as the Division of Perceptual Studies at the University of Virginia, Medical School, and research institutes such as the Windbridge Institute in Tucson, Arizona.   The kinds of ostensibly paranormal phenomena studied by empirical researchers are often adduced as evidence for life after death. Arguments historically purporting to show this I designate “classical empirical arguments” for postmortem survival. While the arguments may differ in various respects, they have a common generic structure: certain data are said to be evidence for survival because the hypothesis of survival allegedly provides the best explanation of the data. The explanatory argument for survival is found in most of the published books and articles favorable to survival as far back as Richard Hodgson’s late nineteenth-century articles on the evidence for survival based on the trance mediumship of Mrs. Leonora Piper.

For several years now I’ve expressed a general disappointment in how survivalists have developed and presented the empirical case for survival. In my previous blog, I said that the nub of my critique is the inadequacy of survivalist arguments purporting to show that the survival hypothesis is the best explanation of the relevant data. I argue that this inadequacy is rooted in the widespread failure of survivalists to acknowledge, much less critically engage, the large number of auxiliary assumptions that must be enlisted for the survival hypothesis to do explanatory work. Survivalists are unsuccessful at showing that the survival hypothesis actually explains anything largely because of their suppression of required auxiliary assumptions. Furthermore, this suppression creates the additional illusion that survivalists have successfully “ruled out” rival hypotheses, such as the living-agent psi hypothesis that purports to explain the data in terms of psychic functioning (extra-sensory perception and psychokinesis) in living persons. Since classical empirical arguments for survival depend on the survival hypothesis explaining the data better than various proposed counter-explanations, the suppression of auxiliary assumptions perpetuates the illusion that survivalists have shown that the survival hypothesis is the best explanation of the data.

Anticipating a central theme in my forthcoming Philosophical Critique of Empirical Arguments for Postmortem Survival (Palgrave Macmillan, 2015), in this blog I outline a few of the widespread conceptual problems that plague survival literature, and I show how they each converge on the suppression of auxiliary assumptions.

1. The Generic Explanatory Survival Argument

To appreciate the conceptual failures in survival literature, it’s important to have a general idea of what the argument for survival is supposed to look like. The argument survivalists make or (as is more often the case) wish to make is an explanatory argument. The basic idea is that the phenomena of near-death experiences, alleged communications from the deceased through mediums, claims to past-life memories (and correlated phenomena), or some such ostensibly paranormal phenomenon provide data that constitute evidence for survival because survival explains these phenomena. More precisely, the phenomena are evidence for survival because the survival hypothesis is the best explanation of the data.

Where “E” ranges over some (narrow or broad) strand of data from paranormal phenomena, explanatory survival arguments are, structurally or formally speaking, two-tier arguments consisting of an explanatory inference and an evidential inference.

1. There is some evidence E.

2. The survival hypothesis S explains E.

3. No other competing hypothesis C explains E (as well as S explains E).


4. S is the best explanation of E.


5. E is evidence for S.

Let’s call this the “generic explanatory survivalist argument” (hereinafter, GESA). GESA is a two-tier argument: (1) through (4) is a standard form of an inference to the best explanation, and (4) to (5) constitute an evidential inference – explanatory power converts to evidential cash value. This is the kind of argument you’ll find implicitly or explicitly adopted in the literature, for example in Almeder (1992), Braude (2003), Carter (2012), Fontana (2005), Gauld (1982), Griffin (1997), Lund (2009), Paterson (1996), Stevenson (1974), and Tucker (2005, 2013).

As a generic argument, there are two important more specific issues GESA does not address but which are essential to actual empirical survival arguments.

First, nothing is said above about how good E is as evidence for S, that is, the degree of evidential support E offers for S. And here survivalists differ. Some contend that E increases or raises the probability of S. Others take the view that E is evidence favoring S over some the rival hypotheses C, and thus E makes S more probable than C. Both of these views are, of course, compatible with S having a very low net plausibility or probability. Still other survivalists take a stronger position and claim that E renders S very probable, or at least probable to degree N, where N > ½, and so survival is at least more probable than not. This latter view is particularly prominent among survivalists who maintain that the empirical arguments provide enough evidence to rationally justify belief in survival.

Second, nothing is said in GESA about explanatory criteria. So GESA is silent on what would be required for S (or some other hypothesis C) to explain E or to be the best explanation of E. Typically, explanatory criteria at least include S’s leading us to expect E, or S’s better leading us to expect E than does C (“predictive power” in the broad sense, or “accommodation” to evidence). Some survivalists, however, wish to roll in other qualities such as simplicity and the need for independent support. I’ll comment on these below.

GESA allows a fairly succinct statement of my main criticism of classical empirical survival arguments. As I explained in my previous blog, the survivalist who sports GESA (or some specific version of it) faces something of a dilemma. He can effectively argue in favor of premise (2), but only if he explicitly incorporates a range of auxiliary assumptions about the nature and character of the afterlife. Alternatively, and this is what survivalists typically do, he can effectively argue for premise (3), but only if he shelves the assumptions required to be justified in affirming premise (2). Consequently, what the survivalist cannot (consistently) do is effectively argue for both (2) and (3). In this way, the survivalist loses his reasons (and hence justification) for affirming (4) – the survival hypothesis is the best explanation of the data.

But let’s see how survivalist strategies of argument conveniently mask this dilemma.

2. “Lazy Testing” – Evading the Burden of Explanatory Reasoning

Speaking in the context of arguments that purport to show the existence of an intelligent designer of the universe, philosopher of science Elliott Sober has identified a form of explanatory reasoning that he aptly calls “Lazy Testing”:

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 but are unable to say what testable predictions H makes. (Sober 2008: 353)

Sober’s “lazy testing” diagnosis is quite appropriate as a way of characterizing a widespread pattern of reasoning within pro-survival literature. Most survivalists allege, often with great emphasis, that the survival hypothesis is a testable hypothesis, and yet in the literature this claim is given short shrift and we’re left wondering quite rightly whether it’s true at all. What we typically find in the relevant literature is an accumulation of data, consisting largely of testimonial claims and descriptions of the methods/conditions of their verification, all recounted with an impressive narrative. Survival is then something of a sotto voce inference – merely asserted to be the best explanation because all other known explanations fail. One can randomly select a work on survival and it’s likely you’ll find this structure of argument (e.g., Almeder 1992, Carter 2012, Fontana 2008, Tucker 2005, 2013). But as Sober has noted, the alleged vices of rival hypotheses do not confer virtue on one’s preferred hypothesis, but this is all that one would have to hang hope on in the absence of one’s preferred hypothesis having any virtue of its own.

Hence, one problem in the literature is that survivalists fail to show the explanatory virtue of the survival hypothesis itself, but this is masked by the nearly exclusive emphasis on how other hypotheses allegedly fail to account for the relevant evidence. The thing to see here is that arguing in support of premise (3) of GESA is entirely legitimate, and so we should expect survivalists to attempt to debunk the alleged explanatory virtues of rival hypotheses. The crux of the issue, though, is whether (i) survivalists present a positive case for the explanatory virtues of the survival hypothesis (hence offer support for premise (2) of GESA), and whether (ii) the justification for affirming premise (2) involves reasons that are compatible with the reasons offered as a justification for affirming premise (3). Otherwise put, we need to evaluate the “ruling out” of rival hypotheses (in premise (3) of GESA) in the light of what has been established with respect to the explanatory virtues of the survival hypothesis itself (premise (2) of GESA).

The survivalist counter-response is easily anticipated. The survivalist will claim that he does support premise (2), for he points to the evidence being what we would expect if survival were true, the simplicity of the survival hypothesis, the survival hypothesis being falsifiable, and the survival hypothesis being independently supported – each alleged explanatory virtues. Yes. There is no doubt that survivalists make such claims, or proffer such considerations, but the claims either lack adequate development or grounding, or they involve a logical sleight of hand that is masked by the emphasis on considerations used to rule out competitors.

So let’s look with greater scrutiny at survivalist dialectical maneuvers with respect to each of the aforementioned presumed explanatory virtues: survival leading us to expect the data, the alleged simplicity of the survival hypothesis, falsifiability, and independent support.

3. “Suppressed Assumptions” – The Perils of Predictive Power

I invite the reader to peruse Richard Hodgson’s famous “Further Record of Observations of Certain Trance Phenomena” (1897-98). Hodgson says repeatedly that the data provided by trance mediumship are exactly what we would expect if the survival hypothesis were true (and not what we would expect given rival hypotheses), a claim that quickly became a staple of survivalist explanatory claims and so may be found in countless pro-survival books and articles. Yet neither Hodgson nor his survivalist descendants have shown that their quasi-predictive claims are true. They merely assume that such claims are true because they make a large number of assumptions about what surviving persons would be like in the afterlife. If they were to try to show that these predictive claims are true, it would be evident, as E.R. Dodds (1934) later argued, that the survival hypothesis is not a single or simple hypothesis, but a “hydra-headed” hypothesis involving various collateral assumptions about, for example, the powers, knowledge, intentions, and character of survivors.

Therefore, the survivalist is justified in affirming premise (2) of GESA only if the survivalist is justified in a large number of additional assumptions about the afterlife, e.g. if persons were to survive death, they would have the requisite powers and intentions to communicate with the living, could efficaciously exercise such powers, and would have sufficient continuity of memory and character as to be identifiable by living persons as some specific formerly living person. And to show that we’re justified in accepting premise (2) would require showing that we’re justified in accepting the assortment of auxiliary assumptions. In other words, the survival hypothesis only leads us to expect the data if it is what I call a robust survival hypothesis, a simple survival hypothesis (affirming the survival of the self or our individual consciousness) supplemented with additional assumptions.

However, the introduction of a robust survival hypothesis raises the difficult question of the epistemic status of the required auxiliary assumptions. Are we justified in accepting such claims? At all events, what’s transparently clear is that survivalists have not shown that we are justified in accepting such claims, nor even that the survivalist is so justified. And to this extent, the survivalist has not shown that anyone is justified in accepting premise (2) of GESA. Otherwise put, survivalists have not shown that the survival hypothesis explains anything because they’ve not acknowledged, much less independently supported, the range of assumptions without which survival explains nothing.

Now a crucial point to note here, though, is that the suppression of auxiliary assumptions also infects showing that we’re justified in accepting premise (3), and in two ways.

First, in order to justifiably maintain premises (2) and (3), the survivalist’s justification for the survival-friendly auxiliaries (required for the survival hypothesis to have predictive power over the relevant data) must exclude our being justified in rival auxiliary assumptions. These rival auxiliaries would include (i) rival auxiliary assumptions about the afterlife such that if they were true, we would not expect the relevant evidence and (ii) rival auxiliaries that when conjoined with rival non-survival hypotheses would lead us to expect the relevant evidence. The second is particularly important because one way in which survivalists have tried to rule out rival hypotheses is by treating those rival hypotheses in their most stripped down form and then (correctly) arguing that they cannot accommodate the evidence. This is a frequent strategy found in survivalist dismissals of appeals to extra-sensory perception among the living to account for the apparently extraordinary knowledge mediums possess or that young children who claim to remember past lives possess. But this is a logical sleight of hands. Neither a stripped down survival hypothesis nor a stripped down appeal to living-agent telepathy or clairvoyance will lead us to expect the relevant data. Each candidate explanation must be taken in a fairly robust form, and in the case of living-agent psi must be combined with various psychological assumptions (e.g. concerning motivations, the range and capacities associated with dissociative phenomena, rare cognitive skills). So the question is whether there is some robust form of the appeal to living-agent psi that will lead us to expect the data at least as well as some robust survival hypothesis.

Second, it’s equally clear that, when survivalists wish to dismiss appeals to living-agent psi on the grounds that the data are contrary to what such a hypothesis would lead us to expect, survivalists are (perhaps unwittingly) working with their own “bulked up” version of the living-agent psi hypothesis. They adopt various assumptions about the scope, potency, and refinement of psi, and how it might interact with dissociative phenomena and the sudden manifestation of unusual skills. Stephen Braude (2003) has addressed in some detail this problematic feature of survivalist attempts to rule out counter-explanations in terms of living-agent psi. Surprisingly, or perhaps not surprisingly, none of the major pro-survival books since Braude (2003) have as much as addressed Braude’s challenge, and this includes Fontana (2005), Lund (2009), Carter (2012), and Tucker (2005, 2013). What Braude has rightly noted is how suppressed assumptions about psi are operative in survivalist efforts to rule out the appeal to living-agent psi. What’s crucial in the debate, then, is how these assumptions stack up against alternative sets of assumptions that produce a robust living-agent psi hypothesis that does indeed lead us to expect the evidence. Since Braude has himself proposed such a robust living-agent psi hypothesis, it would be incumbent upon survivalists to rule out this robust living-agent psi hypothesis. To date they have failed to do this.

So I’ve been arguing above that survivalists prematurely pop the celebratory cork of having ruled out rival hypotheses by treating those hypotheses in either very simple versions that, like a simple survival hypothesis, radically underdetermine the evidence, or by treating them in a narrow band of robust forms that poorly accommodate the evidence but by virtue of questionable survivalist assumptions. What the literature has failed to produce is a thoroughgoing engagement with robust rival hypotheses that do accommodate the evidence in a way that is comparable to how a robust survival hypothesis may accommodate the evidence. Of course, a precondition of any such evaluation would an acknowledgement of survivalist auxiliary assumptions that permit the survival hypothesis to accommodate the evidence. And so we see that the suppression of auxiliaries is highly salient to both showing that the survival hypothesis explains the data and that it does so in a way superior to competitors.

4. “Suppressed Assumptions” – Simplicity, Falsifiability, and Independent Support

But the suppression of auxiliaries infects survivalist attempts to rule out competitors in another way. Survivalists often support premise (3) in GESA by arguing that the survival hypothesis is simpler than competitors, for example, simpler than an appeal to a living-agent psi hypothesis, which allegedly must be stretched into a “super-psi” hypothesis that requires living persons to accomplish extraordinary psychic feats, e.g. telepathically or clairvoyantly mining and integrating information from multiple sources.

However, when survivalists appeal to the relative simplicity of the survival hypothesis in this context, without exception they are referring to a hypothesis of survival sans auxiliary assumptions, and this simple survival hypothesis is then compared to rival hypotheses in their most robust forms, bulked-up with various auxiliary assumptions. A great example of this is the survivalist discontent with appeals to living-agent psi, which survivalists contend can only explain crucial strands of data by being amped up into a “super-psi” hypothesis, an appeal to living-agent psi supplemented with various auxiliary assumptions that permit psi to have a potency and refinement beyond what has been independently established to exist. That a survival hypothesis (without auxiliaries) is simpler than a robust rival hypothesis is a red herring. What’s at issue is the simplicity of the hypotheses in their mutually robust forms because it’s only in these forms that they would have a claim to predictive power as a central explanatory virtue.

This impacts the interrelated issues of falsifiability and independent support/testability as well. Any hypothesis can be made falsifiable by conjoining it to the right assumptions, so neither the survival hypothesis nor rival hypotheses are prevented from securing this apparent explanatory virtue. My hypothesis of an invisible gardener in my yard is falsifiable given, for example, the added assumption that he attracts blonde women between the ages 21 and 34. What is crucial is that our auxiliary assumptions, without which hypotheses make no predictions, be independently testable. What survivalists must show is that the survival hypothesis is, unlike competitors, genuinely falsifiable since its assumptions are independently testable. They have not done this, but unless one acknowledges the role that auxiliary assumptions play, this requirement is easily bypassed. And it becomes easy to target rival hypotheses as failing to secure a virtue, which not even the survival hypothesis can secure. Consequently, when survivalists claim that the survival hypothesis is falsifiable and independently testable, but appeals to “super-psi” are not, it’s important to see that a simple survival hypothesis is not falsifiable and a robust survival hypothesis requires auxiliary assumptions that are no more independently testable than those employed by rival theories, indeed probably less so.

To the extent, then, that simplicity, falsifiability, and independent testability are considered explanatory virtues, it’s clear that survivalist arguments create an illusion that they successfully rule out rival hypotheses. This is masked by the suppression of auxiliary assumptions, for once these assumptions are introduced the simplicity, independent testability, and falsifiability of rival hypotheses – including robust living-agent psi – are at no more of a disadvantage than the survival hypothesis itself. Or at any rate, until such time as survivalists treat their preferred hypothesis in its appropriate robust form, we have no good reason to accept the survivalist contention that premise (3) of GESA is true. And so we lose our reasons for supposing that the survival hypothesis is the best explanation of the data.

5. Concluding Thoughts

To sum up: the survivalist suppression of auxiliary assumptions creates the illusion that the survivalist has shown that premises (2) and (3) in GESA are true. More precisely stated, it creates a twofold illusion. First, it creates the illusion that the survival hypothesis explains the data because rival hypotheses apparently don’t explain the data. Second, it creates the illusion that rival hypotheses – such as the living-agent psi hypothesis – don’t explain the data (as well as survival) because they lack some virtue the survival hypothesis is presumed to possess, but which, as I’ve argued, dissolves upon more careful scrutiny,

The survivalist suppression of auxiliary assumptions is a fallacy that infects survivalist literature. It’s widespread in its presence, it’s far reaching in its implications for the assessment of the survival hypothesis. The fallacious nature of the suppression of auxiliary assumptions may be put in more systematic terms as follows:

  • It permits survivalists to create an appearance of explanatory virtue for the survival hypothesis by facilitating an exclusive focus on how poorly alternative hypotheses fare. But this lazy testing simply evades the burden of showing that survival explains anything at all.
  • It creates the illusion that survivalists have ruled out the appeal to living-agent psi as a rival hypothesis with at least equal explanatory power. This is an illusion because:
    1. The process of “bulking up” hypotheses (generating robustness) can easily accommodate evidence, and this principles holds equally for the living-agent psi hypothesis and the survival hypothesis, each of which in their suitably robust forms can equally account for the relevant evidence.
    2. Survivalists have not shown that a robust survival hypothesis is simpler than a robust living-agent psi hypothesis (that accounts for the data), and so they have not shown that the relevant kind of survival hypothesis has any advantage over living-agent psi alternatives at this juncture.
    3. Survivalist auxiliary assumptions are not independently testable, and so if the plausibility of a robust living-agent psi hypothesis (that accounts for the evidence) is reduced for this reason, the same applies mutatis mutandis to a robust survival hypothesis (that accounts for the evidence). Thus, the survival hypothesis has no advantage here.
    4. A robust living-agent psi hypothesis is no less (trivially or non-trivially) falsifiable than a robust survival hypothesis, given the right sort of auxiliary assumptions, so the survival hypothesis has no advantage at this juncture.

Hence, for the above reasons I maintain that survivalist fails to show that anyone is justified in accepting premises (2) and (3) of GESA. It follows that we have a defeater for the conclusion (4). So, we should not accept the survivalist claim, on the basis of (2) and (3), that the survival hypothesis is the best explanation of the data. And this defeat will transfer to other explanatory arguments for survival that are specific instances of GESA, which make use of the explanatory criteria discussed above.

In my forthcoming book, I propose a formalization of the classical arguments that drops the reference to explanatory power and unpacks the arguments purely in terms of (Bayesian and Likelihood) confirmation measures. I argue that these arguments also fail, which suggests that the problems associated with auxiliary assumptions are not limited to explanatory survival arguments but apply more broadly to empirical arguments for survival, at least those based on paranormal phenomena.


Works Referenced or Cited

Almeder, R. (1992). Death and Personal Survival. Lanham, MD: Rowman & Littlefield.

Braude, S. (2003). Immortal Remains: the Evidence for Life after Death. Lanham, MD: Rowman & Littlefield.

Carter, C. (2012). Science and the Afterlife Experience: Evidence for the Immortality of Consciousness. Rochester, VT: Inner Traditions.

Dodds, E.R. (1934). “Why I Do Not Believe in Survival.” Proceedings of the Society of Psychical Research 42: 147–72.

Fontana, D. (2005). Is There an Afterlife? A Comprehensive Overview of the Evidence. Blue Ridge Summit, PA: O Books.

Gauld, A. (1982). Mediumship and Survival: A Century of Investigations. London: William Heinemann Ltd.

Griffin, D.R. (1997). Parapsychology, Philosophy, and Spirituality: A Postmodern Exploration. Albany: State University of New York Press.

Lund, D.H. (2009). Persons, Souls, and Death: A Philosophical Investigation of an Afterlife. Jefferson, NC: McFarland.

Paterson, R. (1995). Philosophy and the Belief in a Life after Death. New York: Saint Martin’s Press.

Sober, E. (2008). Evidence and Evolution: The Logic Behind the Science. Cambridge: Cambridge University Press.

Stevenson, I. (1974). Twenty Cases Suggestive of Reincarnation, 2nd Ed. Charlottesville, VA: University of Virginia Press.

Tucker, J. (2005). Life Before Life: Children’s Memories of Past Lives. New York: Saint Martin’s Griffin.

Tucker, J. (2013). Return to Life: Extraordinary Cases of Children Who Remember Past Lives. New York: St. Martins Press.

Comments are closed.

Post Navigation